JOBS WORKING PAPER Issue No. 71 Towards A Just Coal Transition Labor Market Challenges And People’s Perspectives From Wielkopolska Luc Christiaensen, Céline Ferré, Maddalena Honorati, Tomasz Gajderowicz and Sylwia Wrona TOWARDS A JUST COAL TRANSITION LABOR MARKET CHALLENGES AND PEOPLE’S PERSPECTIVES FROM WIELKOPOLSKA Luc Christiaensen Céline Ferré Maddalena Honorati Tomasz Gajderowicz Sylwia Wrona © 2022 International Bank for Reconstruction and Development / The World Bank. 1818 H Street NW, Washington, DC 20433, USA. Telephone: 202-473-1000; Internet: www.worldbank.org. Some rights reserved This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: Christiaensen, Luc, Céline Ferré, Maddalena Honorati, Tomasz Gajderowicz, and Sylwia Wrona. 2022. “Towards a Just Coal Transition: Labor Market Challenges and People’s Perspectives from Wielkopolska.” World Bank, Washington, DC. License: Creative Commons Attribution CC BY 3.0 IGO. Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org TOWARDS A JUST COAL TRANSITION Labor market challenges and people’s perspectives from Wielkopolska 1 Towards a Just Coal Transition Labor market challenges and people’s perspectives from Wielkopolska Luc Christiaensen, Céline Ferré, Tomasz Gajderowicz, Maddalena Honorati, and Sylwia Wrona.1 Background papers, valuable data analysis, and comments were provided by Elizabeth Ruppert Bulmer, Maciej Jakubowski,2 and Jan Frankowski, Joanna Mazurkiewicz, Jacub Sokołowski and Piotr Lewandowski.3 October 2022 1 Corresponding authors: Luc Christiaensen, (lchristiaensen@worldbank.org), Jobs Group, World Bank; Céline Ferré (cferre@worldbank.org), Jobs Group World Bank, and Tomasz Gajderowicz (tgajderowic@wne.uw.pl.edu), Department of Economics, University of Warsaw. 2 University of Warsaw. 3 Institute of Structural Research (IBS). 2 Acronyms AI Artificial Intelligence ALMP Active Labor Market Program BGT Burning Glass Technologies BKL Human Capital Balance (Bilans Kapitału Ludziego) CHP Combined Heat and Power CSO Civil Society Organization DCE Discrete Choice Experiment EC Europen Commission EGD European Green Deal EU European Union GDP Gross Domestic Product GHG Greenhouse Gas GUS Statistics Poland (Główny Urząd Statystyczny) IBS Instytut Badań Strukturalnych (Institue of Structural Research) IEA International Energy Agency ILO International Labor Organization IO Input-Output IRENA International Renewable Energy Agency ISCO International Standard Classification of Occupations JTF Just Transition Fund JTM Just Transition Mechanism LO Labor Office NDC Nationally Determined Contribution NGO Non-Governmental Organization NUTS Nomenclature des Unités Territoriales Statistiques (Nomenclature of Territorial Units for Statistics) OECD Organisation for Economic Cooperation and Development OTJ On-the-job PES Public Employment Services PLN Polish Złoty RES Renewable Energy Sources TA Technical Assistance TJTP Territorial Just Transition Plan ZE PAK Zespół Elektrowni Pątnów-Adamów-Konin WTP Willingness to Pay 3 Executive Summary This report illustrates a data-driven and people-centered approach and identifies key policy principles to develop viable job transitions for a just coal transition, with an application to Eastern Wielkopolska in Poland. Coal-related jobs are at the forefront of the disruption brought about by the transition toward a low-carbon economy. Some displaced workers may be able to transition easily to new job opportunities, whereas many others may not. The resulting disruption to jobs and livelihoods may exacerbate the already challenging labor market environment in remote regions and traditional sectors that have not kept pace with broader economic modernization trends. A series of recent World Bank studies closely examined the labor, skills, and policy implementation challenges brought about by the transition out of coal in three Polish regions: Wielkopolska, Silesia, and Lower Silesia.4 This paper zooms in on the labor market challenges regarding the just coal transition in Wielkopolska and the affected people’s perspectives and opportunities. Particular attention goes to analyzing spatial, expectation, and skill-related mismatches between current and possible future jobs and their bearing on the viability of job transitions (including through reskilling). To make the complementary social protection and labor market support most effective, a series of principles regarding the design and implementation of this support are also advanced. Finally, an artificial-intelligence-powered, excel-based job-matching tool is developed and illustrated to help policymakers. identify reskilling needs, attract investments, and assist caseworkers in developing “viable-job-transition-pathways” for affected workers. Eastern Wielkopolska is the region most advanced in the transition out of coal. As Europe’s largest coal producer, Poland is today at the vanguard of the European coal transition. Within Poland, Eastern Wielkopolska is the region most advanced in the transition out of coal: finding viable job transitions is of imminent importance. As of February 2022, the ZE PAK Capital Group, one of the country’s largest lignite conglomerates, had closed most of its extraction sites, the last three power plants were planned to shut down by 2024, and all lignite mining activities in KWB Konin by 2030. Unlike in other parts of Poland, mine workers are only a minority of the coal-related workforce in Wielkopolska. They make up only 20 percent of ZE PAK’s 3,800-person workforce. Power plants employ a quarter of ZE PAK’s workers. The remaining 2,100 employees work in two subsidiaries operating within the group. They provide support services for mining extraction and power plants. Beyond ZE PAK, there are an estimated 250 workers in other segments of the coal value chain in the region. The affected coal-related workforce in Wielkopolska lives concentrated around the mines, calling for a geographic, or local economic development approach to brokering a just coal transition, rather than a sectoral approach. Importantly, with already weak labor market indicators, and coal a major source of municipal finance, four municipalities in the Konin powiat are particularly at risk of changes in ZE PAK’s labor demand. In all four municipalities . 4 The work has been supported by a grant from the Directorate-General for Energy of the European Union. . 4 (Kazimierz Biskupi, Wierzbinek, Kleczew, and Wilczyn) (i) finding a job is difficult, as reflected by high levels of registered unemployment; (ii) ZE PAK is the predominant purveyor of jobs, reflected in the large share of ZE PAK employees in total employment (between 19 and 39 percent); and (iii) households’ incomes are highly dependent on ZE PAK wages, as suggested by the significant share of ZE PAK employees over total working-age population. Even households with no members working for ZE PAK Capital Group are indirectly affected by mining and energy sector developments. Between a fourth (Wierzbinek) and a half (Kazimierz Biskupi) of all household survey respondents with no ZE PAK employee depend at least partly on the mining and energy sector; 11 percent have at least one household member working in the mining or energy sector in firms other than ZE PAK. This happens in a context where the regional economy of Wielkopolska performs broadly on par with the rest of Poland. ZE PAK workers are similarly skilled as other workers in Wielkopolska (and Poland more broadly), but non-ZE PAK workers in at-risk municipalities are substantially less skilled. Overall, ZE PAK employees score similarly to other Poles in Wielkopolska and Poland, for most skills, even though higher for maintenance and repair skills, and somewhat lower for advanced digital skills (-0.4 point) (World Bank Skills and Preference Survey 2021). This holds for both higher and lower-educated workers and bodes well for their chances in the local labor and regional labor markets. Non-ZE PAK workers in at-risk municipalities, on the other hand, not only score worse than workers elsewhere, but they also score below the ZE PAK workers (including the older ones) who, in addition to enjoying targeted transition support, . on the local labor market. This makes non- will thus be better equipped to take up new jobs ZE PAK municipality workers particularly vulnerable to the upcoming transition out of coal and potential displacement effects, further compounding the effects of the already lower economic dynamism and more limited employment opportunities offered by the communities they live in. Coal-related workers in Wielkopolska display a high aversion to relocation and commuting; they also put a premium on job security and continued use of their current competencies. Only 12 percent of ZE PAK workers would be willing to take up a job in Turek, which is 50 km away or a 50-minute drive from Kazimierz Biskupi; and 42 percent indicated that they would not be interested. Forty percent replied that it depends on the job offer. The DCE model estimates that an additional hour of commute is worth PLN 1,036, which is equivalent, on average, to one-fifth of a monthly salary at ZE PAK. They evaluate the financial compensation needed for working abroad, ceteris paribus, at PLN 5,055 per month and for moving to another region within Poland at PLN 2,705, which correspond respectively to a monthly and half a monthly average salary at ZE PAK. The resistance to traveling or moving for work substantially reduces the resilience of the local labor force to adjust to sudden or longer-term structural changes in the economies. It is also an indication of the broader welfare loss associated with the coal transition beyond the income from coal-related jobs itself. Non-ZE PAK municipality workers are even more reluctant to relocate for, or commute to, work. Furthermore, coal-related workers are willing to reskill, but value job security and a position 5 consistent with their competencies and experience. Residents of the four municipalities are ready to accept lower job entry conditions if it holds prospects for good progression. International best practice principles can guide the design and implementation of support packages to increase their effectiveness. When local labor markets are hit by the withdrawal of a core employer or sector, as in Eastern Wielkopolska, existing social assistance and labor market programs and the capacity of local employment offices may be insufficient to address the needs of the directly, and indirectly, affected workers. Adherence to a number of principles in the design and implementation of support programs can make them more effective. Support packages should be comprehensive (including psychological support), properly balanced between income support and re-employment services, tailored to the needs of the affected workers and deployed timely (starting pre-layoff), coordinated with job- generating investments to align training and reskilling programs, and not displace other regularly unemployed such as youth and long term unemployed. Implementation will be . more effective if the different stakeholders are involved throughout and properly informed, if services are well coordinated, often involving the set-up of a one-stop-shop, if labor offices are adequately staffed, or leveraged by hiring private providers, if financing is secured in a timely manner, and if the implementation is accompanied by a performant monitoring and evaluation system with adequate feedback mechanisms. To assist caseworkers in developing “viable-job-transition-pathways” for affected workers and help policymakers identify reskilling needs and attract investments, an excel-based job- matching tool, calibrated on the Polish labor market, is introduced. The job matching tool uses big data techniques to identify positions requiring tasks and skills most similar to the position held by the worker at risk of dismissal. It narrows down options to occupations with demand surplus, as identified by the local labor market barometer published annually by the local labor offices, and with wage offerings sufficient to overcome mobility aversion if the job offers are only available outside the affected municipality or region. Viable transition pathways are available in the current local labor market for blue-collar workers (who account for 70 to 80 percent of coal-related workers), but existing labor demand may not be sufficient to absorb everyone; tertiary-educated specialists may need more reskilling given the demand for high specialization. Most viable trajectories do not point to retraining and upskilling in renewable energy skills, the focus of much of the regional diversification efforts, or in digital jobs, often considered the jobs of the future. Key challenges going forward will be to provide adequate job opportunities for affected workers, especially non-mine workers in the most affected municipalities. This segment of the population deserves special attention: their numbers are non-negligible, they are less skilled, they operate in already lagging local economies, and they are not covered by the social agreements covering mine workers. The “viable-job-matching tool” can be developed further through ground-truthing with caseworkers in local labor offices and as a planning tool to guide re-purposing plans and economic diversification that leverage the available skills and account for the workforce’s key job attribute preferences. 6 7 Table of Contents Acronyms .................................................................................................................................... 3 Executive Summary...................................................................................................................... 4 1 Introduction ...................................................................................................................... 12 2 Coal in Poland – A declining sector with peculiar employment features ............................... 19 2.1 Decarbonization adds impetus to Poland’s longstanding decline of coal production ............................ 19 2.2 Concomitantly, employment in the mining sector has fallen sharply to 88,000 jobs ............................. 20 2.3 Recent estimates put the number of coal-related jobs between 145 and 218 thousand ....................... 21 2.4 Large and increasing wage premia in coal-related jobs, especially for the lower skilled ....................... 24 3 Coal-related employment in Wielkopolska: limited and geographically concentrated ......... 26 3.1 Concentration in few lagging and depopulating districts within an average region ............................. 26 3.2 Excess demand for higher skilled workers in the region and substantial skills mismatch ...................... 31 3.3 Coal workers are a declining, blue-collar workforce concentrated in one firm ...................................... 35 3.4 Coal-related jobs are especially important to Konin powiat within Eastern Wielkopolska .................... 41 4 The type of jobs coal-related workers can and would like to do........................................... 44 4.1 Local understanding of coal-related workers’ skills and job preferences is needed ............................... 44 4.2 ZE PAK workers similarly skilled as others; workers in affected municipalities, less skilled ................... 46 4.3 Job attributes coal-related workers value .............................................................................................. 50 5 Policy principles for a successful transition of affected workers ........................................... 58 5.1 The existing social protection and labor systems are not fully adequate ............................................... 58 5.2 Five principles to guide the design of additional social protection and labor support ........................... 59 5.3 Five principles for successful implementation ........................................................................................ 61 6 Assisting the transition with an AI-powered job-matching tool............................................ 63 6.1 Ingredients needed to assess the viability of individual job and labor force transitions ........................ 63 6.2 Towards a “viable-job-matching” decision tool tailored to the Polish labor market ............................. 65 6.3 Five transition pathways relevant to the mining and energy sector ...................................................... 67 6.4 Demonstrated proof of concept, but further validation and development needed. .............................. 73 7 Conclusions ....................................................................................................................... 73 Bibliography .............................................................................................................................. 78 Annex 1: Methodology to estimate the indirect impact of mine closure ...................................... 82 8 Annex 2: Skills and preference survey questionnaires ................................................................. 84 ZE PAK survey questionnaire ............................................................................................................................ 84 Municipalities survey questionnaire ................................................................................................................ 90 Annex 3: Skills and preference survey methodology.................................................................... 99 DCE methodology............................................................................................................................................. 99 Econometric model ........................................................................................................................................ 103 Annex 4: Preference survey – sampling frame ........................................................................... 106 Reweighting the sample from ZE PAK ............................................................................................................ 106 Reweighting the municipalities’ sample ........................................................................................................ 108 Annex 5: The viable job matching tool - methodological considerations .................................... 110 Tables Table 1: Middle-skilled occupations in the medical sector, manufacturing and drivers make the bulk of occupations where jobseekers exceed vacancies................................................................................. 32 Table 2: Previous estimates of indirect impact of mines closures........................................................ 82 Table 3: JRC (2018) Estimates ............................................................................................................... 83 Table 4: Job attributes – ZE PAK DCE design 1 .................................................................................... 100 Table 5: Job attributes – ZE PAK DCE design 2 .................................................................................... 101 Table 6: Job attributes – municipalities DCE design 1 ........................................................................ 102 Table 7: Job attributes – municipalities DCE design 2 ........................................................................ 103 Table 8: Education distribution in the population, unweighted sample, and after weights adjustment for non-response................................................................................................................................. 107 Table 9: Gender distribution in the population, unweighted sample, and after weights adjustment for non-response ...................................................................................................................................... 107 Table 10: Age distribution in the population, unweighted sample, and after weights adjustment for non-response ...................................................................................................................................... 107 Table 11: Gender distribution in the population, unweighted sample, and after weights adjustment for non-response ...................................................................................................................................... 109 Table 12: Age distribution in the population, unweighted sample, and after weights adjustment for non-response. ..................................................................................................................................... 109 Table 13: Identification of similar professions with the designed tool .............................................. 111 9 Figures Figure 1: Most of the coal employment in Europe is concentrated in Poland ..................................... 14 Figure 2: Employment in the mining sector was divided by 4.5 between 1989 and 2019 ................... 20 Figure 3: Polish workforce directly and indirectly employed in mining and power generation .......... 22 Figure 4: Employees in coal-related professions enjoy wage premia of 50 percent or more .............. 25 Figure 5: Mining is concentrated in Eastern Wielkopolska, within the Konin subregion ..................... 27 Figure 6: Konin subregion, and especially Konin powiat, are substantially lagging within an otherwise average performing region ................................................................................................................... 28 Figure 7: Employment is skewed towards traditional sectors .............................................................. 29 Figure 8: The highest intensity of unemployment in Wielkopolska occurs around power plants and mines ..................................................................................................................................................... 30 Figure 9: Eastern Wielkopolska has seen its workers leave and the birth rates decline ...................... 31 Figure 10: Most job offers in Wielkopolska are for higher skilled individuals ...................................... 33 Figure 11: 4 in 10 job offers are in traditional sectors of activity ......................................................... 34 Figure 12: Low skilled blue-collar workers are most often characterized by overeducation............... 35 Figure 13: PAK KWB Konin is the only active mining site left in Wielkopolska..................................... 36 Figure 14: 1 in 5 coal-related jobs are in the mines, but 9 in 10 jobs are provided by ZE PAK (including subsidiaries) .......................................................................................................................................... 37 Figure 15: Over the past decade, employment at ZE PAK contracted by half ...................................... 39 Figure 16: Half of ZE PAK workers will have retired by 2030 ................................................................ 40 Figure 17: Many ZE PAK employees are lower-educated, middle-aged and male ............................... 40 Figure 18: Four municipalities display a combination of high dependency on ZE PAK and poor labor market indicators .................................................................................................................................. 43 Figure 19: ZE PAK employees have transversal skills mostly similar to other Poles; workers in at-risk municipalities are less skilled across the board .................................................................................... 47 Figure 20: Lowest skills among men, especially those not working for ZE PAK ................................... 49 Figure 21: Example of a choice card ..................................................................................................... 50 Figure 22: ZE PAK employees list age and lack of local jobs as the biggest obstacles to finding a job 52 Figure 23. ZE PAK employees’ preferences – DCE 1 ............................................................................. 53 Figure 24. ZE PAK employees’ preferences – DCE 2 ............................................................................. 55 Figure 25. Preferences of the residents of municipalities – DCE 1 ....................................................... 56 Figure 26. Preferences of the residents of municipalities – DCE 2 ....................................................... 57 Figure 27: Examples of transition pathways for an energy engineer ................................................... 68 Figure 28: Examples of transition pathways for a lignite mining technician ........................................ 70 10 Figure 29: Examples of transition pathways for a truck mechanic ....................................................... 71 Figure 30: Examples of transition pathways for an electrician ............................................................. 72 Figure 31: Examples of transition pathways for an electrical machine fitter ....................................... 73 Boxes Box 1: Five lessons from past mine closure in Poland and the U.S. ................................................... 18 Box 2: Estimating the number of indirect coal-related jobs in Poland .............................................. 23 Box 3: ZE PAK subsidiaries and subcontractors .................................................................................... 38 Box 4: Retraining ZE PAK workers for opportunities in Renewable Energies (RE) ............................... 64 11 1 Introduction Following the Paris Agreement (2015)5, the European Union (EU) adopted “climate neutrality by 2050”, while also committing to leave no one behind. The aim is to reduce net greenhouse gas (GHG) emissions in the EU to zero by 2050, while also ensuring no one place or person is left behind. This ambition has been converted into law in 2021 through the adoption of the European Climate Law. The Law also sets the intermediate target of reducing GHG emissions by 55 percent by 2030 (compared to 1990 levels). To achieve the intermediate target, a set of concomitant legislative proposals and policy initiatives has further been advanced (the “Fit for 55” package). They include, among others, a reduction of the GHG emission caps of power plants through a modification of the Emission Trading System.6 To alleviate the socio-economic impacts of the transition in the most affected, mostly carbon- intensive regions, the EU further provides targeted support to help mobilize 100 billion euro in these regions over the 2021-2027 period. These efforts are governed under the Just Transition Mechanism (JTM).7 With carbon-neutral energy production at the forefront of reducing GHG emissions, accelerating the coal-phase out is a key area of policy action. The shift to climate neutrality requires a shift to carbon neutral energy production as well as the transformation of production and consumption patterns towards greater energy efficiency and lower resource use.8 With carbon dioxide the primary contributor to GHG emissions9, the shift away from fossil fuels (coal, petrol, gas) is at the forefront of the transition. Coal is the most carbon-rich energy source10, making the accelerated phase-out of coal a primary area of policy action. Over the past decades, coal consumption in the EU has come down substantially already (from 1,100 Mt in 1990 to 437 Mt in 2021).11 Yet, solid fossil fuels (i.e., coal and coal derived 5 https://unfccc.int/process-and-meetings/the-paris-agreement/the-paris-agreement 6 There are also many other measures such as a reduction of Member States’ emission reduction targets, reversing the declining trend of carbon removals through land-use change and reforestation, and increasing efficiency of energy use in transport (including in shipping and aviation), buildings, and industry through a combination of investments, CO2 emission standards, energy taxation, and carbon border adjustment mechanisms. 7 https://ec.europa.eu/info/strategy/priorities-2019-2024/european-green-deal/finance-and-green-deal/just- transition-mechanism_en 8 Beyond energy production, other heavily affected sectors include energy-intensive industries (such as steel, chemicals, plastics), agriculture, waste management and transport. 9 In 2019, CO2 accounted for 81.6 percent of EU GHG emissions. 10 The amount of CO2 emitted per million British thermal units (Btu) for coal, petrol and gas are respectively 229/215 (anthracite or hard coal/lignite), 161/157 (diesel/gasoline), and 139/117 (propane/natural gas) pounds (https://www.americangeosciences.org/critical-issues/faq/how-much-carbon-dioxide-produced-when- different-fuels-are-burned). 11 In 2021, total (hard and brown) coal consumption in the EU stood at 437 Mt (of which 160 Mt hard coal and 277 Mt brown coal). This is down from 1,100 Mt in 1990 (of which 400 Mt hard coal and 700 Mt brown coal). Most coal use is for power production (49 percent of hard coal and 92 percent of brown coal in 2020). 12 solids)12 still account for 12.6 percent of the EU’s 2020 electricity generation (Eurostat, 2022)13, but makes up 62 percent of electricity and heating CO2 emissions (EIA, 2019)14. Coal regions and workers are particularly exposed to the energy transition, exacerbated by their economies’ strong dependence on coal and the attractive labor contracts typically enjoyed by coal mine workers. Coal mines and their associated activities (including power generation) tend to be spatially concentrated, around the extraction sites. They are often the main employer in the local economy, providing financially attractive and secure employment for many semi-skilled workers, especially those working in the mines. With few attractive alternatives present in the often lagging and undiversified local economies, coal mine closures tend to heavily affect their workers and communities. The high reservation wage and historical and cultural attachment to mining further impede labor mobility (across sectors and space). Labor unions often successfully manage to negotiate transition packages to somewhat mitigate the social costs of mine closures. Even so, the negotiated packages typically do not (or only partially) apply to workers in the supporting industries or the non-coal-related workers in the local economies, leaving many exposed. They also don’t compensate for the broader economic loss. As a result, across the world, transitions out of coal have been fraught with major social challenges, especially in coal producing regions (World Bank, 2018a; World Bank, 2021; Ruppert Bulmer et al., 2021). In recognition of this challenge, a specific financial instrument has been created by the EU under the JTM, the Just Transition Fund (JTF), to mitigate the social and employment effects of the coal transition in carbon-intensive regions. For the EU’s coal phase out and its broader climate transition, a successful transition out of coal in Poland is particularly important. After Germany, Poland is the EU’s second largest coal producer and consumer.15 It further employs about half of Europe’s coal related labor force (Figure 1). What happens with coal production, coal workers and coal regions in Poland is a first order matter for brokering a just coal and energy transition in Europe. The transition out of coal in Poland started long time ago. But coal remains of strategic importance to the Polish economy. In 2020, over 40 percent of the country’s total energy supply (TES) and 70 percent of its electricity generation come from coal and lignite (IEA, 2022), the highest rate in Europe. Coal also continues to employ about 88,000 people directly in the mines, down from about 444,000 in 1989 (415,000 in hard coal; 29,000 in lignite). Europe’s commitment to stop its fossil fuel imports from Russia following Russia’s invasion of Ukraine will slow down 12 Solid fossil fuels cover various types of coal (such as hard and brown coal) as well as solid products derived from coal (such as coke which is used in the steel industry). 13 Eurostat consulted on 13 May 2022 (online data code NRG_BAL_PEH) (https://ec.europa.eu/eurostat/databrowser/view/NRG_BAL_PEH__custom_2712428/default/table?lang=en) 14 Solid fuel is defined as coal, peat and oil shale (EIA, 2019). 15 Total (hard and brown) coal consumption in Poland amounts to 117.6 Mt in 2021 or 26.9 percent of the EU’s total (hard and brown) coal consumption (Eurostat, 2022). Most of it is for power generation and heating. Total (hard and brown) coal consumption in Germany is 153.8 Mt in 2021 (36.8 Mt hard coal (all imported) and 127 Mt brown coal). https://ec.europa.eu/eurostat/statisticsexplained/index.php?title=Coal_production_and_consumption_statisti cs 13 Poland’s coal phase-out in the near future to ensure energy security in Europe,16 but Poland remains committed to complete coal mine closure by 2049. Figure 1: Most of the coal employment in Europe is concentrated in Poland 160 Employment (thousand workers) 155 140 120 100 80 60 40 20 27 22 18 16 15 4 3 3 3 2 0 Poland Romania Czechia Serbia Bulgaria Germany Greece Slovakia Slovenia North Spain Macedonia Note: employment in hard coal and lignite, as registered under sector B05 of NACE Rev.2 classification. Source: Eurostat (2019), table lfsa_egan22d. In 2021, coal was still extracted in six regions in Poland : hard coal in Silesia, Lesser Poland, and Lublin; and lignite in Greater Poland (i.e., Wielkopolska), Lower Silesia, and Łódź. Most of the extraction sites are in the Upper Silesian basin, but only about half of the coal seams are economically viable. In addition to hard coal, Poland exploits several lignite deposits: three sites, exclusively surface mines, are currently operating, in Łódź, Lower Silesia, and Wielkopolska. In contrast to Poland’s hard coal mines, which operate at an average working depth of 600 meters, the lignite mines consist almost entirely of mine-mouth power plants. This report, which focuses on Eastern Wielkopolska, is part of a larger set of regional labor market studies in coal transition regions in Poland. Funded by the European Commission, these studies cover the three regions most advanced in their coal phase-out and economic diversification process: Wielkopolska, Silesia, and Lower Silesia. Together, these regions account for 86 percent of all Polish coal mining employment, the preponderance in Silesia. Each region also qualifies for JTF support. Among the three regions, Eastern Wielkopolska is the region most advanced in the transition out of coal, with most mining related activities concentrated in one conglomerate. The area is economically heavily dependent on coal and energy production, 16 Before Russia’s invasion of Ukraine on February 24, 2022, Poland’s coal phase -out was on an accelerating path, given the high cost of extraction and the challenge for Polish coal-fired power plants to provide electricity within politically acceptable price margins. 14 and it is characterized by the presence of the most polluting plants in the region (GHG emissions in Eastern Wielkopolska amounted to 7.62 million Mg in 2019, about 62 percent of all emissions in Wielkopolska). ZE PAK Capital Group is the biggest employer in the region and the largest private energy group in Poland, consisting of entities engaged in lignite extraction, power generation from conventional and renewable sources, heat production and electricity trading. Lignite mining activities are planned to shut down in KWB Konin (operating in the area of Konin and Kola) by 2030, and in three power plants operated by the ZE PAK Capital Group by 2024.17 The region’s ambition is to move from coal-based energy production towards green energy with three specializations: (i) renewable energy, (ii) electromobility and (iii) hydrogen technologies. The Konin sub-region in Eastern Wielkopolska is also identified as one of the coal regions eligible for JTF support.18 The report presents a “people” perspective on the transition, giving voice to those affected; it assesses their skills and prospects for future jobs within the context of their local labor markets. It begins by evaluating the number of workers (directly and indirectly) affected by the mine closures within the broader context of the local and regional labor markets within which they operate. It subsequently maps out their skills and job attribute preferences and aspirations. It categorizes the task content of all jobs within Poland and combines this with information on the affected workers’ skills and job attribute preferences to identify viable job transition pathways for affected workers within the current local and broader regional/national labor markets. To inform these different areas of inquiry, the study uses administrative and secondary labor force data from national labor force surveys and local labor market barometers from the public employment offices. To assess the skills and reveal the job attribute preferences of affected workers, the study further conducted surveys and discrete choice experiments with representative samples of the ZE PAK employees and the non-coal workers in the most affected surrounding communities. Finally, the study also reviewed the lessons from international experience regarding the institutional design and practical implementation of a just transition of affected workers out of coal. The findings provide an important piece of the “just coal transition” policy puzzle in Poland, but abstract from labor demand related issues. Among the lessons learned from past mine closures (Box 1) are the importance of timely analysis and broad consultation as well as the need for a comprehensive approach, addressing both labor supply and demand side issues. This report abstracts from the employment opportunities new investors and employers may provide (including through repurposing of the mining lands and assets), i.e., the labor demand side of the challenge, including the investments and policy reforms needed to attract them. As such it produces only part of the information needed. In particular, it provides policymakers and potential investors with an in-depth characterization of the local labor pool, 17 KWB Adamów finished closing down in February 2021. 18 Coal regions may access and disburse JTF resources based on their Territorial Just Transition Plans (TJTP). Only projects in eligible JTF regions at NUTS3-level qualify for JTF support. 15 in terms of numbers and demographics, their job attribute preferences and aspirations, as well as their skill profile (cognitive/non-cognitive; routine/non-routine). This helps investors and policymakers with investment decisions and the design, targeting and content of their active labor market programs (ALMPs). The matching tool developed for the report, which identifies viable job transition pathways within the current labor market for those affected, can be readily adapted to identify skill and aspiration gaps vis-a-vis future job opportunities as the mines and their lands are repurposed and the local economies are restructured. A series of principles to guide the choice and operational implementation of transition measures are further presented. Overall, the report finds that the impact of the transition will be concentrated in a few municipalities that are highly dependent on coal extraction, where workers display a strong preference for continuity and stability, and where the most at-risk individuals will be non- mine workers. In Eastern Wielkopolska, a limited number of workers will be directly and indirectly affected by the transition: 3,800 workers were still employed by ZE PAK, and an additional 250 workers were employed by ZE PAK subcontractors in 2020. But, coal-related employment is spatially concentrated around the mines, in four municipalities where ZE PAK is the dominant employer. Those municipalities display sluggish labor markets, even though they are embedded in a more dynamic regional economy (Wielkopolska). Non-mine workers in affected municipalities deserve special attention, especially as they are less skilled, are not covered by social agreements, and face limited transition prospects as they operate in heavily affected and already lagging local economies. ZE PAK employees and residents of the four most highly affected municipalities in Wielkopolska want to remain in municipalities where they live (and are willing to give up a lot for that); work in similar positions/sector of activity; and they value job security. The job-matching tool developed for this study proposes viable transition pathways and reskilling options under current local labor market conditions, but the number of potential positions available may not be sufficient to absorb all workers who may lose their job in the future. Core principles for the effective transition of affected workers out of coal concern among others the appropriate balance between income support mechanisms and re-employment services to maintain incentives for continued engagement in the labor market and avoid displacement of others. Monitoring and evaluation systems accompanying implementation are often deficient. Future analysis should complement and link the information in this report with the development of a regional, economic diversification strategy, an in-depth analysis of the local investment climate to reorient coal-related businesses and attract new domestic and foreign investment, and an assessment of the number and type of jobs the regional strategy is likely to generate (locally in the mining communities, and/or elsewhere in the region or country). While new employment opportunities should come from all sectors of the economy, the energy sector can still be a catalyst for regional development, including in Eastern 16 Wielkopolska.19 Overall, the key labor market challenge is to create a more supportive environment for firms to thrive and create employment and to provide adequate opportunities for workers to build skills relevant to alternative, more competitive sectors of work. The report proceeds as follows. Section 2 elaborates on the extent and nature of coal production and employment in Poland by way of background. This is followed by an analysis of the local labor market features and performance in Eastern Wielkopolska and the current (and historical) role of coal related activities within it (Section 3). The skills and job attribute preferences and aspirations of Eastern Wielkopolska’s affected workers are reviewed in Section 4. Section 5 presents a series of principles derived from international good practice to guide the implementation of a just transition of labor out of coal in practice. Based on a decomposition of all jobs in Poland (including those related to coal) into their composite tasks, a practical, data-driven tool to identify viable job transition pathways for different types of affected workers is further introduced and discussed in Section 6. The application is illustrated based on the jobs available in the current labor market and also takes workers’ job attribute preferences and skill profiles into account. The section further illustrates how the job matching tool can be used when information about future jobs and associated skill needs becomes available, for labor market planning and ALMP programming (such as the nature and design of reskilling programs) as well as for individual case counseling of affected workers. Section 7 concludes. 19 At the global level, new jobs in transition-related technologies and sectors are expected to outweigh job losses in fossil fuels and nuclear energy (IRENA, 2021). 17 Box 1: Five lessons from past mine closure in Poland and the U.S. First, timing and pace are key: transition takes a long time and preparation should start early. When many workers, businesses and communities are implicated, fundamental change to an industry cannot happen quickly, even with the best advance planning and post-closure transition policies in place. The timing and speed of transition are subject to political economy dynamics. Uncertainty around commodity prices makes it difficult for communities to transition because prices affect both willingness and capacity to diversify toward other industries. Where actors are public (e.g., Poland), governments have the power to act quickly but risk the future support of the electorate. Where actors are private but unions are strong and/or regulatory authority is weak or captured by private interests (e.g., the U.S.), boom and bust cycles can be exacerbated, which could create obstacles to both the design and implementation of effective transition policies. Finally, severe social dislocation in a context of fragile local economic viability may also pass a point of no return. The risk is higher where long-term dependence on coal has delayed acceptance of transition. Second, transition requires a comprehensive approach with complementary initiatives, policies, and incentives to sway the many actors along the coal value chain, including those with vested interests like utility monopolies and manufacturers of mining equipment. Labor demand (attracting firms) and labor supply (e.g., reskilling) challenges must be addressed simultaneously, with inappropriate or delayed repurposing of the mining lands and assets often leaving many missed opportunities. Third, economic diversification is essential and requires help from both local and higher-level government with respect to planning and financial resources. Advance planning, investment in infrastructure, addressing environmental degradation and attracting private investment are key ingredients of economic diversification, requiring significant local and regional institutional capacity and coordination, including for repurposing of the mining lands and assets. Fourth, transition support packages need to be inclusive, to help workers who are directly or indirectly affected by mine closure to find new jobs. Addressing worker transition challenges effectively requires a solid understanding of the scope and nature of the potential impacts of transition. Policymakers need to understand the ways in which a future transition away from coal may affect the livelihoods of both coal and non-coal workers and their surrounding communities, in order to implement policies and programs for managing transition effectively. Transition assistance programs targeting formal mine workers fall short of meeting the needs of informal or subcontracted workers in and around the mines. Even large mine operators employ a significant share of their workforce on temporary and/or informal contracts. Informal coal sector workers are at greater risk than their formal counterparts and less equipped to weather income shocks. Packages can include jobs search support, reskilling, counseling, entrepreneurship training, and/or access to financial resources. Fifth, the advantages of inducing voluntary job separations through generous compensation packages are offset by the risk of inflicting long-term damage on local economies. High reservation wages dampen local labor demand and economic recovery through diversification. This can undermine public fiscal health and ultimately diminish local institutions and social capital. Too- generous severance packages can induce labor exit, which can accelerate a coal community’s decline. Source: Ruppert Bulmer et al. (2021). 18 2 Coal in Poland – A declining sector with peculiar employment features 2.1 Decarbonization adds impetus to Poland’s longstanding decline of coal production Poland’s coal production has more than halved since the 1990s, driven by reduced economic viability. Over the past three decades total coal production in Poland has more than halved, from 229 Mt in 1990 to 108 Mt in 2020. Most of the reduction happened in hard coal (from 147 Mt in 1990 to 55Mt in 2020). This is consistent with the broad downward trend in coal production in the Europe Union. In fact, Poland is by now virtually the only hard coal producer left in the EU (producing 96% of all EU hard coal production; the Czech Republic produces the remainder). Hard coal reserves in Poland are managed through a few state- owned conglomerates. Hard coal mostly consists of steam coal for electricity generation and heating (72 percent).20 It is deep-mined and fully mechanized with over 90 percent of coal produced by longwall systems (Euracoal, 2020). Nonetheless, only half of the Polish coal seams are still considered economically viable (at least until the recent price hike following Russia’s invasion of Ukraine on Feb 24, 2022).21 Poland further exploits several lignite deposits, which consist almost entirely of mine-mouth power plants. Productivity is around 6,800 tons per employee, which is much lower than the levels observed in Germany (the largest EU lignite producer) and Greece (EURACOAL, 2020). Finally, state-owned mines or units of integrated mines (whether hard coal or lignite) that are unprofitable and slated for phase-out are transferred to the public entity Spółka Restrukturyzacii Kopalń (SRK) to manage their eventual closure. In 2021, there were 12 mines or units of mines managed by the restructuring company (SRK, 2021). In the May 2021 “social agreement” between the government and mining trade unions, Poland presented a decarbonization plan that lays out the timetable for hard coal mines closure.22 Following the European Green Deal and in line with the “social agreement”, the coal sector in Poland will gradually reduce its output capacities and close its mines, over the next three decades, with the last mines closing in 2049. In the energy sector, the country wants to replace its coal-fired power generation capacity with increased production from renewable energy, including a significant portion from offshore wind, coupled with natural- gas-fueled power plants and especially nuclear energy, to stabilize output in an increasingly RES-dependent power grid. Poland is facing a historic challenge to transition quickly from an energy model based on coal extraction – in place for the past 70 years – to a zero-emission 20 The rest mostly consists of coking coal. 21 As a result, until recently, part of Poland’s hard coal consumption was imported (in 2020, 10.6 Mt out of the 65.6 Mt of hard coal consumed in Poland was imported, much of it from Russia). Brown coal or lignite is little traded. 22 https://www.gov.pl/web/aktywa-panstwowe/umowa-spoleczna. 19 future. This is further compounded by the February 2022 invasion of Ukraine by Russia, which undermines the potential of natural gas as transition fuel.23 2.2 Concomitantly, employment in the mining sector has fallen sharply to 88,000 jobs Since 1989, as coal production declined, employment in the coal mining sector 24 has fallen by 80 percent. Over the past three decades, Poland’s coal mining sector lost approximately 300,000 jobs, down to 92,600 in 2019 (Figure 2) and subsequently to 87,600 by end-2020. Most of the reduction in employment happened in the first decade, during Poland’s transition from a centrally planned model to a market economy (1989-2002). Coal extraction now represents only 1 percent of total employment in Poland. But it still represents 5 percent for the largest extracting region (Silesia),25 and it can represent up to half of employment at the local level, rendering coal-intensive municipalities particularly vulnerable, as discussed further below. Figure 2: Employment in the mining sector was divided by 4.5 between 1989 and 2019 500,000 28,393 450,000 415,740 Total number of workers 400,000 350,000 300,000 250,000 200,000 9,301 150,000 83,300 100,000 50,000 0 1989 1994 1999 2004 2009 2014 2019 Hard coal Lignite Note: Totals exclude employment in coal processing. Source: INSTRAT, 2021. The sharp reduction in coal mining employment was the result of hard coal mine closures in the early 1990s. Between 1990 and 1995, employment reduction was achieved almost entirely through a hiring freeze. The remaining cuts were achieved through a combination of 23 Russia has traditionally been an important supplier of gas for Poland, which now needs to be sourced from elsewhere. Ongoing investments to expand natural-gas-fueled power generation are continued, but the extension of the lifetime of some coal-fueled blocks is considered as alternative for constructing new natural- gas powered power plants. 24 Note that “coal mining sector” refers to hard coal and lignite mining activities. 25 Eurostat, 2020. Indicator: lfsa_egan22d. 20 hiring freezes, expanded access to early retirement, and – beginning in 1998 – voluntary exits facilitated by a generous Mining Social Package comprising cash benefits (severance pay and allowances), activation instruments (training, scholarships, retraining contracts, support for starting a business) and instruments aimed at employers (contribution refunds, reimbursement of employment costs). By 2002, coal mine jobs had fallen by nearly two-thirds to 164,000. This was followed by a slower pace of contraction, largely due to attrition (Baran et al. 2020). Whereas lignite extraction operated with a relatively stable workforce of around 24,000 until then, it began to decline steadily from 2002 onward, falling below 10,000 by 2015 (see Figure 2). Recent years have seen some limited new hiring, for example during the 2008 economic and financial crisis, and more recently during the 2021 energy crisis. 2.3 Recent estimates put the number of coal-related jobs between 145 and 218 thousand Two groups of coal-related employment are considered, those directly employed in the mines by the mining companies and mining conglomerates, and those deriving their jobs and incomes from coal-related activities. The number of direct coal jobs can be obtained from the mines with relative precision and is estimated at 87,600 at the end of 2020 (80,000 in hard coal; 7,600 in lignite extraction) (INSTRAT, 2021). But each coal mining job also generates additional jobs in the coal value chain, i.e., in the firms and businesses supplying goods and services to the operations of the mines. Coal mines further provide the inputs for coal-fired power stations. Finally, much of the demand for locally produced goods and services in the local communities (and the related jobs) originates in the salary mass of the coal and coal-related workers (the so-called induced effects). The latter three employment effects (linked to the coal value chain, coal-fired power stations and the induced effects) are much harder to estimate. Appropriate disaggregated data is hard to come by and studies use different methods and different sectoral and geographic coverage, resulting in a wide range of estimates. The results are reviewed briefly here, under the broad umbrella of indirect effects (as opposed to the direct effects). Recent studies using input-output tables estimate the total number of jobs indirectly linked to Poland’s coal mining sector between 57,000 to 130,000 (Figure 3, Box 2). It is common to calculate the number of indirect jobs related to a sector using input-output (IO) tables. IO tables are a set of national (or regional) economic accounts disaggregated by sector. They are a snapshot of flows of products and services in the economy for a single year: they identify and disaggregate all the monetary flows between industries (inter-industry expenditure flows), between consumers and industries, and between industries and suppliers of factors in the economy. The results suggest that every coal mining job generates between 0.6 and 1.4 additional jobs, or an additional 57,000 to 130,000 jobs indirectly linked to coal mining. This puts the estimated total number of coal-related jobs in Poland today between 144,600, and 217,600 (Figure 3).Nonetheless, these studies only provide orders of magnitudes at best, and should be treated as such. Estimating the indirect impact of mines closure on employment through IO models is limited by the simplifying assumptions of the methodology. IO models rely on complicated yet simplified systems of equations, and the emphasis on the 21 product side of the economy doesn’t explain why inputs and outputs follow a specific economic pattern. First, the assumption of linear equations relating to one industry’s output may not be realistic: since the factors are largely indivisible, it is not always necessary to increase the inputs proportionally to increase the output. Second, IO models do not allow substitution between factors, and in the long-run, replacement potential between inputs may be relatively high. Figure 3: Polish workforce directly and indirectly employed in mining and power generation 158 Direct - Hard coal extraction Direct - Lignite extraction Direct - Other coal and lignite (beyond extraction) 130 Power plants (PP) Upper-bound 112 70 Upper-bound 110 88 Lower-bound 96 Lower-bound 8 (including Inter-regional subcontractors) 57 80 49 Recipients Suppliers (including intermediaries = Intra-regional coal-fired PP, heating, and 13 coking plants) Recipients Recipients (power plants) (power plants) Coal Power Kiewra Alves Dias Frankowski and Ingram Mining plants et al (2019) et al (2018) Mazurkiewicz et al (2020) (2020) DIRECT JOBS Mining and Power Plants INDIRECT JOBS Note: Each square represents 1,000 jobs. All estimates exclude induced jobs. “Hard coal extraction” and “Lignite extraction” estimates are reported by INSTRAT (2021) and refer to administrative data shared by mining conglomerates. However, Eurostat reports a larger number of individuals employed in “coal and lignite” (table lfsa_egan22d), namely 158 thousand: “Other coal and lignite (beyond extraction)” refer to the differential between the two estimates. Source: Alves Dias et al (2018), Eurostat (lfsa_egan22d, 2019), IBS (2020), Ingram (2020), INSTRAT (2021), Kiewra et al (2019). 22 Third, IO models require huge amounts of data, which may not be accurate or up-to-date, and are rarely disaggregated below the national level: IO tables are only available every five years in Poland (the latest year of information published by GUS is 201526). As a result, while providing some benchmarks, the results of IO models can only provide orders of magnitudes at best, and should be treated as such, as inputs into a broader debate. Below the report argues that the spatial concentration of coal mines and coal mining related activities in certain municipalities is likely the more important finding, indicating that the effects of coal mine closures are hard felt locally. Box 2: Estimating the number of indirect coal-related jobs in Poland Using IO tables from 2015, Kiewra et al (2019) and Alves Dias et al (2018) estimate that between 57 and 88 thousand persons respectively are indirectly linked to coal mining in Poland.27 The estimations rely on the use of IO tables and multipliers from the EU Joint Research Center, originally developed for predicting the effects of a change in the final demand in one sector on other related sectors (Thissen and Mandras, 2017). The reasons for the difference in their estimates are not immediately clear. Both Alves Dias et al (2018) and Kiewra et al (2019) use the IO tables from 2015 and they both consider employment by coal recipients or users (particularly coal-fueled power plants). To calculate the number of indirect jobs generated in the coal supply chain, they both assume that the ratio of mining-related value-added to mining-value added corresponds to the ratio of mining- related jobs to mining jobs. Kiewra et al apply this ratio to the 2017 number of mining jobs; the base year of mining employment used by Alves et al (2018) is not clear. In terms of reporting, Alves Dias et al report separately also the intra- and inter-regional impacts, respectively 49 and 39 thousand jobs, though both effects should be included by Kiewra as the estimates are national. Frankowski and Mazurkiewicz (2020) further refine Kiewra et al’s work, estimating that Poland’s coal mining sector generates between 96 and 112 thousand indirect jobs, 41 thousand of which are generated among suppliers. As in Kiewra et al, indirect jobs are calculated as the share of value added transferred to the mining industry in the total value added generated in a given sector, multiplied by the number of employees in that sector. The paper differs in that employment data was retrieved from Eurostat data for 2018, and that employment among other recipients than coal-fired power plants (in manufacturing, gas, steam, and air conditioning (respectively NACE Rev.2 C and D) were added, proportionally to the share of coal in each sector. Finally, Ingram et al (2020) provide a slightly larger range of estimates, from 110 to 130 thousand, using a specific survey of 207 companies.28 This study specifically considers the timeline of mines closure: under an optimistic scenario, 26,667 jobs would be indirectly affected by 2030, half of what is estimated under a plausible scenario (50,580). The different results, along with estimates of the workforce directly employed in the mining sector and in the power plants, are illustrated in Figure 3. 26 https://stat.gov.pl/en/topics/national-accounts/annual-national-accounts/input-output-table-at-basic- prices-in-2015,5,3.html 27 These estimates include power generation, equipment supplies, services, and R&D. 28 Unspecified sampling framework. 23 Additional bottom up estimates of the indirect effect of the mine closure on employment in the coal value chain of Wielkopolska are presented in section 3. In addition to national- level estimates, Alves Dias et al (2018) estimate the indirect impact of the transition at NUTS- II level29: a total of 8,090 jobs are indirectly linked to coal mining in Wielkopolska, 3,447 of which through intra-regional and 4,643 through inter-regional linkages. This report adds to this using a simple bottom-up approach: the ZE PAK Capital Group shared their list of subcontractors (6 entities in total), whose total workforce was estimated using administrative data (section 3.4). Final estimates of the workforce indirectly affected by the mines closure were disaggregated between those working on sites located in Wielkopolska, and those working on sites outside of Wielkopolska. The impact of the transition on households living in the municipalities most affected by mines closure is further refined in section 3.4. 2.4 Large and increasing wage premia in coal-related jobs, especially for the lower skilled Wage premia of 50 percent or more are not uncommon for employees in coal-related professions, especially for lower-skilled workers and in state-owned mines or mining conglomerates. Hourly wages in coal-related occupations tend to be substantially higher than wages in other sectors, both on average and when comparing within the same occupational categories (Figure 4).30 If coal-related workers were to occupy a similar occupation in a sector other than mining or energy production, their average gross salary would be 20 to 40 percent lower. Put differently, most workers earn at least 50 percent more than those in similar occupations outside the mining sector. The wage premium earned by coal-related workers is particularly high for the lower skilled occupations (craft workers, plant operators and especially elementary occupations) and in public mines and mining conglomerates, where the average sectoral wage gap vis-à-vis other professions was PLN 3,065 in 2018 compared to 1,339 in private mines.31 When controlling for demographic characteristics such as gender, age, and years of education, in addition to occupation, miners’ wage premia are even higher. The logarithm of the earnings regressed by applying the Mincer’s Equation (1974) , which regresses earnings on the worker’s educational attainment, including additional variables indicating sociodemographic (age, sex) and controlling for working time indicates that miners earn 50 to 60% more than other workers with similar education and work experience. 29 The NUTS classification (Nomenclature of territorial units for statistics) is a hierarchical system for dividing up the economic territory of the EU for the purpose of the collection, development and harmonization of European regional statistics. NUTS-I refers to major socio-economic regions, NUTS-II to basic regions for the application of regional policies, and NUTS-III to small regions for specific diagnoses (https://ec.europa.eu/eurostat/web/nuts/background). 30 61 occupations from the International Standard Classification of Occupations (ISCO-08), which are directly affected by the Just Transition, were identified. These occupations, which are directly related to mines and coal- fueled power plants, are listed in different sectors of activity: management of power plants and mines (4 professions), mining in general (24 professions), quarry mining (11 professions), energy production (15 professions), and mining Infrastructure (7 professions). 31 In high-skilled professions in privately-owned mines, the premium does not exceed 20% on average, while in the public sector it is similar to this for low-skilled workers 24 Figure 4: Employees in coal-related professions enjoy wage premia of 50 percent or more Managers Professionals Technicians 0.30 0.15 0.20 0.20 0.10 0.10 0.05 0.10 0.00 0.00 0.00 0 5,000 10,000 15,000 20,000 0 5,000 10,000 15,000 20,000 0 5,000 10,000 15,000 20,000 Craft workers Plant operators Elementary occupations 0.40 0.40 0.40 0.30 0.30 0.30 0.20 0.20 0.20 0.10 0.10 0.10 0.00 0.00 0.00 0 5,000 10,000 15,000 20,000 0 5,000 10,000 15,000 20,000 0 2,500 5,000 7,500 10,000 Mining and quarrying All other sectors Note: Average gross wages. Average monthly earnings were calculated as total salaries and benefits for 2018 divided by 12 and number of hours worked. Source: Author’s computations using Structure of Earnings Data (BDL, access: 1.11.2021). Between 2004 and 2018, the wage32 gap vis-à-vis other sectors generally widened further. The average salaries of mining engineers, technicians, and miners’ more than doubled over that period: from PLN 4,617 to PLN 8,926 for engineers, from PLN 3,927 to PLN 8,032 for technicians, and from PLN 3,112 to PLN 6,700 for miners. While in 2004 a mining engineer earned PLN 1,563 more than an engineer in another sector, in 2018, the wage gap widened to PLN 4,495. Similar increases were observed for technicians and miners. On the other hand, average salaries of other types of occupations in the mining sector grew at a slower pace, more closely aligned with national wage trends (GUS, 2020).33 These results are consistent with the World Bank (2018) study which finds that coal wages are not only high but have been rising faster over time, even when comparing coal and non-coal workers with the same demographic characteristics, including educational attainment.34 32 Average gross monthly earnings were calculated as total salaries and benefits for 2018 divided by 12. 33 The average wage for the Polish economy grew by 100 percent between 2004 and 2018 (in nominal terms). 34 Different factors likely underpin such wage differences. They may simply be compensation for the health hazards and difficult working conditions of coal-related employment. Given the high capital intensity of the mining sector, they may also reflect higher labor productivity. Finally, through effective unionization coal-related workers may have more effectively leveraged their bargaining power for faster wage growth, especially the lower-skilled workers in deep mining who are also most exposed to health hazards and difficult working conditions. 25 Finally, the types of occupations, contract terms, compensation and working conditions can vary widely between coal mine workers, and across the coal value chain. Employees of mining conglomerates benefit from the most protective contracts and social packages (in case of dismissal), negotiated by often powerful unions. They enjoy good contract terms (high wages, open-ended contracts, sometimes guaranteed employment within the conglomerate), generous health and unemployment insurance, early retirement eligibility, and pensions indexed to their high wages. They often also benefit from subsidized access to energy (free coal or subsidized heating). Within mining conglomerates, underground miners (hard coal) receive more generous compensation than surface miners (lignite). On the other hand, employees of supporting industries within the coal value chain (suppliers and service providers) tend to have less secure contracts, lower wages, and less protective social insurance than those directly employed by the mining conglomerates, even though their working conditions are on average likely still better than those outside mining and quarrying (Figure 4).35 3 Coal-related employment in Wielkopolska: limited and geographically concentrated 3.1 Concentration in few lagging and depopulating districts within an average region Wielkopolska’s coal mining activity is concentrated in Eastern Wielkopolska, within the Konin subregion. The region of Wielkopolska (Wielkopolska Voivodeship) represents the NUTS-2 administrative level in Poland. It is divided into the city of Poznań and 5 subregions (NUTS-3 level), including the Konin subregion. The Konin subregion consists of 8 districts (powiats), including Konin powiat and Konin municipality. The mining area of Eastern Wielkopolska is a subset of the Konin subregion, and includes 5 powiats only (Konin, Turek, Kolski, Słupecki, and the municipality of Konin) (Figure 5).This is where Wielkopolska’s mining and mining-related activities are concentrated, with coal dependency particularly high in 4 municipalities in the Konin powiat (see further below). 35 Given definitional issues, the exact wage premium by occupation between those indirectly employed in coal and those employed in other sectors is hard to discern from the Structure of Earnings Data; they do not clearly distinguish between those directly employed in mining and those indirectly employed. 26 Figure 5: Mining is concentrated in Eastern Wielkopolska, within the Konin subregion Konin subregion Eastern Wielkopolska Wielkopolska Poland Wielkopolska Note: Eastern Wielkopolska is a subset of Konin subregion (8 powiats) and consists of 5 powiats. Eastern Wielkopolska’s economy and labor market lag behind the rest of Wielkopolska (Figure 6). Average GDP per capita in the Konin subregion is high and substantially higher than in the rest of Poland36. It underscores the importance of the mining and power generating sector for the regional and local economy. Yet, labor market participation within the Konin subregion is low. Almost half of the working age population in the Konin subregion is inactive, and the registered unemployment is 6 percent of the working-age population.37 Or, only about 45 percent of the working-age population in the Konin subregion is actually working, 8 percentage point less than in the Wielkopolska region as a whole.38 Among those employed, many more are working in traditional (i.e., natural-resource related) sectors (61 versus 50 percent), especially in agriculture (31 versus 18 percent; Figure 7). Unsurprisingly, wages in the Konin subregion are on average also lower (Figure 6).39 Lower wages with higher inactivity and unemployment have further combined to considerably higher poverty. 21 percent of the population in the Konin subregion has been classified as poor in 201140, compared to 11 percent in Wielkopolska (GUS, 2020). The economy and labor market of the broader 36 Separate, disaggregated statistics for Eastern Wielkopolska are not available. 37 Unemployment rates are not available at the subregion level: instead, registered unemployment, as a share of the working-age population (15-to-64-year-olds) is reported. 38 With 43 percent of the working-age population in Wielkopolska inactive and the registered unemployment rate 4 percent (Figure 6), the employment rate (i.e., the employed share of the working-age population) is 53 percent. This assumes that the registered unemployed are in fact not working and looking for a job, which may not be true for all of them. Yet there may also be some unemployed, looking for a job, that are not registered as unemployed. 39 Higher earnings of those employed in “traditional” coal-related jobs in the Konin subregion--the mining conglomerate ZE PAK is the largest employer in Eastern Wielkopolska—do not compensate fully for the lower wages among those in the other traditional sectors such as agriculture, resulting overall in lower wages on average. 40 Poverty estimates below the NUTS-II level, or voivodeship, cannot be calculated from existing datasets, here EU-SILC. Numbers for Konin subregion (NUTS-III level) were obtained from Szymkowiak et al. (2017), who use the Small Area Estimates imputation technique. 27 Wielkopolska region are not performing much worse than the rest of Poland, however. Wielkopolska has in fact a higher GDP/capita and a higher employment rate than Poland (lower inactivity and unemployment share), but with employment slightly more concentrated in traditional sectors, wages are also somewhat lower on average and poverty rates slightly higher (11 vs 9 percent). Yet, overall, the differences are small. Figure 6: Konin subregion, and especially Konin powiat, are substantially lagging within an otherwise average performing region KP PL KS GP GDP PER CAPITA 45 50 55 60 65 70 (PLN per month) AV. GROSS SALARY (‘000 PLN per month) KP KS GP PL 45 50 55 60 65 70 EMPLOYMENT IN TRADITIONAL SECTORS PL GP KS KP (Share, %) 40 45 50 55 60 65 70 GP PL KS KP INACTIVITY RATE 40 45 50 55 60 65 70 (Share, %) GP KS PL KP REGISTERED 0 2 4 6 8 10 UNEMPLOYMENT (Share, %) PL GP KS POVERTY RATE 5 10 15 20 25 (Share, %) KP = Konin Powiat KS = Konin Subregion GP = Greater Poland PL = Poland Note: Traditional sectors focus on the extraction and processing of natural resources, including extraction of raw materials - mining, fishing, and agriculture. They are characterized by high labor-intensity. Source: GUS (2020) excluding GDP per capita (GUS, 2019), and 2011 poverty estimates for Konin subregion (Szymkowiak et al., 2017). 28 Figure 7: Employment is skewed towards traditional sectors Wielkopolska Eastern Wielkopolska 18% 21% 25% 33% 1% 2% 16% 33% 22% 29% Agriculture, forestry, fishing Industry and construction Rest Trade, repair of motor vehicles, transportation and storage, accommodation and catering, ICT Financial and insurance activities, real estate activities Note: Traditional sectors focus on the extraction and processing of natural resources, including extraction of raw materials - mining, fishing, and agriculture. They are characterized by high labor- intensity. Source: GUS, 2019. Within Eastern Wielkopolska, the Konin powiat lags most behind and is also most vulnerable – many of Wielkopolska’s coal-related workers reside here and its economy heavily depends on coal. Konin powiat performs worse on all economic and labor market indicators. Average GDP per capita in the powiat of Konin (which excludes the city of Konin) is only 72 percent of the region’s average, one of the lowest in the region (GUS, 2020). Labor market participation is also substantially lower than in the Konin subregion and the region as a whole (close to 60 percent of the working age population is inactive and the unemployment share is as high as 9.3 percent). Unemployment is especially high among youth, with 49 percent of women under the age of 35 registered unemployed at the end of December 2020 (36 percent for young men). Among those employed, even more people are in the traditional sectors than in the Konin subregion, and wages are on average even lower than in the Konin subregion. Four municipalities located within Konin powiat stand out in particular: Kazimierz Biskupi, Wierzbiniek, Kleczew and Wilczyn. They are most at risk to be affected by the transition out of coal: they display already weak labor market indicators and with ZE PAK the largest employer in these municipalities, providing employment to between 20 and 40 percent of those at work, they are highly vulnerable to changes in ZE PAK’s labor demand (Figure 8). 29 Figure 8: The highest intensity of unemployment in Wielkopolska occurs around power plants and mines Unemployment shares (%) <1% 1-2% 2-3% 3-4% 4-5% >5% Note: Unemployment shares are computed as the share of unemployed individuals in the total working-age population (active + inactive). As such, they are different from unemployment rates, which are defined as the share of unemployed working-age individuals among active (unemployed + employed) working-age individuals. Source: GUS, 2021. Finally, and unsurprisingly, over the past decades, Eastern Wielkopolska has also seen its population, number of children, and work force decline. With too many people inactive and/or unemployed and little diversification to more modern activities (in manufacturing or services) higher wages for those with jobs in the mining sector have not been able to prevent a contraction and ageing of the population. During 2000-2020, the population of Konin municipality, for example, declined, by about 12 percent, from 82,640 residents in 2000 to 72,539 in 2020 (GUS). For the most affected municipalities the population decline during 2000-2020 amounts to 18 percent in Kazimierz Biskupi, 6 percent In Wierzbinek, 3 percent in Kleczew and Wilczyn. As can be seen from the lower number of 30-45-year-olds in Eastern Wielkopolska compared to the rest of Wielkopolska (Figure 9), outmigration contributed importantly. This is likely further compounded by a declining birth rate (as reflected by the lower number of 0-14-year-olds). While outmigration and declining birth rates reduce the need for future jobs, the reducing labor supply and the lack of newly educated workers may also deter investors and hinder economic diversification. 30 Figure 9: Eastern Wielkopolska has seen its workers leave and the birth rates decline 85+ 80-84 75-79 70-74 65-69 60-64 55-59 50-54 45-49 40-44 35-39 30-34 25-29 20-24 15-19 10-14 5-9 0-4 20,000 15,000 10,000 5,000 0 5,000 10,000 15,000 20,000 Female (Wielkopolska) Male (Wielkopolska) Female (Eastern Wielkoposlka) Male (Eastern Wielkopolska) Note: Wielkopolska’s population is scaled to match Eastern Wielkopolska’s population. Source: GUS, 2020. 3.2 Excess demand for higher skilled workers in the region and substantial skills mismatch Labor demand in Wielkopolska exceeds supply in most professions. Wielkopolska’s labor market is tight, with job creation far exceeding job destruction.41 Five percent of all employers were looking for workers in 2019 (GUS, 2020), and according to the occupational barometer,42 employers report difficulty filling vacancies in most occupations. A few occupations have more workers than jobs, mostly in middle-skilled professions related to the medical sector, and low and middle-skilled occupations related to industry, manufacturing, and transport (Table 1). The situation on the labor market may change due to the influx of refugees - the influx of Ukrainian citizens to Poland may turn out to be an injection of human capital for some industries struggling with staff shortages. According to the information from the Border Guard, more than 3.64 million refugees from Ukraine have already crossed the Polish- Ukrainian border since the beginning of Russia's invasion. Due to data limitations, the labor market analysis is carried out for Wielkopolska and not Eastern Wielkopolska per se. Yet, this also helps shed light on the broader labor market opportunities available to coal workers. 41 Throughout 2019, 68 thousand new jobs were created, and 30 thousand jobs were liquidated (GUS, 2020). 42 Poznań Regional Labor Office, 2020. “Monitoring of Professions in Deficit and Surplus in the Greater Poland Voivodeship in 2019.” Professions in deficit, surplus and balan ce are identified by a group of experts, which gathers every year. As such, it is not a quantitative comparison of the number of registered jobseekers and registered vacancies. 31 Table 1: Middle-skilled occupations in the medical sector, manufacturing and drivers make the bulk of occupations where jobseekers exceed vacancies Surplus Balance Auxiliary medical staff Managers of other types of services, not Laboratory diagnostics specialists classified elsewhere Glasses opticians Advertising and marketing specialists Driving instructors Conductors and related Constructors and cutters of clothing Craftsmen of wooden and related materials Operators of pulp and paper production equipment Fitters and service technicians of electronic devices Sellers in marketplaces and bazaars Tourist guides and tour leaders Information office workers Note: Excess demand is not displayed, as it concerns all remaining occupations (neither in surplus, nor in balance). Source: Poznań Regional Labor Office, 2020. “Monitoring of Professions in Deficit and Surplus in the Wielkopolska Voivodeship in 2019. Job vacancies in Wielkopolska are concentrated among professions requiring higher skills. Reviewing the jobs on offer in 2020 across the different fora (on and offline) and agencies (public and private)43, the Poznań Regional Labor Office concluded that labor demand in Wielkopolska has been more concentrated in skilled occupations. For example, most online job offers in Wielkopolska target specialists (37 percent), followed by technicians and other associate professionals (23 percent), representatives of public authorities, senior officials, and managers (13 percent) (Figure 10). Only 3 percent of online job offers cater to “Workers doing simple jobs” (Poznań Regional Labor Office, 2020). But not all job offers are published online and private sector firms are much more likely to advertise vacancies through the Internet than public sector entities (respectively 87 and 53 percent). The latter are more likely to cooperate with Public Employment services (PES) (43 percent compared to 31 percent in the private sector) and aimed at somewhat lower skilled workers (Poznań Regional Labor Office, 43 PES job offers accounted for less than half (41 percent) of all job offers in Wielkopolska in 2018, with the market share of private employment agencies continuously increasing. 32 2019). Job offers registered with PES tend to be concentrated in traditional sectors, such as industrial processing and construction (Figure 11). Figure 10: Most job offers in Wielkopolska are for higher skilled individuals Specialists 37% HIGH-SKILLED Technicians and WHITE COLLARS associate professions 22% Representatives of public authorities, senior officials, and managers 13% Salesman and service workers 9% LOW-SKILLED WHITE COLLARS Office workers 6% Industrial workers HIGH-SKILLED and craftsmen 4% BLUE COLLARS Machine and device operators and assemblers 4% LOW-SKILLED Workers doing simple jobs BLUE COLLARS 3% Source: Poznań Regional Labor Office, 2020. “Research on Job Offers on the Internet for Monitoring Occupations in Deficit and on the Local Labor Market.” Across occupations, soft skills are increasingly sought after. Employers in the region, as elsewhere, look particularly for entrepreneurship skills, initiative, and creativity. Knowledge of foreign languages, the ability to search for information, negotiate, plan one’s work, and communicate are also very much valued.44 Skills mismatches are significant (60 percent of jobs), with undereducated workers filling a fair number of high skilled white-collar occupations and overeducated workers filling most low skilled blue-collar occupations. Given data limitations, the degree of skills mismatch is measured by comparing the corresponding education level of occupations, as a proxy for their skill requirement, with the workers’ actual educational attainment (Figure 12). Two points stand out. High-skilled white-collar occupations are partially filled by undereducated workers (11 percent of specialists, 52 percent of technicians and associate professionals, and 41 percent of managerial jobs). These are also occupations in high and excess demand in Wielkopolska (Figure 10), leading employers to underfill the positions. The labor market for high skilled white-collar works is tight. Low-skilled blue-collar workers, on the other hand, are 44 Poznań Regional Labor Office, 2020. “Professional Activation and Job Placement in the Greater Poland PSZ.” 33 overwhelmingly overeducated, suggesting insufficient vacancies for medium skilled or high skilled blue-collar workers, as alluded to above, leading them to accept jobs below their qualifications. The labor market for low skilled blue-collar workers is sluggish. Fewer skills mismatches are observed among low-skilled white collar and high-skilled blue-collar workers, with both over- and underfilling occurring. The situation in Wielkopolska is similar to this of the rest of Poland, where two in three occupations (65 percent) are over- or underfilled, compared to 60 percent in Wielkopolska. Figure 11: 4 in 10 job offers are in traditional sectors of activity Industrial processing Scientific activities Construction Education Transport/warehouse Healthcare and social assistance Business/admin services Retail Public admin, defense Accommodation Other Agriculture, forestry Note: Job vacancies listed by the Public Employment Services (PES). Source: Poznań Labor Office, 2020. “Professional activation and job placement in the Wielkopolska Public Employment Services.” Skills mismatch exacerbates labor market sluggishness in coal-dependent regions. The importance of skills mismatches when looking for a job is confirmed by ZE PAK employees (World Bank Skills and Preference Survey, 2021). ZE PAK employees mentioned that the lack of jobs in the region was the biggest obstacle to finding a job. This suggests both a lack of job offers (as demonstrated by the local labor market indicators) as well as a mismatch of competencies (consistent with the high value they put on finding a job commensurate with their competencies, when evaluating different job attributes (section 4)). 34 Figure 12: Low skilled blue-collar workers are most often characterized by overeducation 100 59 89 48 59 65 77 88 11 19 Share of workers (%) 80 60 40 20 0 Managers Professionals Technicians Clerical Service and Skilled Craft and Plant and Elementary and associate support sales workers agricultural, related machine occupations professionals workers forestry and trades operators fishery workers and workers assemblers High-skilled white collars Low-skilled white collars High-skilled blue collars Low-skilled blue collars Undereducated Overeducated Rightly educated Note: Data for Wielkopolska. The correspondence between occupations and skills is using the ILO nomenclature. High-skilled white collars are expected to have tertiary education (ISCED 5-8), low- skilled white collars and high-skilled blue collars are expected to have more than upper-secondary non- tertiary education (ISCED 3-4), and low-skilled blue collars are expected to have up to upper-secondary education (ISCED 0-2). Source: Author’s computations using Structure of earnings data (BDL, access: 1.11.2021). 3.3 Coal workers are a declining, blue-collar workforce concentrated in one firm 3.3.1 Coal-related work in Eastern Wielkopolska is concentrated in one conglomerate Most coal-related activities in Eastern Wielkopolska are carried out within one, vertically integrated, privately owned conglomerate, the Zespół Elektrowni Pątnów-Adamów-Konin (ZE PAK) Capital Group. It is the largest private energy group in Poland. As of December 2021, ZE PAK will continue to exploit only one extraction site in Eastern Wielkopolska, however. The Pątnów-Adamów-Konin (PAK) lignite basin has been producing lignite for over fifty years and now generates approximately 2.5 percent of Poland’s electricity needs. Until 2021, there were two active mining sites: Konin and Adamów, operated by two subsidiary companies. The first subsidiary, PAK Kopalnia Węgla Brunatnego Konin SA, consisted of three mines: Jóźwin IIB, Drzewce and Tomisławice, which supplied three mine -mouth power plants, Pątnów I, Pątnów II and Konin.45 Lignite production at Konin is planned to continue till 2030, although only the Tomisławice surface mine will be working after 2020. Following the completion of mining and mine decommissioning at Adamów, the second subsidiary, PAK 45 The third power plant, Konin, uses mainly biogas since the beginning of 2021. 35 KWB Adamów SA, which had been supplying the Adamów power station, stopped operating in the first quarter of 2021 (Figure 13). Figure 13: PAK KWB Konin is the only active mining site left in Wielkopolska Wilczyn Wierzbinek Klecew Tomisławice Kazimierz Jóźwin Biskupi Pątnów Drzewce Konin KONIN Wladyslawów Koźmin POWIAT Adamów Eastern Wielkopolska Municipalities most affected by the mines closure (Preference Survey) Active mining areas Closed mines Headquarter of ZE PAK Power plants Konin Only 1 in 5 coal-related jobs in Eastern Wielkopolska are in mines, but 9 in 10 coal-related jobs in Wielkopolska are provided by ZE PAK (including subsidiaries) (Figure 14). Unlike in other parts of Poland, mine workers are only a minority of the coal related work force in Wielkopolska. They make up only 20 percent of ZE PAK’s 3,800-person workforce. Power plants employ a quarter of ZE PAK’s workers. The remaining 2,100 employees work in two subsidiaries operating within the group that provide support services for mining extraction and power plants. Beyond ZE PAK, there are an estimated 250 workers in other segments of the coal value chain in the region, employed across 6 firms that are external subcontractors of ZE PAK (Figure 14). A full description of the structure of ZE PAK and its subsidiaries is in Box 3. 36 Figure 14: 1 in 5 coal-related jobs are in the mines, but 9 in 10 jobs are provided by ZE PAK (including subsidiaries) ZE PAK mines PAK Górnictwo (subsidiary) 1,404 ZE PAK power plants 706 PAK Serwis (subsidiary) 950 Indirect (within Greater Poland) 759 974 250 Indirect (outside of Greater Poland) Source: IBS, 2021. 37 Box 3: ZE PAK subsidiaries and subcontractors Two subsidiary companies operating within ZE PAK Capital Group are responsible for repairs and services in the consortium: PAK Górnictwo and PAK Serwis. Both enterprises were established around mine and power plant teams involved in the maintenance and renovation of production infrastructure. PAK Górnictwo supports mining activities within ZE PAK Capital Group, offering a wide range of services including repair and maintenance of machines, servicing heavy equipment, rail and road transport of goods, electrical installations, repair and maintenance of electrical devices, and site preparation and land reclamation. Due to the phase-out of lignite mining, PAK Górnictwo will gradually reduce its activity. The company is making efforts to expand the scope of its services to the external market, including the construction of photovoltaic farms and other services (transport, mechanical repairs, reclamation, and development of post-mining areas). At the beginning of 2021, 1,404 persons were employed by the subsidiary, and by 2024, the company expects to employ 323 people, equivalent to a 77 percent reduction in 3 years (ZE PAK administrative data). PAK Serwis supports power generation activities within ZE PAK Capital Group, offering a broad spectrum of maintenance services of technology equipment, electrical, automation and IT systems in power engineering and industrial plants. Furthermore, the company provides comprehensive project management in power engineering and industry. It also carries out specialized tests and diagnostic measurements (such as electrical equipment and installation tests, thermographic testing, leak detection, emissions measurement of gas and dust pollution). The company has a more diversified portfolio of clients that include energy companies (both conventional energy and renewables), as well as industrial companies in Poland and Germany. At the beginning of 2021, the company employed 706 people, two thirds of whom were technicians, one fourth were engineering and professional technicians, and 5 percent worked in administration. By 2024, the company expects to employ 600 people, equivalent to a 15 percent reduction in 3 years (ZE PAK administrative data). In addition to the two subsidiaries operating within ZE PAK, six subcontractors manage the combustion of by-products and the restoration of degraded land, under contracts with ZE PAK. These firms together employ around 250 workers in Wielkopolska, and an additional 950 people elsewhere (IBS, 2021). − Dolina Nidy produces gypsum building materials made of stone obtained from an opencast mine and synthetic gypsum from gas desulphurization installations. The company has two production plants, one of which is located in Wielkopolska on the premises of Zespół Elektrowni Pątnów-Adamów-Konin. At the end of 2020, Dolina Nidy employed 185 persons, most of whom were blue-collar, prime-age workers with general upper-secondary or tertiary education. − Wienerberger has a ceramic plant in Honoratka, in the municipality of Sompolno, and produces ceramic products using by-products from KWB Konin. In 2020, it employed 749 persons. − Siniat produces gypsum and gypsum products for the construction industry. The company is based in Warsaw and has one plant in Konin. In 2020, it employed a total of 268 people throughout the country. − EKOTECH provides ash disposal and combustion waste management services, as well as full or partial revitalization of degraded land. − Continental Road and Borowiak both provide mineral and construction products. Details on their activities or connections with lignite production are unavailable. Source: IBS, 2022. 38 3.3.2 Over the past decade, coal-related employment in Eastern Wielkopolska halved Eastern Wielkopolska will most likely be the first region to completely phase out of coal. Lignite mining has declined at an accelerating pace in recent years. Between 1990 and 2017, lignite production and consumption in Poland fell from approximately 68 to 63 million tons (Polish Geological Institute, 2020). But in 2020, only 47 million tons of lignite were produced in Poland. Most of the decline took place in Eastern Wielkopolska. The decrease in lignite production is reflected in the decline in labor demand. The number of jobs in ZE PAK Capital Group fell from 5,600 in 2017 to 4,000 in 2020 and 3,800 at the beginning of 2021. Until Russia’s invasion of Ukraine in February 2022, this trend was predicted to persist to 2030, when lignite mining and the use of lignite in the power and heating sectors would be discontinued. Figure 15: Over the past decade, employment at ZE PAK contracted by half 9,000 8,700 8,300 8,000 8,000 7,200 7,000 7,000 6,400 Nuber of employees 5,900 6,000 5,200 4,800 5,000 4,000 4,000 3,000 2,000 1,000 0 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Hetmański et al (2021), using ZE PAK administrative data. By 2030, ZE PAK is projected to have lost another 2,000 workers (half its current employment) through natural attrition and early pension schemes. Using ZE PAK administrative data, Hetmański et al (2021) estimate that, based on current pension packages, 38 percent of ZE PAK workers will be retired by 2025, and another 55 percent by 2030, coinciding with the end of coal-based operations (see Figure 16). These estimates are subject to change, depending on faster shutdowns of open-cast mines and power plants, temporary changes in hiring (such as during the recent energy crisis), and the type of workers dismissed (e.g.,, ZE PAK may not select redundant employees based on age, but rather on the positions it requires to run the sites).46 More broadly, following Russia’s invasion of Ukraine in February 2022, the end date and trajectory may be revised to ensure energy security in the short run as Europe decouples its fossil fuel supply from Russia. The Government of Poland’s overall commitment to a complete coal phase out by 2049 remains. 46 Further analysis of administrative data suggests that the current rate of job cuts (on average 150 employees quarterly) will be maintained until 2025, after which operation of Pątnów II Power Plant (still to be determin ed) will require significantly fewer workers than at present. 39 Figure 16: Half of ZE PAK workers will have retired by 2030 300 100% pension rights up to a given Cumulative percentage of number of employees 240 2030 80% employees acquiring 55% (persons) 180 60% year 120 40% 60 20% 0 0% 2019 2024 2029 2034 2039 2044 2049 2054 number of employees (left axis) cumulative percentage (right axis) Source: Hetmański et al (2021), using ZE PAK administrative data. 3.3.3 Older, lower skilled blue-collar male workers are the majority of ZE PAK’s workforce The highest share of employees (40 percent) was aged 50 years or more, only 9 percent of ZE PAK employees were women. However, there are significant gender differences in terms of workers’ age profiles. Men over 50 years of age constituted 35 percent of the entire group of men, whereas the share of women in this age group was significantly higher at 54 percent. The youngest employees (under age 30) constituted only 5 percent of all staff (Figure 17). Compared to the rest of Poland and Wielkopolska, the educational attainment of the ZE PAK workforce is also lower: two-fifths of the employees have a basic or vocational education background, compared to 34 and 28 percent in Wielkopolska and Poland respectively. Figure 17: Many ZE PAK employees are lower-educated, middle-aged and male Education structure Age structure 100% 100% 31 22 80% 36 35 54 60% 38 36 35 50% 40% 41 28 20% 29 34 23 17 15 0% 5 5 6 0% 7 4 Poland Greater ZE PAK Men Women Poland (men) up to 31-40 41-50 >50 Basic Vocational Secondary Tertiary 30 y.o. y.o. y.o. y.o. Note: all data represent year 2019. Source: Hetmański et al (2021), using BDL GUS, ZE PAK administrative data, and own data collection. 3.3.4 ZE PAK employees enjoy higher wages, but have lower household incomes Employees of the coal mining sector enjoy high wages, but they are substantially lower than counterparts in Silesia and in Lower Silesia, where most mines are state-owned (BDL, access: 1.12.2021). Mining engineers and technicians earn PLN 6,590 and PL 5,384 respectively in Wielkopolska, which is about PLN 1,793 and 1997 less than in Silesia, and PLN 40 309 and PLN 1940 less than in Lower Silesia. Similarly, ZE PAK miners’ wages are lower than miners in other regions: miners from ZE PAK earn PLN 4,731 on average (ZE PAK administrative data, 2020), compared to PLN 6,901 for miners in Lower Silesia and PLN 6,960 in Silesia (BDL, access: 1.12.2021). The wage gap of ZE PAK employees vis-à-vis other sectors within the region has also narrowed in recent years, especially among miners (GUS, 2020, and Hetmański, 2021). In 2019, ZE PAK employees earned on average a gross monthly salary of PLN 4,966 (ZE PAK administrative data, 2020), which is on par with Wielkopolska’s thriving capital, Poznań (PLN 5,700), and 30 percent above the regional and subregional averages of PLN 4,700 (Wielkopolska) and PLN 4,200 (Konin subregion) (GUS, 2020). Between 2013 and 2019, ZE PAK salaries increased by 2 percent only, compared to 4.5 percent wage growth in the Konin subregion, and 5 percent at the regional and national levels (ZE PAK administrative data from Hetmański, 2021). Only higher-skilled engineers and technicians saw their earnings increase faster than other sectors of activity in Wielkopolska: higher-skilled engineers experienced the largest increase from 2004 to 2018 (63 percent), and technician’s 56 percent, compared to 51 percent among miners (BDL, access: 1.12.2021). This is consistent with the reported excess demand and observed underfilling of high skilled white-collar workers in Wielkopolska. ZE PAK employees’ high wages do not compensate for low earnings from other household members, resulting in lower per capita household incomes on average (World Bank Preference Survey, 2021). Two thirds of ZE PAK employees live in double-income households. Nevertheless, three-fifths of ZE PAK employees have disposable per capita income below the national average. This is particularly true for employees without higher education (below bachelor), 68 percent of whom are below the national average, compared to 52 percent of employees with higher education (World Bank Preference Survey, 2021). 3.4 Coal-related jobs are especially important to Konin powiat within Eastern Wielkopolska The economy of Konin powiat is quite vulnerable to the transition out of coal, given the high concentration of the economy and employment in the coal value chain work, with little diversification of the economy thus far. Although an increase in entrepreneurship is observed in the Konin subregion, the increase in business entities in the Konin powiat is one of the lowest in the region (Churski, Perdał and Burchardt, 2021). In recent years the position of the Konin subregion has become more competitive in the production of metal products (aluminum smelter and automotive industry), there has also been a development of furniture and textile companies and logistic industry thanks to good location of the region and accessibility of communication. Most of the enterprises operating in the Konin subregion are, however, small economic entities employing up to 9 people (Hołub-Iwan, Orsa-Chomiak, Terlecki, Gutta, Gutta, Gozdek, 2019). ZE PAK Capital Group and its subcontractors account for 7 percent of employment in firms with more than 10 employees located within the municipality of Konin or Konin powiat (ZE PAK administrative data). This high concentration around Konin municipality and powiat is largely due to the fact that one of the largest 41 employers in the region’s coal value chain – namely PAK Serwis, with 1,400 employees – is located in the municipality of Konin. The vast majority of workers employed in Konin powiat have job stability (90 percent have open-ended contracts), and most work in the services sector. Non ZE PAK job opportunities tend to be less appealing, however: the majority are located in small companies, and pay relatively low wages (Hołub-Iwan, Orsa-Chomiak, Terlecki, Gutta, Gutta, Gozdek, 2019). Labor force participation in Konin powiat is low: only one in two working-age adults is employed. Most of the inactive working-age population is older and/or female: two thirds are over 45 years of age, and two thirds are women. Five percent of the working-age population is unemployed, and these tend to be younger and less educated individuals. Forty- two percent of the unemployed are registered with the PES, but they feel moderately prepared to look for a job. Only one in three unemployed workers is aware of institutions or organizations that can help with job search (World Bank Skills and Preference Survey, 2021). ZE PAK plays a dominant role as an employer in several municipalities in Eastern Wielkopolska that neighbor the lignite sites. Ninety-eight percent of all ZE PAK employees live in Eastern Wielkopolska. One third commute from the municipality of Konin, but they only make up 2.6 percent of the working-age population. On the other hand, municipalities located close to the mines are highly dependent on ZE PAK for employment: ZE PAK represents respectively 39, 31 and 29 percent of total employment in Kazimierz Biskupi, Wilczyn, and Wierzbinek (Figure 18). Considering that men make up 91 percent of all ZE PAK employees, about half of the employed men living in a given municipality are employees of ZE PAK (74 percent for Kazimierz Biskupi and 66 percent for Wierzbinek) (Hetmański et al, 2021). Four municipalities in Konin powiat are especially at risk, namely Kazimierz Biskupi, Wierzbiniek, Kleczew and Wilczyn. These municipalities display already weak labor market indicators and are highly vulnerable to changes in ZE PAK’s labor demand. In all four municipalities (i) finding a job is difficult, as reflected by high levels of registered unemployment; (ii) ZE PAK is the predominant purveyor of jobs, reflected in the large share of ZE PAK employees in total employment; and (iii) households’ incomes are highly dependent on ZE PAK wages, as suggested by the significant share of ZE PAK employees over total working-age population (see Figure 18). Even households that have no members working for the ZE PAK Capital Group are indirectly affected by developments in the mining and energy sector. Between a fourth (Wierzbinek) and half (Kazimierz Biskupi) of all household survey respondents with no ZE PAK employee depend at least partly on the mining and energy sector; 11 percent have at least one household member working in the mining or energy sector in firms other than ZE PAK (World Bank Skills and Preference Survey, 2021).47 47 Respondents’ self-assessment, to the question: “To what extent is your household income dependent on the fuel and energy sectors?” Those responding that they were “completely, partially, or slightly dependent” (as opposed to “completely independent”) were considered dependent. 42 Figure 18: Four municipalities display a combination of high dependency on ZE PAK and poor labor market indicators 8% 39% 40% 6% 6% 31% 29% 6% 5% 30% 5% 5% 5% 4% 4% 19% 4% 3% 20% 3% 2% 5% 10% 0% 0% Kazimierz Wierzbinek Kleczew Wilczyn … Konin Biskupi Municipality Share of registered unemployed in total working age population (Left Axis) Share of ZE PAK employees in total working age population (Left Axis) Share of ZE PAK employees in total employment (Right Axis) Note: Eastern Wielkopolska municipalities with ZE PAK employees, ranked from four most dependent on ZE PAK, to the least dependent on ZE PAK (the municipality of Konin). Source: ZE PAK administrative data and GUS (2020). Recent survey results of the non-ZE PAK labor force in these municipalities point to high inactivity rates, high public sector employment and mainly low-skilled jobs. According to the World Bank Skills and Preference Survey (2021), 44 of the (non-ZE PAK) household respondents in these communities, were economically inactive, mostly older people - over 45 years of age, and 5 % were unemployed. The unemployment period varies and amounts to an average of 11 months. Less than half of the unemployed are registered at the Labor Office. Most unemployed people would take up part- or full-time work; however, there is also a group interested only in part-time work (31%) or only full-time work (16%). Starting one’s own business is rarely considered - 11% of respondents took this into account. Wage expectations differ depending on the level of education. Higher education translates into higher salary expectations: a net monthly salary below PLN 3,000 would not be accepted by respondents with higher education. At the same time, 71% of respondents with at least secondary education would be satisfied with a salary of PLN 2,501-3,000. Among the employed, an employment contract is the dominant form of (72%) and mostly for an indefinite period (90%). 15% of working respondents are self-employed, and 5% are employed based on a mandate contract. Only 57 % of the employees work primarily in the private sector; 42% work in the public sector. Household survey respondents usually work in small companies employing up to 10 people (44%) or 11 to 49 people (13%). The positions held rarely require specific education, for 48% of the level of education does not matter in their workplace, and 29% believe that secondary vocational education is sufficient. Employees rarely manage the work of others; only 15% supervise or manage other people's work. Although 16% considered job change in the last 12 months, only 8% have been looking for a job or participated in the recruitment process. 43 4 The type of jobs coal-related workers can and would like to do 4.1 Local understanding of coal-related workers’ skills and job preferences is needed The coal-related workforce displays little mobility traditionally. Mining activities are by nature location-bound, with coal-based industries (coal-fired power stations, coking and steel industries) typically also establishing in the vicinity of the mines, given the bulkiness and cost to transport coal. This has often generated a strong attachment of coal-related workers to the local community and low geographical mobility (the Ruhr valley in Germany, the Nord- Pas-de-Calais basin in France, the Upper-Silesian basin in Poland). Beyond limited geographic mobility, employees of the mining sector usually also display little cross-sectoral mobility. Many spend their entire career working for one employer, the mining conglomerate, which typically also dominates the local labor market. Coal mining jobs often involve semi-skilled production and machine operation occupations, which have limited immediate transferability to other sectors. As they are usually better paid than others holding similar jobs in other sectors, employees in the mining sectors also have high wage expectations (section 3.3.4). As a result, even employees with skills that are easily transferable (e.g., electricians, drivers) have limited incentive to look for work elsewhere. These social and economic forces driving the limited geographic and intersectoral mobility of coal-related workers are compounded by a deeply felt cultural identity, rooted in mining.48 Reasons for this could be traced to the communities’ historical contribution to their nations’ early industrialization and energy security, the emergence of social rights and workers’ movements around mining -related activities, and strong group solidarity and work ethics in general. Does limited mobility of the coal-related workforce across space and sectors still hold today? Low worker mobility limits the pool of potential job opportunities, exacerbating the challenge of a just coal transition. This holds especially when the surrounding districts or regions are economically vibrant, as in many regions in Poland (especially in Lower Silesia, but to a lesser extent also in Silesia and Wielkopolska). Yet, whether workers in Eastern Wielkopolska value mine related work and job opportunities locally to the same extent as has been historically observed in Poland and elsewhere, is an empirical question. Mining has mechanized and automatized over the past decades. Along with it, the skill profile of its workers, and possibly also the link of their identity to mining and the mining community, may have changed.49 Given the different health hazards involved, the mining related identity may also be stronger among deep-coal mining communities (as in Silesia) than among lignite workers (as in Wielkopolska and Lower Silesia). To inform policymakers and investors regarding the potential and desirability of different mine repurposing activities and economic 48 Carley, et al., 2018, Mayer, 2018, Robertson (2006). 49 Findings from the UK, however, suggest that the new technologies in mining that have intensified workplace monitoring and surveillance, have not changed mine workers’ unique sense of identity as coal miners, given their continuing role as autonomous workers that mediate the impact of technology on their working practices (Allsop and Calveley, 2009). 44 diversification strategies, the skills profile and job aspirations of the directly and indirectly affected workforce of Eastern Wielkopolska must be better understood. This also helps develop tailored training and job transition programs and policies for the affected workers and informs other labor market transitions affecting Eastern Wielkopolska, such as the ongoing digitization of the economy and workplace and the transition to more environmentally sustainable economic activities. A new skills and worker job preference survey among Eastern Wielkopolska’s coal-related workers helps shed light. During 2021, a skills and job preference survey was conducted among all ZE PAK workers, i.e., workers directly affected by the coal phase-out, as well as a representative sample of 400 households from four municipalities in Eastern Wielkopolska heavily affected by the coal phase-out (Kazimierz Biskupi, Wierzbiniek, Kleczew and Wilczyn).50 In the latter survey, only non-ZE PAK workers were interviewed. 51 The non-ZE PAK households and workers in these municipalities stand to be indirectly affected given the heavy dependence of their local economies and the local public finance on mining activities. Annex 4 provides detailed information about the survey sampling frames and the sampling as well as the sample selection bias corrections that were applied after the survey. Non- response, and particularly selectivity in non-response, is inherent to any survey. This can undermine the accuracy and representativity of the findings derived from the recorded responses. Using sample bias corrections based on additional information from other sources on key variables for both the respondents and non-respondents, the effect of non-random non-response on the representativity of the findings has been mitigated.52 The survey itself consisted of two components. First, it measured how respondents scored on a range of transversal hard and soft skills and, second, it identified how they value different job characteristics, such as job location, job security, and job compensation, through a discrete choice experiment (DCE). The questionnaire also queried the respondents’ employment status and socio-economic situation to identify heterogeneity in skills and job attribute preferences. The questionnaires for both the ZE PAK workforce and the municipalities were each time field tested and fine-tuned accordingly. They are in Annex 2. 50 See section 3.4 for details. 51 In total, 524 questionnaires were filled by ZE PAK workers, and 404 interviews were conducted among the municipalities’ working-age population excluding ZE PAK workers (109 in Kazimierz Biskupi, 120 in Kleczew, 96 in Wilczyn, and 79 in Wierzbinek). Within each household selected to participate in the municipality survey (simple random selection), all working-age individuals (i.e., people aged 16 to 64 years old), excluding ZE PAK workers, were interviewed. Selected households with only ZE PAK working-age workers were replaced from a pre-selected pool of replacements. 52 All ZE PAK employees received an invitation to participate in the survey; about 1 in 4 participated (or 524 employees). Correction weights were constructed using administrative data from ZE PAK on age, gender, and education. Finally, and where possible, additional adjustments were made to impute missing answers from the 524 ZE PAK participants. For all questions, the response rate was at least 95%. The final results for the municipalities were also reweighted, so that they represent an unbiased sample of the working-age population of the municipalities, excluding ZE PAK employees. 45 4.2 ZE PAK workers similarly skilled as others; workers in affected municipalities, less skilled 4.2.1 Skills measurement and scales The focus is on transversal skills. ILO (2004) shows that individuals are most employable when “they have broad-based education and training, basic and portable high-level skills, including teamwork, problem solving, information and communications technology (ICT) and communication and language skills. This combination of skills enables them to adapt to changes in the world of work.” The better equipped the affected workforce in Eastern Wielkopolska is with these “transversal” skills, the easier the transition to other jobs will be. The skills questionnaire thus focuses on core transversal skills that are not exclusively related to any particular sector or occupation. The analysis is comparative. Rather than assessing the skill profile of the coal-related workers in absolute terms, the report compares them with the rest of the labor force in Poland to assess their comparative (dis)advantage vis-à-vis other workers. Information on universal competencies among workers in Poland has been annually collected through the Polish Human Capital Balance (Bilans Kapitału Ludziego, BKL) survey. It is the largest cross-sectional panel study monitoring the Polish labor market from 2010 to 2022. Self-assessment scales are used, going from 1 (very poor) to 5 (very developed), which will serve as a benchmark for our analysis. Self-skills assessments are easier and less expensive to administer than test-based skill measurement, though the findings on whether they are a good proxy for test-based skill measurement are mixed. By taking a comparative perspective, lingering doubts about the accuracy of self-assessed skills are mitigated.53 The transversal skills of the coal-related workforce are thus assessed by comparison to those of the broader regional and national labor pool, using the same questions and similar survey settings as in the BKL survey. Following skill domains were adopted. They were also queried in the BKL survey and relate to: self-organization and initiative taking, communication with other people, organizing and conducting office work, managerial abilities and organization of other people’s work, fluency in Polish language (oral and written), ability to search for and analyze information and draw conclusions, capacity to maintain assemble and repair technical devices, ability to perform simple calculations, willingness to travel frequently, knowledge of specialized computer programs, ability to code and create websites, artistic and creative abilities, and physical strength. 53 Research by Allen & Van Der Velden (2005), Brown et al. (2015) and Davis et al. (2006) shows mixed conclusions as to whether self-skills assessments can be a good proxy for test-based skill measurement The comparison of the skill profiles among different groups holds, however, as long as potential biases in the self- assessments are unrelated to the group defining features. The report thus mainly operates on the differences in self-assessment, treating them as a relative measure and refraining from treating them as indicators of competencies in absolute terms. 46 4.2.2 Similar skills for ZE PAK workers; lower skills for residents of at-risk municipalities ZE PAK employees display on average similar skills as other Poles. Overall, ZE PAK employees score similarly to other Poles in Wielkopolska and Poland, for most skills, but substantially higher for maintenance and repair skills, and somewhat lower for advanced digital skills (-0.4 point)54 (Figure 19). This pattern holds when controlling for workers’ skill levels (as captured by education and ISCO classification),55 i.e., comparing skill profiles of ZE PAK employees sorted by their educational achievements with those of workers employed elsewhere in Wielkopolska or Poland in occupations requiring similar levels of educations. Figure 19: ZE PAK employees have transversal skills mostly similar to other Poles; workers in at-risk municipalities are less skilled across the board Poland (BKL 2019) Greater Poland Wielkopolska (BKL (BKL 2019) 2019) Municipalities ZE PAK managerial abilities and organization of work of others 4 ART artistic and creative abilities self-organization of work 3 ORGANIZATION organizing and conducting office physical fitness 2 work 1 fluent use of the Polish language 0 contacts with other people in speech and writing FOUNDATIONAL COMMUNICATION performing simple calculations willingness to frequent travels searching and analyzing maintenance, assembly and information and drawing repair of technical devices conclusions knowledge of specialized programs, the ability to write programs or create websites TECHNICAL Source: World Bank Skills and Preference Survey (2021), BKL (2019). Among ZE PAK employees, mine workers display the lowest transversal skills, but they score better on technical skills. Within ZE PAK, miners have the lowest skill profile, far below managers and professional technicians, power plant employees, administrative and repair 54 Both differences are statistically significant. 55 To do so, ZE PAK workers were first divided between those with up to upper-secondary education (low), and those with at least higher education (high). Second, BKL data were split between last 7 groups of ISCO-08 occupations55 (low), and the first two55 (high). The former groups include technicians and associate professionals, craft support workers, service and sales workers, skilled agricultural forestry and fishery workers, craft and related trade workers, plant and machine operators and assemblers, and elementary professions; the latter group consists of managers and professionals. 47 staff. The skills gap is largest for organization at work, digital skills, and simple calculus. On the other hand, miners feel more confident about maintenance, assembly, and repair of technical devices and creative activities. Miners make up about one fifth (18.4%) of the total ZE PAK workforce. To the extent that the self-assessed transversal skill scores provide a good approximation of their skill levels in absolute terms, this suggests that they will face more difficulties in transitioning to other jobs, thus requiring special policy attention. Finally, and most strikingly, non-ZE PAK municipality workers, score lowest on all skill dimensions56 (Figure 19), except creativity. They do not only score worse than workers elsewhere, but also below the ZE PAK workers (including the older ones), who, in addition to enjoying targeted transition support, will thus be better equipped to take up new jobs on the local labor market. This makes non-ZE PAK municipality workers particularly vulnerable to the upcoming transition out of coal, further compounding the effects of the already lower economic dynamism and more limited employment opportunities offered by the communities they live in (section 3.4). 4.2.3 Lowest skills among male municipality workers Male (non-ZE PAK) respondents from the municipalities display the lowest skill sets, except for digital skills (Figure 20). The gap with ZE PAK males is close to 1 point in most dimensions57; the gap with women living in municipalities is smaller, but still statistically ORGANIZATION significant58, with the largest differences in communication and organization. The skills sets for male and female employees in ZE PAK are not systematically lower or higher. They are different. Men have higher technical skills and physical strength; women have better organizational and communication skills. There are also slight differences in foundational cognitive skills (women scoring slightly better), but the differences are not statistically significant. COMMUNICATION FOUNDATIONAL TECHNICAL 56 The lower scores are statistically significant lower for 7 out of 12 the skills examined. 57 The differences are statistically significant for 8 out of 12 skills examined. 58 The differences are statistically significant for 6 out of 12 skills examined. 48 Figure 20: Lowest skills among men, especially those not working for ZE PAK Municiapalities - males Municiapalities - females ZEPAK - males ZEPAK - females managerial abilities and organization of work of others 5 ART artistic and creative abilities self-organization of work 4 ORGANIZATION 3 organizing and conducting physical fitness office work 2 1 fluent use of the Polish language in speech and 0 contacts with other people writing COMMUNICATION FOUNDATIONAL performing simple willingness to frequent calculations travels searching and analyzing maintenance, assembly and information and drawing repair of technical devices conclusions knowledge of specialized programs, the ability to write programs or create websites TECHNICAL Source: World Bank Skills and Preference Survey (2021), BKL (2019). 4.2.4 Overall willingness to reskill to get a new job, especially for hard skills Most workers feel that their skills are suited to their jobs. This holds for 67 percent of the workers in the municipalities and 58 percent of the ZE PAK employees; 27 percent of ZE PAK workers feel that they have the skills to handle more demanding duties. One quarter of women and 10 percent of men indicate that they need further training to do their current job well. For 13 percent of employees, the job performed in ZE PAK is inconsistent with their education and skills, particularly among younger workers (aged 45 or less). The majority of workers (sixty-one percent) would be willing to reskill or acquire new competencies when looking for another job. Consistent with a more traditional learning environment, ZE PAK employees are primarily interested in training that develops hard competences. Sixty-four percent of respondents would be interested in upskilling in hard skills, and almost half would like to upskill in soft skills (assertiveness, communication, etc.) and job search methods. Language courses were the most sought-after training, suggesting that workers are looking to more nationally or globally competitive sectors as a potential source of good-quality jobs. Workers also underlined their willingness to participate in 49 training focused on digital skills – the field where the respondents rate their abilities especially low. 4.3 Job attributes coal-related workers value 4.3.1 The valuation of job attributes requires comparing job packages To properly assess people’s valuation of different jobs and job attributes, they need to be confronted with a clear (discrete) choice between different job packages with clearly defined attributes. Oftentimes, respondents do not or cannot know what future situation they prefer, either because they don't have an answer, or because they do not have enough information to make an informed decision. For example, one may be willing, or not, to move regions for a job, but the answer will depend on many other characteristics of the job offer such as the position, the wage, the type of contract, the benefits, the availability of childcare etc. Hence, instead of asking separately direct questions, such as one’s willingness to move regions, or to change sectors, or to accept a pay cut, the DCE methodology reveals these preferences by asking the participant to choose several times, between two hypothetical options, each of which includes a significant set of characteristics, or attributes. This methodology follows a selection process of probabilistic nature, as described by McFadden (1974). The choice of an alternative that seems more attractive to respondents is determined by a utility that depends on individuals’ observed characteristics and unobserved idiosyncrasies. This process allows modeling respondents’ decision processes and extracting information about preferences (individual utility function parameters) based on pre- composed successive choices in a controlled environment. Figure 21: Example of a choice card Source: World Bank Preference Survey, 2021. The DCE consisted of two designs, one for ZE PAK workers, and one for the four municipalities most affected by the mines closure (Figure 21). Each design consisted of six choice sets, each of them offering two different job descriptions to choose from. These two offers included detailed job characteristics, or attributes: alignment with educational specialization, type of contract, monthly wage, wage increase over the first two years of work, 50 benefit levels, commuting time. The attributes and their levels were based on the literature review and preliminary qualitative work, including focus groups with local stakeholders (including ZE PAK human resource department and trade unions, and social partners in the municipalities surveyed). To prevent cognitive overload, the choice experiment was divided into two designs (DCE 1 and DCE 2) with seven attributes each. The designs had two common, overlapping attributes (net salary and form of employment). See Annex 3 for more detail on the choice cards and their attributes and the econometric procedures applied to estimate the value attached to the different job attributes. 4.3.2 Willing to take up new jobs, with strong preference for stability Direct queries about workers’ disposition towards job alternatives indicate that ZE PAK employees are willing to take up new jobs, but age and lack of suitable local job offers are the two constraints that deter them most (see Figure 22). The vast majority (76 percent) of ZE PAK employees intend to look for a job once they are dismissed, especially those under 45 years old (98 percent). ZE PAK employees seem more mobile than municipality residents (21 percent considered changing jobs in the past 12 months, as compared to 13 percent of those working in the four municipalities); but neither are actively looking for a job (only 5 percent of ZE PAK employees were actively looking for a job or in the process of being recruited for a position; 7 percent of municipality respondents). The main obstacle to find another job is age (43 percent), and lack of local job offers (41 percent). Moreover, 27 percent identify a lack of appropriate contacts and their current level of education as a problem. Despite the common perception that there is a lack of job offers in the area, employees show a high aversion to relocation and even commuting. Only 12 percent of ZE PAK workers would be willing to take up a job in Turek, which is 50 km away or a 50-minute drive from Kazimirez Biskupi; and 42 percent indicated that they would not be interested. Forty percent replied that it depends on the job offer. The declared resistance to traveling or moving for work substantially reduces the resilience of the local labor force to adjust to sudden or longer- term structural changes in the economies. It is also an indication of the broader welfare loss associated with the coal transition beyond the income from coal related jobs itself. The DCE helps probe this further. ZE PAK employees reveal a strong preference for stability, i.e., having a job similar to the one they are holding, in a salaried position, with limited income drop. Nearly 61 percent would look for a position in the same profession, but ZE PAK employees do not discard retraining options (only 12 percent would not consider learning new skills). Most employees have spent their entire career working for ZE PAK and have little experience and appetence for entrepreneurship: less than one-third of ZE PAK employees considered the possibility of self-employment, possibly in construction and renovation, transport, carpentry, electrics, food preparation, and accounting. 70 percent of ZE PAK employees would be interested in a 51 full-time job only; only 27 percent would consider a monthly salary below PLN 4,00059, while 23 percent would accept a monthly wage between PLN 3,000 and 4,00060. Figure 22: ZE PAK employees list age and lack of local jobs as the biggest obstacles to finding a job Age Lack of job offers in the area Lack of appropriate contacts, acquaintances Lack of appropriate certificates Level of education Insufficient experience Need to care of the house Child care Poor health Studying or training Need to take care of family member Need to take care of a farm 0% 10% 20% 30% 40% 50% 60% Source: World Bank Preference Survey, 2021. 4.3.3 ZE PAK employees care mostly about commuting time and job tenure A short commuting time is the most important job feature for ZE PAK employees (see Figure 23). For 95 percent of ZE PAK workers, commuting to work takes less than an hour, for 50 percent, it is up to 30 minutes. The DCE model estimates that an additional hour of commute is worth PLN 1,036, which is equivalent, on average, to one-fifth of a monthly salary at ZE PAK. Women are even more sensitive to extended travel times and value avoiding an additional hour of commute at PLN 1,201 per month. Aversion to commuting is negatively correlated with vulnerability, suggesting that those most at-risk to be affected by the transition are more willing to accept a job further away . Those who claim that the coal transition has a negative impact on their lives are willing to pay less to avoid additional hour of commuting than those who don’t (PLN 477 vs PLN 1,577s per extra hour of commuting). Younger and more educated employees are less averse to commuting; they need to be compensated less than other socio-economic groups (PLN 976 vs 1,318 for an extra hour of commuting). 59 The average monthly wage in the Konin subregion is PLN 4,200 (BDL, access:1.11.2021). 60 It is well below ZE PAK average of PLN 5,500 per month. 52 Figure 23. ZE PAK employees’ preferences – DCE 1 -2300 -1300 -300 700 1700 time to Travel work way) Salary over the (one -1,036 Base: No change next 2 years 20% increase over initial salary 764 40% increase over initial salary 221 Base: Self-employed Labore code based in public sector (permanent) 1,628 Form of employment Labore code based in private sector (permanent) 1,677 Professional preparation, fixed-term employment contract 1,683 Temporary (contract of mandate) in private sector -783 Permanent contract after 6 months, private sector 976 Compliance of Base: Work Related to skills and education (requires entry training) 926 work with education inconsisent with your education and skills (requires Corresponds to skills and education 1,143 complete retraining) Base: No benefits Flexible working time 737 Certified professional courses 797 Transport provided to and from work by bus -233 Benefits Private medical care -318 Childcare services -1,708 Mobile phne or computer 720 Note: a darker color means a significance at 10 percent. Source: World Bank Skills and Preference Survey, 2021. ZE PAK workers have a strong preference against relocating, more pronounced this time among the most vulnerable. They evaluate the financial compensation needed for working abroad, ceteris paribus, at PLN 5,055 per month and for moving to another region within Poland at PLN 2,705, which correspond respectively to a monthly and half a monthly average salary at ZE PAK. Miners expect a slightly higher compensation for working abroad than other ZE PAK workers (by PLN 495). Similarly, older workers and those with lower education require higher compensation to relocate for work. Finally, workers whose household income is entirely dependent on the fuel and energy sector are also reluctant to move places. 53 ZE PAK employees very much value salaried work over self-employment, as well as job security; however, they are ready to take on professional training, especially women, highly-educated, and older workers. ZE PAK employees are not much interested in starting their own business (only 24 percent considered doing so in the past) and tend to prefer any other type of employment (open-ended and term-contracts in the public or private sector). Their preference goes to on-the-job training under a term-contract (for which they would be willing to pay (WTP) PLN 1,683 monthly), followed by a permanent contract in the private sector (WTP = PLN 1,677), and a public sector contract (WTP = PLN 1,628). These preferences are particularly strong among women and older workers (above 45 years of age); however, the oldest workers value public and private sector contracts more than on-the-job training or starting their own business. Those with lower educational attainment value on-the-job training more than any other type of employment, suggesting that they are conscious of the need to improve their competencies. ZE PAK employees prefer to find a position similar to the one they already have, consistent with their competences and experience. They prefer their new job to be consistent with their current level of education and skills (WTP = PLN 1,143), or they are willing to pay PLN 926 to take a job in a similar workplace, even if it requires entry reskilling. Women and workers who find their lives less strongly affected by coal transition are less likely to accept retraining and upskilling. Men are more willing to take up professional courses. The most attractive sector of activity to work in is in renewable energy (RE), for which ZE PAK employees are willing to pay PLN 1,527. The preference for RE is particularly strong among higher-educated workers and younger cohorts, in line with recent trends and changes of attitudes and preference towards greener energy sources among these demographics. Miners have a negative attitude to working in RE but are more likely to accept employment in construction than other ZE PAK professional groups. Other professional groups consider employment in administration, RE, and industrial processing more attractive than mining. Agriculture is the least attractive option, and employees are willing to forego PLN 1,223; this is especially true among younger workers (45 years of age and younger) and those with higher education. Miners are less attached to job security than other groups of workers. ZE PAK employees with higher education, aged 45 years or less, and not miners value safety at work more. The difference in WTP between miners and other worker groups amounts to PLN 3,733, and between those with and without higher education to PLN 2,375. 54 Figure 24. ZE PAK employees’ preferences – DCE 2 -6500 -4500 -2500 -500 1500 3500 Compan y size . 2 Base: Poland (no need to change the Poland (need to change the place of residence) -2,705 Location place of residence) Abroad (need to change the place of residence) -5,055 Base: Not conditio Workin burdensome Dangerous -174 ns g Base: Self-employed Labor code based in public sector (permanent) 1,590 Form of employment Labor code based in private sector (permanent) 2,236 Professional preparation, fixed-term employment contract 2,938 Temporary (contract of mandate) in private sector 354 Permanent contract after 6 moths, private sector 2,096 Base: Mining Construction 765 Industrial processing 1,105 Type of industrial activity Renewable energy (photovoltaics, wind farms) 1,527 Agriculture -1,223 Transport services, repairs, communication -315 Other services (gastronomy, tourism and recreation, etc.) -1,614 Administration (office work) 754 Note: a darker color means a significance at 10 percent. Source: World Bank Skills and Preference Survey, 2021. 4.3.4 Residents of the municipalities simply do not want to move for work Residents of the four municipalities do not want to commute or relocate elsewhere in Poland. For 85 percent of working residents, the travel time to work is less than 1 hour, and for almost every second person, it is less than 30 minutes. The DCE model estimates that an additional hour of commute is worth PLN 5,342 per month, which is more than the average monthly salary in the Konin subregion. Women are even more sensitive to commuting and willing to pay PLN 1,601 more than men to avoid one hour of commute. Interestingly, those with higher education are more sensitive to commuting than lower-skilled individuals. Similarly, municipality residents have a strong preference against relocating: they evaluate 55 the financial compensation for working abroad at PLN 6,252 per month and moving to another region within Poland at PLN 2,862 monthly. Again, women are more reluctant to relocate, they are ready to forego PLN 2,591 more than men to avoid working abroad and PLN 1,941 to avoid relocating to another region in Poland. Respondents whose household income is entirely independent of the fuel and energy sector are more sensitive to relocation than those whose income somehow depends on these sectors’ operation. Residents of the four municipalities are ready to accept lower job entry conditions if it holds prospects for good progression. Residents of the municipalities strongly prefer a job that would guarantee a steep wage increase: they are willing to pay PLN 1,904 to guarantee a wage increase of 40 percent over two years’ time. Figure 25. Preferences of the residents of municipalities – DCE 1 -6000 -4000 -2000 0 2000 time to Travel work way) (one . -5,342 Salary over the Base: No next 2 years 20% increase over initial salary 617 change 40% increase over initial salary 1,904 Base: Fixed-term employment 153 Employment contract for an Contract indefinite period Temporary (contract of mandate) -14 Apprenticeship/internship -794 Base: Work Compliance of Related to skills and education (requires entry training) 1,124 work with education inconsisent with your education and skills (requires complete retraining) Corresponds to skills and education 1,548 Base: No Flexible working time -36 benefits Certified professional courses -1,178 Transport provided to and from work by bus 1,013 Benefits Private medical care 719 Childcare services -1,751 Mobile phne or computer -622 Note: a darker color means a significance at 10 percent. Source: World Bank Skills and Preference Survey, 2021. 56 Residents prefer to find a position with skill requirements similar to the one they currently hold. They prefer their new job to match their current education and skills (WTP = PLN 1,548) or be related to their skills set (WTP = PLN 1, 124). The preference is especially strong for women, older workers, and respondents whose household income is entirely dependent on the fuel and energy sector. Figure 26. Preferences of the residents of municipalities – DCE 2 -7000 -5000 -3000 -1000 1000 Compan y size -1 Base: Poland (no Poland (need to change the place of residence) -2,862 Location need to change the place of residence Abroad (need to change the place of residence) -6,252 work with conditio Workin Base: Not Dangerous ns -1,017 g burdensome Compliance of Base: Work Related to skills and education (requires entry training) 66 education inconsistent with your education and skills (requires complete retraining) Corresponds to skills and education 104 Base: Self- Public sector 224 employed Sector Private sector 995 NGO -117 Base: Mining Construction -773 Industrial processing 645 Type of industrial activity Renewable energy (photovoltaics, wind farms) 473 Agriculture 1,522 Transport services, repairs, communication 917 Other services (gastronomy, tourism and recreation, 538 etc.) Administration (office work) 230 Note: a darker color means a significance at 10 percent. Source: World Bank Skills and Preference Survey, 2021. Municipalities’ respondents value agricultural work more than mining. Contrary to what is observed in the ZE PAK, respondents of the municipalities find work in agriculture more 57 attractive than work in mining; this mainly concerns those whose household income is somehow dependent on the fuel and energy sector. Residents, especially women, do not want to undertake a dangerous job. They would forego PLN 1,294 more than males to avoid it. In addition, respondents whose income is independent of the fuel and energy sector are more reluctant to take on dangerous jobs. These findings give important voice to the affected workforce and help steer repurposing plans and investments to generate attractive employment and develop viable job transition paths. Understanding the strong aversion towards relocation and commuting time, the need for utilizing competencies, which are similar to the rest of the workforce for ZE PAK workers, but substantially less for municipality workers, and the overall willingness to retrain (especially for ZE PAK workers), allow to better identify and generate viable transition pathways compatible with the affected workforce’s skills and job-related preferences. 5 Policy principles for a successful transition of affected workers Mitigating the social and employment impacts of coal mine closures requires adopting systemic and coordinated measures, in accordance with the principles of a just transition, that is to concentrate on geographic areas, industry sectors and workers that are most affected by the energy transition. The precise design of the support package depends on the local context, including the local economy and local labor demand (section 3) and the skills and job-related preferences of the affected workforce (section 4), but also the state of the social protection system, the availability of fiscal resources, and the dynamics of the coal transition, whether rapid or gradual (Cunningham and Schmillen 2021). This section zooms in on the options and principles to operationalize active labor market policies and temporary income support measures to help affected workers and communities in their transition away from coal. 5.1 The existing social protection and labor systems are not fully adequate Existing social protection and labor measures for affected workers include income support and the employment programs and services offered by LOs. Social protection measures for workers vulnerable to the energy transition comprise temporary and permanent income support such as early retirement, available exclusively to the mining sector workers (both coal and lignite miners), severance payments in case of mass layoff61 with amounts proportional to the length of employment, and unemployment benefits for twelve months for those registered with the local LOs. As a last resort measure, social assistance benefits and family allowances provide income support to families with income below a certain threshold. Labor 61 According to the Polish Employment Protection Law, severance payments are mandated for employers employing at least 20 employees if the termination of employment is based on collective dismissal and if the termination of individual employment is due to reasons not attributable to the employee. 58 measures include employment services and programs implemented by local Labor Offices (LO)62 to assist laid off workers to find new jobs. Different forms of income support should be deployed in a sequential fashion, so as not to overlap, and be harmonized across categories of coal workers. According to the employment protection law, ZE PAK Capital Group workers would be entitled to severance payments in case of mass layoff, unemployment benefits upon registration with the local LO, early retirement and outplacement services as agreed with the employer and trade unions. They are not eligible to the social protection benefits under the Social Agreement signed in May 2021, designed for hard coal miners, hence not applicable to ZE PAK Capital Group lignite coal miners and power plant workers. It is important that the overall income support benefits are not too generous to avoid the risk of creating work disincentives as shown in the 1990s’ mine closure in Poland (Ruppert Bulmer et al. 2021). Overall, the generosity of income support for to laid-off coal workers should be compatible with incentives to work and harmonized across categories of coal workers. The quality and scale of labor market programs provided by local LOs may not address the needs of mine workers at-risk of being dismissed. Given the large number of workers at-risk of unemployment in the local labor market (more than 2,000 workers according to ZE PAK estimates), the LO’s implementation capacity is further challenged. The quality of job search assistance, profiling and referral to vocational training performed by job counselors in LOs is already compromised by their high case load. As a result, they may not be able to provide the intensive and individualized assistance that affected workers would need for a successful job transition. Many are already older, worked with one employer only, are reluctant to move, and have high reservation wages and lower foundational skills (section 4). 5.2 Five principles to guide the design of additional social protection and labor support Transition support mechanisms should be comprehensive of both income and re- employment support and inclusive of workers in closely related sectors and geographies. Given the heterogeneity in skill profiles, job preferences, and re-employment readiness across directly and indirectly affected workers, transition support systems must be able to provide the full range of employment services, from individualized counseling, job matching services, support for upskilling and re-skilling to match prospect employers’ needs, financial and non-financial support to start up a small business for those with viable business proposals, income support for hard to re-employ workers and mobility assistance for those willing to accept suitable job offers in other regions in case of lack of limited demand. In addition to the income measures listed above, they should further include complementary 62 Powiat LOs provide employment services to unemployed people, job seekers, and workers at risk of losing their jobs. Based on the jobseekers’ profiles, powiat LOs provide a wide menu of employment services and programs, including job search assistance, job matching, career counseling, identifying, organizing, and financing vocational training, internships, and scholarships, initiating and subsidizing the creation of additional jobs (wage subsidies), and payment of unemployment benefits. 59 social inclusion measures such as health vouchers to support workers’ recovery, psychological counselling, substance abuse and addiction counselling, and rehab support where needed. Importantly, as highlighted in section 3.1, affected workers include not just those employed in the mine conglomerates and power plants, but also those employed in the companies in their supply chain and the workers and their families in the affected municipalities more broadly. Hence, selection rules for targeted programs must be inclusive of workers both directly and indirectly affected. Second, the right balance between income support measures and re-employment services depends on the local labor demand as well as the workers’ characteristics. For example, temporary income support should be considered when there is insufficient local labor demand and when local labor markets are highly depended on mining-related activities. In such contexts, the severance payments and unemployment benefits may not provide enough protection in case of longer-term unemployment spells. Mobility assistance reimbursing the cost of moving to a different region based on a job offer (as implemented in Romania) could be more cost-effective than investments in employment services and training in the region. As not everybody is willing to move and mobility constraints are higher among older miners, it is generally worth targeting it to specific groups, for instance younger coal miners. Third, services must be sufficiently flexible, timely and tailored to the particular needs of the affected workers. International good practice shows that pre-layoff assistance, individualized counseling and sophisticated tools to profile and assess the skills and constraints of affected workers are key determinants of success in supporting job-to-job transitions of coal workers. The transition experience in Germany (Ruhr Valley) and more recently of the government of Alberta, Canada, through its Climate Leadership Plan,63 show that the provision of job-search services should begin as soon as the workers receive notice or even before in order to improve the effectiveness of such measures (OECD, 2018).The pre- layoff assistance is intended to inform and prepare workers ahead of layoffs. Typical services implemented as part of pre-layoff assistance include: (i) establishing strong communication with workers to determine their expectations, attitudes, sectoral and mobility preferences; (ii) worker profiling and skill assessments to help understand workers’ skills and training needs and their readiness to assist them with the right active measure based on the profile; and (iii) the provision of initial in-plant job counseling and job placement services for those who are readily matched to existing job opportunities. Regular employment services in Poland typically do not have the flexibility to engage with affected workers in such an individualized and timely manner. Fourth, ad hoc active labor market measures should be coordinated with investments in economic and local business development and related labor and skill needs, in the short and medium term. The job counseling, job search assistance and the referral system should 63 The Plan introduced an accelerated phasing-out of coal-fired power generators and the introduction of a carbon price. 60 be informed by the analysis of labor market demand in terms of skills and occupations. The provision of training could be delivered in class-based and on-the-job, directly by hiring employers. In case of class-based short-term vocational training it will be important to align the selection of courses to the occupation in higher demand and in high productivity sectors. Section 6 showcases how a newly developed skills-matching tool that has been tailored to the Polish labor market, could help with identifying and developing viable transition pathways for affected workers and, relatedly, inform investors about the skills and job preferences of the available workforce and guide course development based on the assessed re/upskilling needs. Finally, as a general principle, labor market programs targeted at at-risk workers because of the energy transition should avoid displacement effects on other “regular” unemployed, notably youth and long-term unemployed. The provision of public employment services and programs should not come at the cost of displacing other vulnerable unemployed. It will be critical to avoid exacerbating labor market distortions via generous outplacement packages to mine workers who already benefit from higher wages compared to workers in other sectors, widening existing labor market distortions risks slowing the pace of adjustment and local economic recovery. Overall, the generosity of temporary income support for to laid-off coal workers should be compatible with incentives to work and job search and harmonized across categories of coal workers. 5.3 Five principles for successful implementation First, the delivery of support services to ensure a just transition requires the involvement of multiple stakeholders. Strong public-private partnerships and direct dialogue across different levels of government, the private sector, trade unions, NGOs and the CSOs are key elements of success to mitigate the social labor impacts of the coal transition. Employers play a key role to make information on redundant employees available and to deliver pre-layoff services to their workers: when an employer is simultaneously laying off workers and planning to invest in land repurposing projects, he could also play a role in identifying and training workers who could be re-employed, synchronizing the stages of employment reduction and jobs creation. If local LOs’ capacity to deliver individualized counseling, job-search assistance and job-matching services before and after layoff is limited, including because of high case load or limited skills of job counselors, outsourcing to private service providers should be considered. Local LOs are often leveraged to establish arrangements with training providers and to provide regular active labor market programs (ALMPs), such as subsidies for starting a business, job placement, internships/wage subsidies, if the services are deemed suitable to affected workers. Second, establishing a one stop-shop and a referral system may be needed to coordinate the range of services available to affected workers, based on a holistic needs assessment. The one stop-shop could be a separate employment center or an office within the premises of the employer or within the local LO. Under the same roof, affected workers will be 61 informed on types of interventions that are provided as part of the comprehensive package, they will be assessed, profiled and referred to providers of suitable services based on the worker’s profile categorization (skills, attitudes and readiness to work) and/or provided with training vouchers. A referral system would need to be developed to lay out the pathways and operational steps to access support services. Third, the communication on how governments are planning to manage the potential layoffs is critical to ensure the coal transition path is correctly understood and to avoid perception of unfairness. The communication around the alternative measures targeted for affected workers needs to be carefully managed. A communication plan on the outplacement options for redundant workers should be prepared and implemented in affected municipalities once the outplacement program is set and funds are secured. Fourth, financing instruments should be mobilized upfront, before the actual mine and power plants close. Mitigating the social and employment effects of the coal transition has both short-term and long-term costs that should be taken into account and budgeted early on. As EU Member State, Poland could access several funding sources: the Just Transition Fund, the European Social Fund and other European funds. These could be used to complement the national state budget from the Labor Fund and the Guaranteed Employee Benefits Fund. Securing the necessary financing instruments to implement targeted employment services and benefits in a timely manner will be important for timely delivery of services. Finally, a strong monitoring and evaluation system - planned in advance and with clear feedback mechanisms- would allow for continuous learning and program adaptation. Local labor offices have helped manage the transition of workers from Adamów open-pit since 2018, but little is known about the employment status of dismissed workers, whether they found another employment or left the labor force, and their socio-economic status. Going forward, defining and agreeing on key roles and responsibilities for monitoring and evaluation activities on a regular basis will be fundamental to inform programming choices and investments. 62 6 Assisting the transition with an AI-powered job-matching tool 6.1 Ingredients needed to assess the viability of individual job and labor force transitions The (local) availability of jobs, workers’ job preferences, and the extent of skills’ (mis)match, determine the viability of job transitions for Eastern Wielkopolska’s coal-related workforce. Section 4 shows that most ZE PAK workers have similar transversal skills as other workers in the region. At the same time, they are also more competent technically, but less skilled digitally. Municipality workers are less skilled across the board. All workers put a high value on being able to work nearby, in positions that provide job security. They prefer jobs that match their skills but are open to re/up-skilling. Assessing the extent of the coal transition challenge for laborers thus requires assessing the spatial, expectations and skills mismatch, i.e., 1) assessing the availability of alternative job opportunities and their location, 2) determining their attributes (salary, contract type, sector, benefits, …), and 3) identifying the skills required and the workers’ ability and attitude towards re/up-skilling. Trade-offs across these dimensions further complicate the identification of viable job transitions. Most jobs, whether currently available or only available in the future, are unlikely to score equally well on all three dimensions (local availability, job preferences, skills (mis)match). Systematically mapping out the trade-offs will be equally important to assess the individual and overall labor transition challenge. A job may be available, for example, but only in another region, or only under a self-employed as opposed to a salaried contractual arrangement. The better a job scores on each of the three dimensions, i.e., the smaller the spatial, expectations and skills mismatch, the more viable is the transition. At the aggregate or regional planning level, the more well matching jobs that there are already available, that will be generated through appropriate repurposing of the mines or economic diversification, or that could be filled through well designed and well targeted re/up-skilling programs, the smoother will be the overall labor transition. Yet, as highlighted in Box 1 and illustrated in detail for the Appalachia region in the United States, in the past, the labor transition has proven to be an uphill battle for many mining communities.64 Seemingly intuitive job transition pathways often prove unsuccessful. Job transitions considered by stakeholders are regularly restricted to different occupations within the same sectors of activity (e.g., reskilling a belt maintenance technician into a photovoltaic installer), or similar occupations in another sector of activity (e.g., offering a school bus position to a dismissed shuttle car operator). ZE PAK’s recent experience of reskilling workers within the energy sector from brown to green jobs has shown limited take-up, however, mostly due to the reluctance to commute, the limited ability of older displaced workers to retrain in a new occupation, and limited career prospects in the medium-to-long term. Such experience is not 64 In Appalachia, a longstanding coal mining region in the United States, only 4 out of the 420 coal counties covered by the Appalachian Regional Commission managed a successful socio-economic transition out of coal (Labao et al., 2021). 63 uncommon. As Michael Bloomberg said at the 2014 Bloomberg Energy Summit: " You’re not going to teach a coal miner to code. Mark Zuckerberg says you teach them [people] to code and everything will be great. I don’t know how to break it to you . . . but no. ” 65 Box 4: Retraining ZE PAK workers for opportunities in Renewable Energies (RE) ZE PAK started its process of collective redundancies in 2020. To facilitate the transition of dismissed ZE PAK workers to RE opportunities, two three-days specialized trainings were organized to provide participants with photovoltaic installation fitter qualifications. Over 300 workers about to be dismissed were offered the training, 50 registered. They all graduated from the course and 21 graduates were short-listed for a position. In the end only 14 ended up accepting a job offer: 5 as photovoltaic systems electricians, 6 as photovoltaic installers, and 3 as construction engineers. Although the training quality was high, few displaced workers transitioned to photovoltaic installers. Part of the reasons why the installer job offers were not appealing is because of the long distance (up to 120 km from Konin) to reach the photovoltaic installations every day. Second, trained employees were too old to work as installer in a photovoltaic farm. Finally, the photovoltaic sector does not offer long-term sustainable job opportunities: job demand is high at the beginning to build the infrastructure, but then the market and job demand disappear, as maintenance requirements are limited. Source: Honorati, 2022. Obtaining timely and systematic information on the spatial, expectation and skills mismatch of alternative jobs and adjudicating potential trade-offs is challenging. Hiring intentions of employers captured through national employer surveys are frequently used to assess labor demand trends across occupations, sectors, and locations (the number and types of jobs). They provide a powerful source to assess employment opportunities in the nearby future (typically the next 6-12 months), complementing the job vacancies published by public and private employment services. Yet not everywhere are employer surveys implemented on a regular basis, and when administered, the results may overlook employment opportunities provided by micro and small enterprises, they may not be representative at local/district level, or they are not publicly available. Employer surveys also give little guidance regarding future jobs emanating from mine repurposing or regional diversification strategies, either directly, but also indirectly in the supply chains, or through consumption linkages in the broader community. This further complicates proper tailoring of re/up skilling programs to the needs of the local labor market, while generic re/upskilling programs (unlinked to particular job opportunities) often prove to be rather ineffective (European Commission, 2012). Second, to assess job preferences regarding different job attributes, DCEs can be used, but they are typically not widely or systematically applied. The potential of the DCE technique to assess job-related preferences/expectations has been illustrated in the context of coal- 65 https://gigaom.com/2014/04/09/michael-bloomberg-you-cant-teach-a-coal-miner-to-code/ 64 related workers in section 4. How these insights can then be used to help delineate the set of viable job alternatives is elaborated further in this section. Such information is, however, typically not readily available to policymakers, while the more widely used direct questions about attitudes towards different job attributes are deficient. As a result, worker preferences regarding job attributes are often insufficiently accounted for, neither in the planning of re/upskilling programs, nor in the development of repurposing activities or regional diversification strategies, burdening their chances of success as illustrated by the limited job placements following ZE PAK’s renewable energy retraining program (Box 4). Thirdly, and most challenging, is to assess the extent of skills (mis)match across jobs. It has also attracted a great deal of attention among policymakers. A pre-condition for a viable job transition is the worker’s capacity to execute the tasks required (with or without re/upskilling). While this may be easier to assess on a case-by-case basis, from a planning perspective, it is also important to know how the skills of the workforce at risk of dismissal, match with the skills needed in currently available jobs or future alternatives. To assess this requires having a similar skills metric that can be used to both describe the skills profile of the current labor force and the skills needed to perform the currently available or future jobs. This is data and computationally intensive. New artificial intelligence (AI) text mining algorithms are now starting to be used to do so. WEF (2018) and OECD (2021), for example, used online job postings collected by Burning Glass Technologies (BGT) and artificial intelligence (AI) text mining algorithms to analyze similarities between pairs of occupations, in the US (WEF, 2018), and in EU-27 Member States, Australia, Canada, the United Kingdom, the United States, and New Zealand (OECD, 2021), respectively. Online job postings provide real time data and the most up-to-date vacancy descriptions (tasks and skills requirements, location, contract type, salary, etc.), but they may also skew the analysis towards higher skilled occupations and sectors, both of which are more likely to use online channels to advertise job postings. 6.2 Towards a “viable-job-matching” decision tool tailored to the Polish labor market Publicly available Polish and location specific data on occupations and new jobs, DCE information on job preferences for coal-related workers and big data techniques are used to assess the extent of viable job transitions for the coal-related work force as present in the current labor market. Particularly, the section develops and illustrates how a new job matching decision tool anchored in location specific data can help individual case workers, mining conglomerates, public employment services (PES) and local authorities prioritize their actions, time, and investments, including to develop tailored re/upskilling programs. It maps out job transitions opportunities for workers at-risk of losing their jobs in the mining and energy sectors, but the methodology can be applied to anyone looking for another job, looking to upskill and/or improve their wage prospects and job satisfaction. The method can also be readily adapted to different contexts and demand projections, including those of new investors or related to repurposing activities. 65 To identify viable job transitions, the following three constraints were considered : (i) the amount of reskilling due to the transition between two occupations was minimized, (ii) only occupations in growing demand on the local labor market were considered, and finally (iii) potential candidates were further screened using the results of the preference survey described in section 4. First, to assess the skills (mis)match across positions, the task-similarity between all professions was assessed. Job descriptions (including those held by workers at risk through the coal transition) were extracted from the Polish Labor and Labor Offices Job Postings Office using the Selenium web scraper. This resulted in a list of 2.7 thousand key occupations. Occupations were then compared one by one based on their task content, using Latent Semantic Indexing (LSI), and a (task) similarity score (between 0 and 1) was constructed for each pair of occupations (yielding a 2.7k by 2.7k similarity score matrix). A similarity score close to zero signifies two occupations with very little overlap in terms of tasks to perform; a similarity score close to one signifies two occupations with large overlap. A career switch between two occupations with a low similarity score is likely to require substantial reskilling, or may even be impossible, while a career switch between two positions with a high similarity score may require limited retraining and may also be much more feasible, at least from a technical capacity point of view. The similarity score index thus allows to identify occupations which are very similar in task content to those currently conducted by the coal-related workforce, suggesting a high probability of a good match, provided that they meet the expectations of the workers and that they are available on the (local) labor market. For robustness and refinement, an additional metric to assess skill similarity across occupations was introduced. In particular, the types of tasks required by each position were divided into five skills groups: non-routine manual (NRM), non-routine interactive (NRI), routine manual (RM), non-routine manual (NRM), routine cognitive (RC). This classification builds on the measure of routine and non-routine task content developed by Mihaylov and Tijdens (2019), who provide a measure of the Routine and Non-Routine Task Content of 427 Four-Digit ISCO-08 Occupations. Each task in an occupation is assigned to one of the 5 groups (NRM, NRI, RM, NRM, RC). Indexes are created by dividing the number of tasks belonging to a given category by the number of all tasks performed in a given occupation. Then, the occupation is classified to one of the 5 skill groups, i.e., the skills group whose tasks accounts for the largest share in that occupation. The results are reported at the 4-digit level and were integrated in the dashboard, defining the top similar professions (based on the task-similarity score) across those five dimensions. Second, job transitions were restricted to positions with excess demand on the local labor market. Poland does not administer an employer survey on a regular and highly disaggregated basis that could be used to locally investigate labor demand. Yet, it annually produces a more qualitative assessment of labor demand through the so-called occupational barometers. These are developed by the powiat labor offices and used here to identify jobs 66 with excess demand,66 in order to restrict job transitions to occupations in growing demand, or with high potential (Górna – Kubacka, Komasa and Rybak, 2020).67 Finally, the viability of the options is further checked against the value assigned to different job attributes as revealed through the DCE. The DCE results of Section 4 were used to further scrutinize the opportunities short-listed by the similarity scores and the occupational barometers. For instance, ZE PAK workers display a strong aversion to commuting or moving regions. The subset of job offers that were not present locally and implied moving to another part of the country were scrutinized: if the wage increase was below ZE PAK employees’ willingness to pay, the option was discarded. The expected wage for each job was based on the average reported in the Labor Force Survey (2018). For more detail on the method and application, see Annex 5, which also provides screenshots of the data presented in the designed online tool used to develop the different pathways. In what follows, the job with the highest task-similarity score in each of the 5 skill groupings is featured and their viability further assessed based on their availability in the local market (as indicated in the occupational barometers) and the wage such a job earns on average (as reported in Labor Force Survey, 2018). 6.3 Five transition pathways relevant to the mining and energy sector Five positions typically found in the mining and power sector were selected, five associated transition pathways identified, and their viability discussed. Using the methodology described in Section 6.2, five potential transition pathways are each time identified for each position, one for each of the 5 skills grouping discussed (non-routine manual, non-routine interactive, routine manual, non-routine manual, routine cognitive), implying different reskilling efforts needed. Within each skill group, the occupation with the highest task similarity score was retained to capture the most similar job in terms of task to perform. The potential wage differential associated with the career change, and the availability of those jobs locally are also discussed. Most trajectories imply reskilling, and some involve pay cuts, in the short-term, but wages should increase in the longer-term given that all occupations are in growing demand. Most trajectories do not point to retraining and upskilling in renewable energies, the focus of much of the regional diversification efforts, or in digital jobs, often considered the jobs of the future. In fact, the transition from coal-based energy production to renewable energy can often be complicated by the complexity of retraining and upskilling (can dismissed workers be reskilled, and are they willing to?), fundamental differences between workplace 66 Occupational barometers of occupations in surplus, balance or deficit, produced by local Labor Offices were used here to identify occupations in excess demand, but the methodology can be adapted to use other sources of labor market forecasting. The occupational barometers represent a qualitative consensus view of local labor market stakeholders, including local labor offices, the industry and their chambers of commerce, and civil society (labor unions, academics, etc.). 67 https://barometrzawodow.pl/#wielkopolskie 67 characteristics (is the nature of the work aligned with displaced workers’ expectations and preferences), and/or lack of such opportunities on the local labor market (especially in the case of dismissed workers with low geographic mobility). The transition from a mechanic of opencast mining machinery and equipment to fitter of renewable energy devices displays a score of 0.54, which is relatively low. Similarly, the transition from a mining engineer – lignite mining to an engineer of equipment and renewable energy systems displays an even lower similarity score of 0.36, suggesting limited overlap of skills (at least as captured by task similarity). Finally, the transition from a technician of opencast mining and equipment to a technician of renewable energy systems displays a similarity score of 0.26, suggesting that this option would not utilize most of the skills and experience acquired by the displaced workers. 6.3.1 High-skilled white-collar: energy engineer Due to the high degree of specialization of an energy engineer, most transition pathways display low similarity scores, and lower salaries (Figure 27). An energy engineer mainly performs non-routine analytical activities, and rarely executes non-routine interactive tasks. The profession’s average salary is high, and equal to PLN 4,696 per month (BDL, access: 1.12.2021). Three out of five proposed transitions offer a lower salary than an energy engineer’s average monthly salary. Figure 27: Examples of transition pathways for an energy engineer The position with most task overlap is a mechatronics technician, for which the similarity score is 0.6. individuals willing to move to this profession will need to accept a lower salary 68 (PLN 3,744). In this non-routine manual profession, equipment maintenance using modern techniques are especially valuable. This transition pathway was however disregarded as it implies a monthly pay decrease of PLN 953, and a decrease in position prestige from engineer to technician. Ship mechanic technician and chemical reactor controller display reasonably similar task- similarity scores (respectively 0.42 and 0.36), but neither of them is in demand in Wielkopolska. In the case of ship mechanic technician, the results of the preference survey show that the positive wage differential (+PLN 7,640 per month) would be large enough for a dismissed engineer to move to the Polish coast, but the position was discarded as it implies a demotion from engineer to technician. The transition to computer network operator was also discarded as it is associated with a lower monthly pay (-PLN 1,078). The most reasonable transition pathway appears to be a telecommunication engineer. Reskilling is moderate, with a similarity score of 0.40. The energy engineer will need to be able to conduct research, supervise technical modernization of works, introduce digital technology, and implement new solutions, measure, and test equipment. The switch to telecommunication engineer is associated with a monthly wage increase of PLN 2,167, or almost a 50 percent increase. 6.3.2 High-skilled blue-collar: lignite mining technician Due to the large wage premium in the mining sector, most transition pathways for a lignite mining technician imply a large salary decrease (Figure 28). A lignite mining technician mainly performs non-routine analytical activities, and rarely executes non-routine cognitive tasks. The profession’s average salary is high, and equal to PLN 5,596 per month (BDL, access: 1.12.2021). All the proposed transitions offer at least a 50 percent wage decrease, which mostly reflects the high wage premium associated with being a skilled technician in the mining sector. 69 Figure 28: Examples of transition pathways for a lignite mining technician The position with most task overlap is a papermaking technician, for which the similarity score is 0.71. It is also the transition that implies the lowest wage drop, with an average monthly salary of PLN 3,558. In this non-routine manual profession, the worker is supervising 6.3.3 Low-skilled blue-collar: truck mechanic All transition pathways for a truck mechanic are available locally, and involve a slight wage increase and limited reskilling (Figure 29). Truck mechanics mainly perform non-routine manual activities, and rarely execute non-routine analytical tasks. The profession’s average salary is equal to PLN 2,643 per month (BDL, access: 1.12.2021). All proposed transitions offer an equivalent or higher salary. The most reasonable transition pathways are tractor mechanic and fitter of internal combustion engine. The position with most skills overlap is a tractor mechanic, for which the similarity score is 0.99, and offers the same salary. With a similarity score of 0.85, one other position requires limited retraining (fitter of internal combustion engine), with a slight wage increase (PLN 2,995). The other three position require more reskilling (similarity scores between 0.41 and 0.56), with slight wage increases. 70 Figure 29: Examples of transition pathways for a truck mechanic 6.3.4 High-skilled blue-collar: electrician Transition pathways for electricians involve very limited reskilling, and limited wage changes (Figure 30). Electricians mainly perform non-routine manual activities, and rarely execute routine cognitive and non-routine analytical tasks. The profession’s average salary is equal to PLN 2,889 per month (BDL, access: 1.12.2021). All proposed transitions offer salaries within a PLN 500 range difference. The position with most skills overlap is electromechanical engineer. This position has a similarity score of 0.95 and offers a slightly higher salary. In the new position, the engineer will have to install, assemble, repair electrical devices using electric installation and locksmith tools, power tools and basic electrical meters. 71 Figure 30: Examples of transition pathways for an electrician 6.3.5 Low-skilled blue-collar: electrical machine fitter Transition pathways for an electrical machine fitter involve limited reskilling, and no wage changes (Figure 31). Electrical machine fitters mainly perform non-routine manual activities, and rarely execute routine cognitive and non-routine analytical tasks. The profession’s average salary is equal to PLN 2,625 per month (BDL, access: 1.12.2021). All proposed transitions offer similar salaries. The position with most skills overlap is electrical engineer, for which the similarity score is 0.83, and offers a slightly higher salary (PLN 3,129). 72 Figure 31: Examples of transition pathways for an electrical machine fitter 6.4 Demonstrated proof of concept, but further validation and development needed. As such, this section has demonstrated the power of data-driven approaches for finding solutions to job disruptions more broadly, including job transition pathways and reskilling opportunities that might not be immediately apparent. To be fully operational, further ground truthing and data manipulations will be needed, including to better inform the adjudication of trade-offs and to operationalize it as a planning tool. This section has focused instead on showing proof concept and illustrating the tool with a series of examples of individually viable transition paths as observed in the current labor markets and given the preferences of the coal-related labor force in Eastern Wielkopolska. The underlying findings are also available as an online dashboard for further manipulation. 7 Conclusions A geographically focused, or local economic development approach to just coal transition, rather than a sectoral approach, is called for. Climate change is forcing a transition to carbon neutrality, putting many jobs at risk, in and especially around the mines. As Europe’s largest coal producer, Poland is at the forefront of the European coal transition. Within Poland, Eastern Wielkopolska is the region most advanced in the transition out of coal; finding viable job transitions is of imminent importance. As of February 2022, ZE PAK Capital Group, one of the country’s largest lignite conglomerates, had closed most of its extraction sites, the last 73 three power plants were planned to shut down by 2024, and all lignite mining activities in KWB Konin by 2030. Workers in fossil fuel sectors and carbon intensive industries, and the towns68 and regions where they are concentrated, will be hardest hit. They are at the forefront of the transition. The energy sector can potentially be a catalyst for regional development and employment, but the path to achieving this economic stimulus role is not certain. Globally, new jobs in transition-related technologies and sectors are expected to outweigh job losses in fossil fuels and nuclear energy (IRENA, 2021).69 Conversion into wind or solar parks, for example, could arguably provide re-employment opportunities for some coal workers after an adjustment of skills, since electrical and mechanical skills, experience of working under difficult conditions and sophisticated safety experience are in demand in the wind and solar energy industries. Similarly, the re-use of closed mines for geothermal energy or hydropower applications could provide jobs and socioeconomic benefits to post-mining communities. However, the distributional impact of the transition may not guarantee successful job transitions across all skill sets: medium-skilled occupations display the highest opportunities for retraining and upskilling measures for the green transition, but low- and high-skilled workers are likely to find fewer opportunities (ILO, 2018). The success of these transitions within the mining and energy sectors are also subject to the jobs-related aspirations of the dismissed workforce and their willingness to participate in retraining and upskilling programs. More broadly, the availability of viable job transitions will depend on the extent to which new economic activities can be developed that make maximum use of the existing skill pool, inside and outside the energy sector, and including through appropriate re-purposing of the mines and their assets. Analysis combining large and representative datasets, econometric techniques, and machine learning can be used to develop well-tailored, realistic, and acceptable retraining and reskilling programs. Firstly, the discrete choice experiment (DCE) showed how voice can be given to displaced workers, whose preferences are typically ignored, partly because it is challenging to capture them adequately. Workers likely to be affected by the coal transition express strong mobility aversion (both to relocation and longer commuting time), consistent with the strong preference for in situ job creation, a sentiment also voiced during stakeholder consultations. The DCE results also showed how workers’ relatively high reservation wages could be addressed by offering jobs with a set of desired non-pecuniary benefits such as a clearly defined career development path (with earnings increases) and jobs compatible in skills and 68 Especially Kazimierz Biskupi, Wierzbiniek, Kleczew and Wilczyn, which display already weak labor market indicators, and are highly vulnerable to changes in ZE PAK’s labor demand. 74 education, or employment sector and type of contract. Despite significant heterogeneity in terms of job attribute preferences, the vast majority of affected workers demonstrate an overall willingness to work and reskill as needed. Secondly, leveraging machine learning and online job descriptions allowed to identify plausible transition pathways for workers affected by the transition. Incorporating the DCE results further helped to narrow down available and acceptable options that not only imply good matches in terms of skill and task similarity but that are also in excess demand in the local labor market and within the range of desired job attributes. Oftentimes, job transition pathways considered by stakeholders are unrealistic. The possibility of a transition from brown jobs to green occupations, for example, is often treated overly optimistically in the transformation process, but trajectories to retrain and upskill the dismissed workforce to renewable energy jobs display low similarity scores, suggesting that displaced workers would not utilize most of the skills and experience they had acquired pre-transition. Most viable transition pathways identified for typical occupations at ZE PAK are in fact in mechanics, with easier transitions for non-sector-specific workers, and more difficult transitions for higher- skilled specialized employees. The discrete choice experiment (DCE) run for miners and other workers in the mining towns will be essential to design tailored coal transition policies that leave no one behind. Following best practice principles in the design and implementation of complementary social protection packages can increase the likelihood of success. When local labor markets are hit by the withdrawal of a core employer or sector, such as mining, existing social assistance and labor market programs and the capacity of local employment offices may be insufficient to address the needs of the directly and indirectly affected workers. A number of principles related to the design and implementation of the complementary support programs can make them more effective. In particular, support packages should be comprehensive (including for example psychological support), properly balanced between income support and re-employment services, tailored to the needs of the affected workers and deployed timely (starting pre-layoff), coordinated with job generating investments to align training and reskilling programs, and not displace other regularly unemployed such as youth and long term unemployed. Implementation will be more effective if the different stakeholders are involved throughout and properly informed, if services are well coordinated and communicated, often involving the set-up of a one-stop shop, if labor offices are adequately staffed, or leveraged by private employment agencies, if financing is secured in a timely manner, and if implementation is accompanied with a strong monitoring and evaluation system with adequate feedback mechanisms. More broadly, the methodologies, techniques and policy principles presented above could also be used to assist public officials in designing effective labor market transition plans during factory closure in single employer dominated labor markets. 75 First the techniques and findings that emerged from the DCE on coal related workers can also be used to inform other labor market transitions, such as the ongoing digitization of the economy and workplace and the transition to more environmentally sustainable economic activities.  Doing so requires further ground truthing and operationalization of the tools. More broadly, the DCE and job matching tool illustrated here in the context of the just coal transition are just an example of the type of methodology that could be used in similar contexts of large-scale lay-offs in areas with single employer dominance. Second, the identification of viable job transition pathways demonstrates the potential power of data-driven approaches for finding solutions to job disruptions, including reskilling opportunities that might not be immediately apparent. Job opportunities were mapped out for workers at-risk of losing their jobs in the mining and energy sectors, but the methodology can be applied to anyone looking to upskill and improve their wage prospects and job satisfaction: the methodology is not limited to the geography and sector presented here, and can be feasibly adapted to different contexts and demand projections: instead of matching displaced workers with current opportunities as identified by local Labor Offices, labor market forecast, development strategies, or investors’ interest in moving to the region (including Special Economic Zones) can be fitted in the job transition pathway tool developed. Despite potential delays in the closure of Poland’s coal mines in the short run, in view of the Russian invasion of Ukraine, the diagnostic and methodology developed for this report remain valid. The transition away from coal is as relevant as ever. With Poland already importing a fair amount of coal from Russia, the Russian invasion is likely to accelerate the development of carbon neutral energy sources in the medium run, while securing energy supplies with the use of coal in the near-term.70 In the short-run, the European Commission proposed to reduce Poland’s dependence on natural gas and coal imports from Russia, extending the lifetime of coal-fueled blocks, while accelerating the rollout of renewable energy sources.71 The changes in energy policy, as well as the influx of Ukrainian refugees may have an impact on an already tight labor market.72 The key findings emerging from the skills and preference surveys are expected to remain unchanged, however, in the near term: skills similarity for ZE PAK workers and lower skills among non-ZE PAK workers in the affected municipalities, reluctance to commute or relocate, preference for job security, willingness to switch sectors if employees can keep deploying their skills, willingness to take up lower paying job when there are credible wage growth prospects, etc. The matching tool developed to identify optimal transition pathways can be adapted to include new occupations that may arise in the near future, or otherwise respond to changing labor demand. Overall, the imperative to take a local economic development as opposed to a sectoral perspective, 70 https://www.iea.org/news/how-europe-can-cut-natural-gas-imports-from-russia-significantly-within-a-year 71 https://www.politico.eu/article/coal-not-taboo-as-eu-seeks-russian-gas-exit-says-green-deal-chief/ 72 The initial wave of refugees, constituted mainly of women and children, is concentrated in the capital city of Warsaw. 76 inclusive of the affected non-ZE PAK workers in the affected communities, will remain unchanged. 77 Bibliography Allen, J. P., & Van Der Velden, R. (2005). The role of self-assessment in measuring skills. Allsop, D., and M., Caveley, 2009. Miners’ Identity and the Changing Face of the Labour Process Within the UK Coal Mining Industry. Qualitative Research in Accounting and Management 6,1/2: 57-69. Alves Dias, P., K. Kanellopoulos, H. Medarac, Z. Kapetaki, E. Miranda Barbosa, R. Shortall, V. Czako, T. Telsnig, C. Vazquez Hernandez, R. Lacal Arantegui, W. Nijs, I. Gonzalez Aparicio, M. Trombetti, G. Mandras, E. Peteves, and E. Tzimas, 2018. “EU Coal Regions: Opportunities and Challenges Ahead,” JRC 112593, Publications Office of the European Union, Luxembourg. Brown, G. T., Andrade, H. L., & Chen, F. (2015). Accuracy in student self-assessment: directions and cautions for research. Assessment in Education: Principles, Policy & Practice, 22(4), 444-457. Baran, J., A. Szpor, J. Witajewski-Baltvilks, 2020. „Low-carbon transition in a coal-producing country: A labour market perspective,” Energy Policy, 147, 111878. Barometr zawodów, https://www.barometrzawodow.pl/#wielkopolskie, access: 1.12.2021. Carley, Sanya, Tom P. Evans, and David M. Konisky. 2018. “Adaptation, Culture, and the Energy Transition in American Coal Country.” Energy Research & Social Science 37 (March): 133–39. Churski P., R, Perdał, M. Burchardt, 2021. Czasoprzestrzenna analiza zmian lokalnego rynku pracy podregionu konińskiego w konsekwencji sprawiedliwej transformacji,” Uniwersytet im. Adama Mickiewicza w Poznaniu, Poznań. Cunningham, W. and A., Schmillen. 2021 The Coal Transition : Mitigating Social and Labor Impacts. Social Protection and Jobs Discussion Paper 2105. World Bank, Washington, DC. Daly, A., S. Hess, K. Train, 2012. “Assuring finite moments for willingness to pay in random coefficient models,” Transportation, 39(1), 19-31. Davis, D. A., Mazmanian, P. E., Fordis, M., Van Harrison, R. T. K. E., Thorpe, K. E., & Perrier, L. (2006). Accuracy of physician self-assessment compared with observed measures of competence: a systematic review. Jama, 296(9), 1094-1102. Deville, J.C., and C.E. Sarndal, 1992. “Calibration Estimators in Survey Sampling,” Journal of the American Statistical Association 87: 376–382. EBRD, 2020. “Insights from Historical Cases of Transition,” Background Paper for the EBRD Just Transition Initiative. EURACOAL, 2020. “Country Profile – Poland,” https://euracoal.eu/info/country- profiles/poland [accessed on 11/29/2021]. 78 European Commission, 2012. “Adapting PES training policy to better service demand. Small scale Study 2012.” European Job Mobility Laboratory. Eurostat, 2022. “Coal production and consumption statistics,” https://ec.europa.eu/eurostat/statistics- explained/index.php?title=Coal_production_and_consumption_statistics# [accessed 13 May 2022]. Frankowski, J., J. Mazurkiewicz, J. Sokołowski, P. Lewandowski, 2020. „Zatrudnienie w górnictwie węgla kamiennego w Zagłębiu Górnośląskim,” IBS. Górna – Kubacka A., M. Komasa, M. Rybak, 2020. “Barometr zawodów 2021,” Wojewódzki Urząd Pracy w Krakowie. https://arrtransformacja.org.pl/wp- content/uploads/2021/09/Raport-Barometr-Zawodow-2021.pdf BDL, https://bdl.stat.gov.pl/bdl/start, access: 1.12.2021. Hetmanski, M., D. Kiewra, D. Iwanowski, and P. Czyżak, 2021. “Just transition in Eastern Wielkopolska – Diagnosis and Recommendations,” Mimeo, INSTRAT. Hołub-Iwan J., I. Orsa-Chomiak, T. Michał, K. Gutta, T. Gutta, P. Gozdek, 2019. „Stan i perspektywy rozwoju inteligentnych specjalizacji w subregionie konińskim,” EU- Consult sp. z o.o., http://iw.org.pl/stan-i-perspektywy-rozwoju-inteligentnych- specjalizacji-w-subregionie-koninskim-raport/ Honorati, Maddalena, 2022. Options to support workers through a transition away from coal in Eastern Wielkopolska. World Bank Group, Jobs Group. Mimeographed. ILO (2004). R195 - Human Resources Development Recommendation (no. 195). https://www.ilo.org/dyn/normlex/en/f?p=NORMLEXPUB:12100:0::NO::P12100_INST RUMENT_ID:312533 Ingram, T., K. Bartuś, M. Baron, and Ł. Bielecki, 2020. “Ekspertyza Dotycząca Konsekwencji Likwidacji Kopalń Węgla Kamiennego dla Sektora Okołogórniczego oraz Sytuacji Społeczno-Gospodarczej w Polsce,” Centrum Badań i Rozwoju, Uniwersytet Ekonomiczny w Katowicach, Katowice. IRENA, 2020. “Global Renewables Outlook: Energy Transformation 2050 ,” International Renewable Energy Agency, 212. IRENA and ILO, 2021. “Renewable Energy and Jobs – Annual Review 2021,” International Renewable Energy Agency, International Labor Organization, Abu Dhabi, Geneva. Kiewra D., A. Szpor, and J. Witajewski-Baltvilks, 2019. “Sprawiedliwa Transformacja Węglowa w Regionie Śląskim. Implikacje dla Rynku Pracy,” IBS Research Report 02/2019. Lancaster, K. J., 1966. “A new approach to consumer theory,” Journal of political economy, 74(2), 132-157. 79 Lobao, L., M. Partridge, O. Hean, P. Kelly, S-H. Chung, E.N. Ruppert Bulmer, 2021. “Socioeconomic Transition in the Appalachia Coal Region: Some Factors of Success,” Washington, D.C.: World Bank Group. http://documents.worldbank.org/curated/en/531201635134585522/Socioeconomic -Transition-in-the-Appalachia-Coal-Region-Some-Factors-of-Success Martinez-Fernandez,C., N. Kubo, A. Noya and T. Weyman, 2012. “Demographic Change and Local Development: Shrinkage, Regeneration and Social Dynamics,” OECD, Paris. Mayer, Adam. 2018. “A Just Transition for Coal Miners? Community Identity and Support from Local Policy Actors.” Environmental Innovation and Societal Transitions 28 (September): 1–13. McFadden, D.,1974. “Conditional Logit Analysis of Qualitative Choice Behavior,” in P. Zarembka (editor), Frontiers in Econometrics, New York: Academic Press. McFadden, D., 2001, “Economic choices,” American economic review, 91(3), 351-378. McFadden, D., K. Train, 2000. “Mixed MNL models for discrete response,” Journal of applied Econometrics, 15(5), 447-470. Mihaylov, E., K. G. Tijdens, 2019. “Measuring the Routine and Non-Routine Task Content of 427 Four-Digit ISCO-08 Occupations,” Tinbergen Institute Discussion Paper 2019- 035/V, http://dx.doi.org/10.2139/ssrn.3389681 OECD, 2018. “A review of “Transition Management” strategies: Lessons for advancing the green low-carbon transition.” OECD, 2021. “OECD Skills Outlook 2021: Learning for Life,” OECD Publishing, Paris. https://doi.org/10.1787/0ae365b4-en. Pacifico D., 2014. “Sreweight: A Stata Command to Reweight Survey Data to External Totals,” The Stata Journal. 14(1):4-21. Piggot, G., M. Boyland, A. Down, and A. Raluca Torre, 2019. “Realizing a Just and Equitable Transition away from Fossil Fuels,” Stockholm Environment Institute, Stockholm. Poznań Regional Labor Office, 2018. “Ocena sytuacji na wielkopolskim rynku pracy i realizacji zadań w zakresie polityki rynku pracy w 2017 roku,” Poznań. https://bip.umww.pl/artykuly/2823922/pliki/20180404121452_1010z1.pdf. Poznań Regional Labor Office, 2019. “Zapotrzebowanie wielkopolskiego rynku pracy na kwalifikacje i kompetencje zawodowe w 2018 r.,” Samorząd Województwa Wielkopolskiego, Poznań, https://wuppoznan.praca.gov.pl/documents/161065/2074094/Zapotrzebowanie%2 0wielkopolskiego%20rynku%20pracy%20na%20kwalifikacje%20i%20kompetencje%2 0w%202018%20r.%20Raport%20skr%C3%B3cony/b4ebed35-6fd5-4398-863b- ac57cbf5170e?t=1552474262895. 80 Polish Green Network, 2019. “Just Transition of Eastern Wielkopolska. Challenges from the Civil Society Perspective – Analysis and Recommendations,” Alliance of Associations Polish Green Network, Warsaw. Revelt, D., K. Train, 1998. “Mixed logit with repeated choices: households' choi ces of appliance efficiency level,” Review of economics and statistics, 80(4), 647-657. Robertson, D., 2006. “Hard as The Rock Itself: Place and Identity in the American Mining Town.” University Press of Colorado, Boulder, California, US. Ruppert Bulmer, E., K. Pela, A. Eberhard-Ruiz and J. Montoya, 2021. “Global Perspective on Coal Jobs and Managing Labor Transition out of Coal: Key Issues and Policy Responses,” Washington, DC: World Bank. © World Bank. https://openknowledge.worldbank.org/handle/10986/37118 License: CC BY 3.0 IGO Strietska-Ilina, O., K. McCoshan, A. McCoshan, G. Montt, M., Harsdorff, B. Tumurchudur-Klok, A. Podjanin, S. Broek, M. Hudepohi, T. Mahmud, 2019, Skills for a greener future: A global view. Geneva: International Labour Organisation. https://www.ilo.org/wcmsp5/groups/public/--- ed_emp/documents/publication/wcms_732214.pdf Szymkowiak, M., A. Młodak, Ł. Wawrowski, 2017. „Mapping poverty at the level of subregions in Poland using indirect estimation,” STATISTICS, 609. Thissen, M., G. Mandras, 2017. “The importance of network: Consistent Interregional Trade and Transport flows between European NUTS 2 regions,” JRC. IPTS Working Papers. Train, K., M. Weeks, 2005. “Discrete choice models in preference space and willingness -to- pay space,” In Applications of simulation methods in environmental and resource economics (pp. 1-16). Springer, Dordrecht. Valliant, R., J.A. Dever, 2018. “Survey Weights: A Step-by-Step Guide to Calculation,” College Station, TX: Stata Press. Wirth, P., B. Mali, and W. Fischer, 2012. “Post-Mining Regions in Central Europe. Problems, Potentials, Possibilities.” White, I.R., P. Royston, and A.M. Wood, 2011. “Multiple Imputation Using Chained Equations: Issues and Guidance for Practice,” Statist. Med., 30: 377-399. World Bank, 2018a. “Managing Coal Mine Closure: Achieving a Just Transition for All.” World Bank, Washington, DC. © World Bank. https://openknowledge.worldbank.org/handle/10986/31020 License: CC BY 3.0 IGO World Bank, 2018b. “Poland Energy Transition: The Path to Sustainability in the Electricity and Heating Sector,” The World Bank. World Economic Forum (WEF), 2018. “Towards a Reskilling Revolution - A Future of Jobs for All,” World Economic Forum, Geneva. 81 Annex 1: Methodology to estimate the indirect impact of mine closure Table 2: Previous estimates of indirect impact of mines closures Study Authors Year Methodology and assumptions Estimations Joint Alves 2018 I-O (Eurostat)73. The estimation of Poland: 48,746 (intra- Researc Dias et al. indirect employment in the coal regional). 87,760 (inter- h Centre sector relied on the use of input- regional) (JRC) output tables and multipliers Upper Silesia: 22,106 (intra- regional), 34,536 (inter- developed by the EU Joint regional) Research Center, originally, for Lower Silesia: 1,698 predicting the impacts of a change (intra-regional) 3,045 in the final demand of one sector (inter-regional) on other related sectors (Thissen Wielkopolska: 3,447 and Mandras, 2017). Indirect (intra-regional); 8,090 employment was estimated by (inter-regional) applying the same multipliers to note: Inter-regional the number of coal direct jobs. The figures include Intra- indices used, besides extending the regional effect and, supply-chain coverage to all therefore, differences are sectors that might be impacted by due to the inter- regional changes in coal mining and coal trade between NUTS 2 power plants activities, are regions. assessed at intra-regional level, and consider spill-over effects at inter-regional level. Institute Kiewra, 2019 Input-Output (Central Statistical 56,700 – suppliers for Szpor, Office, base year 2015) 13,000 – recipients (coal- Structur Witajews Indirect jobs are calculated as the fired power plants) al ki- share of value added transferred to Researc Baltvilks the mining industry in the total h (IBS) value added generated in a given section multiplied by the number of employees in this sector (according to data for 2017) Frankows 2020 Input-Output (Central Statistical 96,000 – 112,000 ki, Office; base year 2015) subcontractors 41,000- Mazurkie Indirect jobs are calculated as the 56,700 wicz share of value added transferred to intermediaries 55,300 the mining industry in the total (coal-fired power plants, value added generated in a given heating and coking plants) section multiplied by the number of employees in this sector 73 The share of added value associated with mining to the mining added value is the same as the share of mining- related jobs to the number of mining jobs. 82 (according to Eurostat data for 2018) + additions of employment in Section C and D, proportionally to the share of coal in a given sector Universi Ingram 2020 Survey (207 companies; sampling The total number of ty of et.al. frame not specified) indirect jobs: 110,000 – Economi 130,000 cs in Number of affected Katowic indirect jobs (until 2030): e Optimistic scenario: 26,667 Plausible scenario: 50,580 Pessimistic scenario: 75,876 Table 3: JRC (2018) Estimates Direct jobs impacted by coal Indirect Jobs PP retirement Direct Inter- Impact Intra- regional 2025 2030 regional (total) Poland 112,500 48,746 87,760 mines 99,500 2,077 5,276 power plants 13,000 Silesia 82,459 22,106 34,536 mines 79,548 315 1,558 power plants 2,911 Lower Silesia 1,926 1,698 3,045 mines 1,108 480 102 power plants 818 Wielkopolska 3,400 3,447 8,090 mines 2,079 748 321 power plants 1,306 Source and methodology: JRC (2018), using EURACOAL database. 83 Annex 2: Skills and preference survey questionnaires The World Bank Skills and Preference Survey consisted of two parts: a skills and employment questionnaire and a Discrete Choice Experiment (DCE). The questionnaire included detailed questions to determine the respondents’ employment status and socio - economic situation. Questions were grouped into thematic blocks. For ZE PAK employees, the blocks consisted of (i) general information, (ii) employment characteristics, (iii) job search, and (iv) self-assessment of competencies. For the municipalities, the blocks consisted of (i) general information, (ii) professional situation, and (iii) self-assessment of competencies. For municipalities the number of questions in the general information part differed depending on the identified professional situation of the respondents (employed, unemployed, inactive). ZE PAK survey questionnaire Good morning! You are about to start the survey on work and professional preferences, carried out at the request of the World Bank Group to support the transformation of the labor market related to the Just Transition process. All your statements will be kept in complete confidence and will only be used anonymously, along with other people’s responses. We are counting on your honest answers. MODULE G: GENERAL INFORMATION G1 Please enter the year of birth |__I__I__I__| G2 What is your gender? 1. Woman 2. Man G3 What is your marital status? 1. Married 2. Divorced or separated 3. Widowed 4. Single G4 Do you have children? 1. Yes 2. No G5 Please indicate your level of education 1. Primary 2. Vocational 3.Secondary(high school or technical college) 4. Post-secondary 5. BA/BSc/Engineering degree 6. Master’s degree 7. Tertiary education with at least a Ph.D. degree G6 How many people are in your …................................................................................. household, including you? G7 What range is your total monthly net 1. PLN 1201–1500 income (per month) for you and all 2. PLN 1501–1800 people in your household from all 3. PLN 1801–2100 84 sources? Please include both your 4. PLN 2101–2500 earnings and any other income such as 5. PLN 2501–3000 pensions, scholarships, alimony, 6. PLN 3001–4000 benefits, rental income, etc.? 7. PLN 4001–5000 8. PLN 5001–6000 9. PLN 6001–7000 10. PLN 7001–8000 11. PLN 8001–9000 9. PLN 9001–10 000 10. above PLN 10 000 11. Refuse to answer -> go to G9 G8 How many people does this income a) ............................................................................ come from? b) Not applicable G9 How many members of your household, a) ............................................................................ apart from you, work in the ZE PAK b) Not applicable Capital Group? G10 How many of your household members, apart from you, work in a company that a) ............................................................................ is the main partner of the fuel and b) Not applicable energy company? G11 To what extent is your household 1. Dependent completely income dependent on the fuel and 2. Partially dependent energy sector? 3. Dependent slightly 4. Completely independent G12 To what extent do you feel that the process of just transformation may have a negative impact on your life? Mark on a rating scale from 1 to 5, where 1 means not at all and 5 very much. MODULE P: EMPLOYMENT CHARACTERISTICS P1 What is your total work experience in ….................................................................................. years (consider all your employers)? P2 Please indicate the number of all your past employers (i.e., places where the ….................................................................................. employment period was at least 3 months) Now, we would like to ask about your main workplace P3 What professional group do you Mining represent? Energy production Service/repairs Technical group, supervision, and management staff Administrative and economic employees P4 What is the name of your work ….................................................................................. position? P5 You job is: 1. Permanent, for an indefinite period 2. Fixed-term work 85 P6 Do you supervise or manage the work 1. Yes of other people? 2. No P7 In your opinion, what level of education 1. Tertiary education with Ph.D. degree is the most appropriate for the work 2. M.Sc. degree you do in your place of work? 3. BA/BSc degree 4. Engineering 5. Post-secondary 6. Secondary general education 7. Secondary vocational 8. Lower than secondary 9. The level of education does not matter P8 In your opinion, which field of study is 1. Field of study that I am studying/graduated the most appropriate for your current from main job? 2. Related to the field of study, I am studying/graduated from 3. Different field of study from the one I am studying/graduated from 4. The field of study does not matter 5. I do not know; I cannot define P9 Which of the following best describes 1. I need further training to do my job well. your skills in relation to the current job 2. My current skills are well suited to my in your place of work? responsibilities/duties. 3. I have the skills to handle more demanding duties. P10 Referring to your competencies related to the job you currently perform, please indicate which of the following two statements better describes the nature of your current job. It does not It requires require many any additional additional competencies competencies 1 2 3 4 5 MODULE Q: JOB SEARCH Now, we would like to ask about your future work. Q1 In the last 12 months, have you 1. Yes considered changing your job? 2. No Q2 Have you recently been looking for a 1. Yes job / participated in the recruitment for the new job? 2. No Q2a If yes, please indicate when Q3 Do you know institutions or 1. Yes, organizations that support job search? 2. No Q3a If yes, please provide examples Q4 How do you assess your chances of 1. High finding a job outside the fuel and 2. Moderate energy sector (mining sector)? 3. Low 86 4. None 5. I do not know Q5 Which of the following factors may be 1. Childcare an obstacle for you in the process of 2. Taking care of another family member looking for a job? (many answers are 3. Taking care of the house possible) 4. Taking care of a farm 5. Poor health 6. Age 7. Lack of appropriate certificates and authorizations 8. Level of education 9. Insufficient experience 10. Study or training 11. There are no job offers in the area 12. Lack of appropriate contacts, acquaintances Nie wymaga Wymaga wielu Q6 Please rate how prepared you are (on a scale from 1 to 5, where 1 means very żadnych bad, 2-bad, 3- innych, average, 4 -good, and 5-very well) to: dodatkowych dodatkowych kompetencji kompetencji a) search for and select job offers 1 Nie wymaga 2 3 4 Wymaga5 wielu żadnych      innych, dodatkowych dodatkowych b) prepare application documents kompetencji kompetencji Nie wymaga 1 2 3 4 Wymaga5 wielu żadnych innych,  dodatkowych     dodatkowych c) participate in interviews kompetencji kompetencji 1 2 3 4 5      Q7 Would you like to undertake training in the area of: a) Job search 1. Yes 2. No b) Soft skills (such as assertiveness, 1. Yes communication, etc.) 2. No c) Hard qualifications 1. Yes 2. No d) Other (what kind?) ….................................................................................. Q8 What kind of training do you find the 1. Job search most useful for you? 2. Soft skills (such as assertiveness, communication, etc.) 3. Hard qualifications 4. Other (what kind?) ……………….. Q8a If other is marked, please indicate which other training you find most useful for you. Q9 After the end of your current 1. Yes Nie wymaga Wymaga wielu żadnych innych, employment, do you intend to look for 2. No dodatkowych dodatkowych a job and take up a new job? kompetencji kompetencji 1 2 3 4 5 Q10 1. Only part-time  job     87 2. Depending on the offer, a part-time or full-time Which form of employment would you job be interested in if you were looking for 3. Only full-time job a new job? 4. Not applicable Q11 When looking for a job, would you be 1. Definitely not willing to learn a new job/reskill? 2. Rather not 3. It's hard to say 4. Rather yes 5. Definitely yes 6.Not applicable Q12 Would you be willing to take up 1. Yes employment in Turek? 2. It depends on the job offer 3. No 4.Not applicable Q13 Given that many jobs offer lower wages 1. PLN 1201–1500 than your current employment sector, 2. PLN 1501–1800 what net monthly salary (in hand) for a 3. PLN 1801–2100 full-time job would be satisfactory for 4. PLN 2101–2500 you / would you agree to accept? 5. PLN 2501–3000 6. PLN 3001–4000 7. PLN 4001–5000 8. PLN 5001–6000 9. PLN 6001–7000 10. PLN 7001–8000 11. PLN 8001–9000 12. PLN 9001–10 000 13. above PLN 10 000 14. Not applicable Q14 Have you ever considered the 1.Yes possibility of starting your own business? 2.No Q14a If yes, what kind of business did you consider establishing? …............................................................................. MODULE K: COMPETENCIES Different types of work require different skills and abilities. It is often the case that our capabilities are relatively high in one or two areas, while in others, they are much lower. Below is a list of the various skills. For each of them, I will ask you to assess the level of your own skills in this respect on a 5-point scale, where 1 is low, 2 is basic, 3 - medium, 4 - high, and 5 - very high. How do you rate your skill level? Competencies Very Low Basic Average High high self-organization of work and showing initiative 1 (Planning and timely implementation of activities 1 2 3 4 5 at work, effectiveness in achieving the goal) 88 1.1 independent decision making 1 2 3 4 5 1.2 entrepreneurship and showing initiative 1 2 3 4 5 creativity 1.3 1 2 3 4 5 (being innovative, coming up with new solutions) 1.4 resistance to stress 1 2 3 4 5 1.5 timely implementation of planned activities 1 2 3 4 5 contacts with other people, both with co-workers 2 1 2 3 4 5 as well as clients and subordinates. 2.1 cooperation in a group 1 2 3 4 5 2.2 easy contact with colleagues or clients 1 2 3 4 5 2.3 being communicative and communicating clearly 1 2 3 4 5 2.4 solving conflicts 1 2 3 4 5 3 organizing and conducting office work 1 2 3 4 5 managerial abilities and organization of work of 4 1 2 3 4 5 others 4.1 coordinating the work of other employees 1 2 3 4 5 disciplining other employees - bringing them to 4.2 1 2 3 4 5 order. 5 availability 1 2 3 4 5 5.1 willingness to travel frequently 1 2 3 4 5 5.1 willingness to work flexible hours 1 2 3 4 5 fluent use of the Polish language in speech and 6 writing (linguistic correctness, rich vocabulary, 1 2 3 4 5 ease of expression). searching and analyzing information and drawing 7 1 2 3 4 5 conclusions 7.1 quick summarizing a large amount of text 1 2 3 4 5 7.2 logical thinking, fact analysis 1 2 3 4 5 7.3 constantly learning new things 1 2 3 4 5 maintenance, assembly and repair of technical 8 1 2 3 4 5 devices 9 performing calculations 1 2 3 4 5 9.1 performing simple calculations 1 2 3 4 5 9.2 performing advanced mathematical calculations 1 2 3 4 5 10 computer operation and use of the Internet 1 2 3 4 5 10.1 basic knowledge of the MS Office type package 1 2 3 4 5 knowledge of specialized programs, the ability to 10.2 1 2 3 4 5 write programs or create websites 11 artistic and creative abilities 1 2 3 4 5 12 physical fitness 1 2 3 4 5 MODULE D: JOB CHARACTERISTICS Finally, we would like to ask about your job characteristics. 89 Please indicate which answer best Work inconsistent with your education and skills D1 describes the compliance of your work Work related to your education with your education. Work matches your skills and education Flexible working time Certified professional courses Please indicate which of the following Transport provided to and from work by bus non-wage benefits you receive in your Private medical care D2 current workplace. (Many answers are Childcare services (kindergartens close to the possible.) workplace, financed by the employer) Mobile phone or computer I am not getting any of the listed non-wage benefits Up to 15 min Please indicate which time of the 15 min – 30 min D3 following is the closest to your 30 min – 1 hour commuting time to work (one way). 1 hour – 1.5 hour 1.5 hour – 2 hours Municipalities survey questionnaire Good morning! You are about to start the survey on work and professional preferences, carried out at the request of the World Bank Group in order to support the transformation of the labor market related to the Just Transition process. All your statements will be kept in complete confidence and will only be used anonymously, along with other people’s responses. We are counting on your honest answers. MODULE G: GENERAL INFORMATION G1 Please enter the year of birth |__I__I__I__| G2 What is your gender? 1. Woman 2. Man G3 What is your marital status? 1. Married 2. Divorced or separated 3. Widowed 4. Single G4 Do you have children? 1. Yes 2. No G5 Please indicate your level of education 1. Primary 2. Vocational 3. Secondary (high school or technical college) 4. Post-secondary 5. BA/BSc/Engineering degree 6. Master’s degree 7. Tertiary education with at least a Ph.D. degree 90 G6 How many people are in your household, …................................................................................. including you? G7 How many household members, apart from you, work in the sector dependent ………………………………………………………………………………. on the fuel and energy sector? G8 To what extent is your household income 1. Dependent completely dependent on the fuel and energy sector? 2. Partially dependent 3. Dependent slightly 4. Completely independent MODULE N: PROFESSIONAL SITUATION N1 What is your total work experience in …............................................................................... years (consider all your employers)? N2 Please indicate the number of all your past employers (i.e., places where the ….................................................................................. employment period was at least 3 months) N3 Last week, did you do any work that 1.      Yes-> go to P1 brought you income or helped with a 2.      No business that your relative owns? N4 Last week, did you do any paid work for 1.      Yes-> go to P1 at least 1 hour? 2.      No N5 During the analyzed week, did you have a 1.      Yes-> go to N5a, and next to P1 job but did not perform it temporarily? 2.      No-> go to N6 N5a What was the reason you did not 1. Disease perform work in the last week? 2. Annual leave 3. Maternity leave 4. Parental leave 5. Business interruption 6. Difficult weather conditions 7. Working time system 8. Studying, improving qualifications 9. Unpaid leave 10. Other reasons, what…? N6 Are you looking for a job? 1.   Yes -> go to module Q (Unemployed) 2.   No -> go to module RR (Professionally inactive) MODULE P: EMPLOYED We would now like to ask about your main workplace. We consider the main place of work to be the place where you spend the most time doing your work. If you work the same number of hours in two or more places, consider the main one where you earn more. P1 How many hours have you worked in the 1.      I__I__I__I hours last seven calendar days in your main 2.      I do not know place of work? 3.      Refusal to answer P2 How many hours do you usually work per 1.      I__I__I__I hours week at your main place of work? 2.      I do not know 91 3.      Refusal to answer P3 What is your employment status at your 1.      I am self-employed ->go P4 current principal place of work? 2.      I am employed under a contract of employment- > go to P5 3.      I am an employee on the basis of a civil law contract (contract of mandate/contract for specific work)-> go to P6 4.      I am an employed person without a written contract of employment-> go to P7 5.      I am an employed person without a written contract of employment -> go to P6 6.      Other, what? -> go to P8 P4 In the last seven days, did you hire any 1.      Yes-> go to P6 employees? 2.      No-> go to P6 P5 Your main job is: 1.      permanent, for an indefinite period 2.      temporary P6 The institution/company that is your main 1.      public company/institution place of work is a 2.      private company/institution 3. NGO P7 How many people work in an 1.      From 1 to 10 people institution/company which is your main 2.      From 11 to 49 people place of work? 3.      From 50 to 100 people 4.      From 101 to 250 people 5.      From 251 or more 6.      I do not know 7.      Refusal to answer P8 What work did you do in the last seven calendar days? What is the name of your profession or occupation?   P9 What is the name of your work position?   P10 What do you do in your job? What are your main responsibilities?   P11 Do you supervise or manage the work of 1.      Yes other people? 2.      No P12 Indicate how much was your net earnings 1.      up to PLN 600 (in hand) earned in the previous month in 2.      PLN 601 – 800 your main place of work? 3.      PLN 801 – 1000 4.      PLN 1001 – 1200 5.      PLN 1201 – 1500 6.      PLN 1501 – 1800 7.      PLN 1801 – 2100 8.      PLN 2101 – 2500 9.      PLN 2501 – 3000 10.   PLN 3001 – 4000 92 11.   PLN 4001 – 5000 12.   PLN 5001 – 6000 13. PLN 6001 – 7000 14. PLN 7001 – 8000 15.   PLN 8001 – 9000 16. PLN 9001 – 10 000 17.   Over PLN 10 000  18. Refusal to answer P13 Agriculture, forestry, hunting and fishing Mining and quarrying Industrial processing Production and supply of electricity, gas, steam, hot water and air for air conditioning systems Water supply sewerage, waste management and remediation activities Construction  Wholesale and retail trade; repair of motor vehicles, including motorcycles What industries does your company Transport and warehouse management operate in? Please describe the industry in Activities related to accommodation and catering detail, indicating both general and specific services functions of the workplace. Information and communication Financial and insurance activities   Activities related to servicing the real estate market Professional, scientific and technical activities Business administration and support activities Public administration and national defense; compulsory social security Education Health care and social assistance Cultural, entertainment and recreational activities Other service activities Households with employees; households producing goods and providing services for their own needs Organizations and extraterritorial teams P14 In which month and year did you start 1.      I__I__I month working at your main workplace? 2. I do not know/ I do not remember 1.      I__I__I__I__I year 2.      I do not know/ I do not remember P15 In your opinion, what level of education 1. Tertiary education with Ph.D. degree is the most appropriate for the work you 2. M.Sc. degree do in your main place of work? 3. BA/BSc degree 4. Engineering 5. Post-secondary 93 6. Secondary general education 7. Secondary vocational 8. Lower than secondary 9. The level of education does not matter P16 In your opinion, which field of study is 1. Field of study that I am studying/graduated the most appropriate for your current from main job? 2. Related to the field of study, I am studying/graduated from 3. Different field of study from the one I am studying/graduated from 4. The field of study does not matter 5. I don’t know, I cannot define P17 Which of the following best describes 1. I need further training to do my job well. your skills in relation to the current job in 2. My current skills are well suited to my your place of work? responsibilities/duties. 3. I have the skills to handle more demanding duties. P18 Referring to your competencies related to the job you currently perform, please indicate which of the following two statements better describes the nature of your current job. It does not It requires require many any additional additional competences competences 1 2 3 4 5 P19 In the last 12 months, have you 1. Yes considered changing your job? 2. No P20 Have you recently been looking for a job 1. Yes / participated in the recruitment for the new job? 2. No To what extent do you agree with the statement: I feel burned out / burned out professionally? (1 - I strongly disagree, 2 - I tend to disagree, 3 - I have no opinion, 4 - I tend to P21 disagree, 5 - I strongly agree) I strongly disagree I strongly agree 1 2 3 4 5 MODULE U: UNEMPLOYED Q1 How long have you been looking for a job? 1.      |__I__I__| (Please enter the number of months) Q2 Are you currently registered as 1.      Yes unemployed at the Labor Office? 2.      No Q3 Are you looking for a job due to…. : 1.      Loss of job 2.      Resigning from work 3.      Willingness to return to work after a break 4.      Willingness to take up the first job in life 94 Q4 Do you know institutions or organizations 1. Yes, that support job search? Please provide examples …......................................... 2. No If yes, write what institutions or Q4a ………………………………………………………………………………… organizations you know Q5 When looking for a job, would you be 1. Definitely not willing to learn a new job/reskill? 2. Rather not 3. It's hard to say 4. Rather yes 5. Definitely yes Q6 How do you assess your chances of 1. Very good finding a job outside the fuel and energy 2. Good sector? 3. Moderate 4. Bad 5. Very bad 6. I do not know Q7 Which of the following factors may be an 1. Childcare obstacle for you in the process of looking 2. Taking care of another family member for a job? (Many answers are possible) 3. Taking care of the house 4. Taking care of a farm 5. Poor health 6. Age 7. Lack of appropriate certificates and authorizations 8. Level of education 9. Insufficient experience 10. Study or training 11. There are no job offers in the area 12. Lack of appropriate contacts, acquaintances Nie wymaga Wymaga wielu Q8 Please rate how prepared you are (on a scale from 1 to 5, where 1 means żadnych very bad and 5 very innych, well) to: dodatkowych dodatkowych kompetencji kompetencji a) Search for and select job offers 1 Nie wymaga 2 3 4 Wymaga 5 wielu żadnych      innych, dodatkowych dodatkowych b) Prepare application documents kompetencji kompetencji Nie wymaga 1 2 3 4 Wymaga 5 wielu żadnych innych,  dodatkowych     dodatkowych c) Participate in interviews kompetencji kompetencji 1 2 3 4 5      Q9 Would you like to undertake training in the area of: a) Job search 1. Yes 2. No b) Soft skills (such as assertiveness, 1. Yes communication, etc.) 2. No c) Hard qualifications 1. Yes 2. No 95 d) Other (what kind?) ….................................................................................. Q10 Training in which area do you find the 5. Job search most useful for you? 6. Soft skills (such as assertiveness, communication, etc.) 7. Hard qualifications 8. Other If other, please indicate what training you Q10a find most useful for you. If it was possible to finance any training Q11 from state funds, what training would you like to undertake? Q12 Which form of employment are you 5. Only part-time job interested in? 6. Depending on the offer, a part-time or full-time job 7. Only full-time job 1. Yes Would you consider taking up Q13 2. Depends on the job offer employment in Turek? 3. No Q14 What net monthly salary (in hand) for a 1. PLN 1201–1500 full-time job would be satisfactory for 2. PLN 1501–1800 you / would you agree to accept? 3. PLN 1801–2100 4. PLN 2101–2500 5. PLN 2501–3000 6. PLN 3001–4000 7. PLN 4001–5000 8. PLN 5001–6000 9. PLN 6001–7000 10. PLN 7001–8000 11. PLN 8001–9000 12. PLN 9001–10 000 13. above PLN 10 000 14. Not applicable Q15 Have you ever considered the possibility 1.Yes of starting your own business? 2.No -> go to Module K Q15a If yes, what kind of business did you consider establishing? …............................................................................. MODULE K: COMPETENCIES Different types of work require different skills and abilities. It is often the case that our capabilities are relatively high in one or two areas, while in others, they are much lower. Below is a list of the various skills. For each of them, I will ask you to assess the level of your own skills in this respect on a 5-point scale, where 1 is low, 2 is basic, 3 - medium, 4 - high, and 5 - very high. How do you rate your skill level? Competencies Very low basic average high high 96 self-organization of work and showing initiative 1 (Planning and timely implementation of activities at 1 2 3 4 5 work, effectiveness in achieving the goal) 1.1 independent decision making 1 2 3 4 5 1.2 entrepreneurship and showing initiative 1 2 3 4 5 creativity 1.3 1 2 3 4 5 (being innovative, coming up with new solutions) 1.4 resistance to stress 1 2 3 4 5 1.5 timely implementation of planned activities 1 2 3 4 5 contacts with other people, both with co-workers as 2 1 2 3 4 5 well as clients and subordinates. 2.1 cooperation in a group 1 2 3 4 5 2.2 easy contact with colleagues or clients 1 2 3 4 5 2.3 being communicative and communicating clearly 1 2 3 4 5 2.4 solving conflicts 1 2 3 4 5 3 organizing and conducting office work 1 2 3 4 5 managerial abilities and organization of work of 4 1 2 3 4 5 others 4.1 coordinating the work of other employees 1 2 3 4 5 disciplining other employees - bringing them to 4.2 1 2 3 4 5 order. 5 availability 1 2 3 4 5 5.1 willingness to travel frequently 1 2 3 4 5 5.1 willingness to work flexible hours 1 2 3 4 5 fluent use of the Polish language in speech and 6 writing (linguistic correctness, rich vocabulary, ease 1 2 3 4 5 of expression). searching and analyzing information and drawing 7 1 2 3 4 5 conclusions 7.1 quick summarizing a large amount of text 1 2 3 4 5 7.2 logical thinking, fact analysis 1 2 3 4 5 7.3 constantly learning new things 1 2 3 4 5 maintenance, assembly and repair of technical 8 1 2 3 4 5 devices 9 performing calculations 1 2 3 4 5 9.1 performing simple calculations 1 2 3 4 5 9.2 performing advanced mathematical calculations 1 2 3 4 5 10 computer operation and use of the Internet 1 2 3 4 5 10.1 basic knowledge of the MS Office type package 1 2 3 4 5 knowledge of specialized programs, the ability to 10.2 1 2 3 4 5 write programs or create websites 11 artistic and creative abilities 1 2 3 4 5 97 12 physical fitness 1 2 3 4 5 MODULE D: JOB CHARACTERISTICS Finally, we would like to ask about your job characteristics. Please indicate which answer best Work inconsistent with your education and skills D1 describes the compliance of your work Work related to your education with your education. Work matches your skills and education Flexible working time Certified professional courses Please indicate which of the following Transport provided to and from work by bus non-wage benefits you receive in your Private medical care D2 current workplace. (Many answers are Childcare services (kindergartens close to the possible.) workplace, financed by the employer) Mobile phone or computer I am not getting any of the listed non-wage benefits Up to 15 min Please indicate which time of the 15 min – 30 min D3 following is the closest to your 30 min – 1 hour commuting time to work (one way). 1 hour – 1.5 hour 1.5 hour – 2 hours 98 Annex 3: Skills and preference survey methodology DCE methodology To analyze preferences towards the workplace, a Discrete Choice Experiment (DCE) was applied. The concept of discrete choice methods is derived from the theory of economic value (Lancaster, 1966) and random utility theory (McFadden, 1974; McFadden, 2001). The utility refers to the level of satisfaction a person achieves from a particular consumption structure – here the consumed good is defined as the workplace. In the discrete choice method, respondents select an alternative that seems to be more attractive to them or decide not to make a choice. This selection process is of probabilistic nature (McFadden, 1974). The choice is determined by a utility that depends on the observed characteristics (attributes), individual preferences regarding these attributes (men may choose different than women; mine workers may choose different than above ground personnel etc.) and unobserved idiosyncrasies. This process allows modeling worker's decision processes and extracting information about preferences (individual utility function parameters) based on pre-composed successive choices in a controlled environment. The design of choice situations (composition of attributes) is based on an effective Bayesian method that allows obtaining the most accurate estimates with minimal respondents' choices (utility balance). The designing process is done using the simulation with obtained priors from the pilot study. Over the study, the design is optimized several times to minimize D-error (a scaled measure of the determinant of the Fisher Information Matrix, which summarizes how good a design is at extracting information about preferences). The choice of attributes and levels was based on literature review and preliminary qualitative work. Prior pilot-testing enabled researchers to refine the attribute levels and their wording. To avoid cognitive overload resulting from too many attributes, two DCE designs with two overlapping attributes were prepared. To estimate the monetary equivalents of a change in job components a monetary attribute (monthly net salary) was included. The designs prepared for ZE PAK employees and municipality residents differed slightly. 99 Table 4: Job attributes – ZE PAK DCE design 1 Attribute Levels of attributes Net monthly salary • PLN 3000 • PLN 4000 • PLN 5000 • PLN 6000 • PLN 7000 • PLN 8000 Contract • Self-employed • Public sector, employment contract for an indefinite period • Private sector, employment contract for an indefinite period • Professional preparation, fixed-term employment contract • Private sector, mandate contract (contract regulated by the Civil Code, not the Labor Code) • Private sector, contract for an indefinite period after 6 months of work Change in financial conditions within • No change the next 2 years from starting work • 20% increase over initial salary • 40% increase over initial salary Non-wage benefits • No • Flexible working time • Certified professional courses • Transport provided to and from work by bus • Private medical care • Childcare services (kindergartens close to the workplace, financed by the employer) • Mobile phone or computer Compliance of work with education • Job incompatible with skills and education (requires complete reskilling) • Related workplace (requires entry training) • Job corresponds to your skills and education Travel time to work (one way) • 0 (remote work) • 15 min • 30 min • 1 hour • 1.5 hour • 2 hours 100 Table 5: Job attributes – ZE PAK DCE design 2 Attribute Levels of attributes Net monthly salary • PLN 3000 • PLN 4000 • PLN 5000 • PLN 6000 • PLN 7000 • PLN 8000 Contract • Self-employed • Public sector, employment contract for an indefinite period • Private sector, employment contract for an indefinite period • Professional preparation, fixed-term employment contract • Private sector, mandate contract (contract regulated by the Civil Code, not the Labor Code) • Private sector, contract for an indefinite period after 6 months of work Company size • Employing 10 people • Employing 100 people • Employing 250 people • Employing 500 people Industry sector • Mining • Construction • Industrial processing • Renewable energy (photovoltaics, wind farms) • Agriculture • Transport services, repairs, communication • Other services (gastronomy, tourism and recreation, etc.) • Administration (office work) Compliance of work with education • Job incompatible with skills and education (requires complete reskilling) • Related workplace (requires entry training) • Job corresponds to your skills and education Working conditions • Not burdensome • Burdensome (physically strenuous or dangerous) Job location • Poland (no need to change the place of residence) • Poland (need to change the place of residence) • Abroad (need to change the place of residence) 101 Table 6: Job attributes – municipalities DCE design 1 Attribute Levels of attributes Net monthly salary • PLN 3000 • PLN 4000 • PLN 5000 • PLN 6000 • PLN 7000 • PLN 8000 Contract • Labor code based (permanent) • Fixed-term employment • Temporary (contract of mandate) • Apprenticeship / internship Change in financial conditions within • No change the next 2 years from starting work • 20% increase over initial salary • 40% increase over initial salary Non-wage benefits • No • Flexible working time • Certified professional courses • Transport provided to and from work by bus • Private medical care • Childcare services (kindergartens close to the workplace, financed by the employer) • Mobile phone or computer Compliance of work with education • Job incompatible with skills and education (requires complete reskilling) • Related workplace (requires entry training) • Job corresponds to your skills and education Travel time to work (one way) • 0 (remote work) • 15 min • 30 min • 1 hour • 1.5 hour • 2 hours 102 Table 7: Job attributes – municipalities DCE design 2 Attribute Levels of attributes Net monthly salary • PLN 3000 • PLN 4000 • PLN 5000 • PLN 6000 • PLN 7000 • PLN 8000 Sector • Self-employed • Public • Private • Non-governmental organization Company size • Employing 10 people • Employing 100 people • Employing 250 people • Employing 500 people Industry sector • Mining • Construction • Industrial processing • Renewable energy (photovoltaics, wind farms) • Agriculture • Transport services, repairs, communication • Other services (gastronomy, tourism and recreation, etc.) • Administration (office work) Compliance of work with education • Job incompatible with skills and education (requires complete reskilling) • Related workplace (requires entry training) • Job corresponds to your skills and education Working conditions • Not burdensome • Burdensome (physically strenuous or dangerous) Job location • Poland (no need to change the place of residence) • Poland (need to change the place of residence) • Abroad (need to change the place of residence) Econometric model The choice is determined by a utility which depends on the observed characteristics (attributes) and unobserved idiosyncrasies and for individual, i , resulting from choosing an alternative, j , in the situation, t , can be expressed as: U ijt = Xijt β + eijt . (1) Where Xijt , corresponds to observed job attributes with the vector of parameters, β , and eijt , refers to the stochastic component, standing for the unobservable factors. Assuming that 103 the stochastic component ( eijt ) follows an independent and identical extreme value (type I) distribution, the probability of, i, respondent selecting an alternative, j, from set of ,J, alternatives takes the form of conditional logistic regression: exp ( Xijt β ) P ( j |J ) = , (2)  exp ( Xikt β ) J k =1 which, depending on the selection and levels of attributes, enables to derive the maximum likelihood estimator of the utility function. As the respondents want to maximize their utility while making a choice, this probability represents an optimization problem. The designated formula 3 is known as the multinomial logit model (MNL). Train (2003) proved that choice behavior logit models have several limitations. Namely, MNL express the preference variation that relates to observed characteristics of the decision- maker – this is called systematic taste variation as opposed to random taste variation. The second limitation is that the logit cannot handle situations where unobserved factors are internally correlated over time. One of the ways to relax these restrictions is to allow for preference heterogeneity and, possibly, correlations between the alternatives and choices across time. This can be done by including the individual-specific parameters, βi , which leads to more general models. (McFadden and Train, 2000). To account for these limitations in addition to the MNL the data will be analyzed with a random parameter logistic regression called the mixed logit model (RPL or MIXL). Mixed logit is expressed as the integrals of standard logit probabilities over a density of parameters, where is an individual and is an alternative. e  n xni Pni =   (  b, )d , (3)  j e n nj  x Here in the random parameter model, we will use the panel specification formulated by Revelt and Train (1998). In the baseline assumption all non-price coefficients will be assumed to follow a normal distribution but the coefficients related to income will be assumed to follow a log-normal distribution. In this case, assuming lognormal distribution for the income coefficient is plausible from a behavioral perspective and the model accountability. Assuming log-normal distribution restricts respondents to have positive income sensitivity, in addition, this 104 assumption guarantees that the resulting distributions of WTP are useful and meaningful i.e., have finite moments (Daly et al., 2012). Given above it is convenient to introduce modification of the basic utility function defined above which is to use the so-called estimation in WTP space (Train and Weeks, 2005): Uijt =  ( pijt + Yijt b  ) + eijt =  ( pijt + Yijt β ) + eijt . (4) where is the stochastic component, is the time attribute, and the vector of parameters is a vector of implicit prices for the non-monetary attributes .In the survey, respondents were asked to choose between two hypothetical alternative job offers described by several attributes or decide not to make a choice. This selection process is of probabilistic nature (McFadden, 1974). The choice of an alternative that seems to be more attractive to respondents is determined by a utility that depends on individuals’ observed characteristi cs and unobserved idiosyncrasies. This process allows modeling workers’ decision processes and extracting information about preferences (individual utility function parameters) based on pre-composed successive choices in a controlled environment. 105 Annex 4: Preference survey – sampling frame The skills and preference survey was collected among two different strata, ZE PAK employees, and working-age adults of the four most affected municipalities. The results are presented independently for these two groups of individuals. Results for ZE PAK workers were reweighted to present representative and unbiased estimates of the current workforce employed by ZE PAK. Results from the four most affected municipalities were similarly reweighted to present representative and unbiased estimates of the working-age population (i.e., aged 15 to 64 years old), excluding employees of ZE PAK residing in these municipalities. Reweighting the sample from ZE PAK Steps to calculate weights for ZE PAK sample 1. Imputing missing information on gender, age and education using chained equations and additional information provided from HR database (salary, family status) 2. Estimating propensity score using logit regression and information on gender, education, and age 3. Calculating non-response adjusted survey weights as weight = 1/propensity score Weights were calculated to transform the self-selected sample of ZE PAK workers into a representative and unbiased sample. The original sample of ZE PAK workers is a self-selected sample of respondents: all ZE PAK employees received an invitation to participate in the survey, but around 1 in 4 participated (or 524 employees). Some did not respond to all questions, but the proportion of missing data was below 5 percent for all key variables. Missing data were imputed using chain equations method and all collected data, including salaries and family status (White et al., 2011). Thus, the final data are available for all 524 respondents and for these dataset non-response weights were calculated. Information about gender, education level, and age was provided by ZE PAK’s human resources department for all their employees. Based on this information, it was possible to assess non-response patterns in the sampled data. In general, the sample characteristics in terms of gender, age and education differed from the population of all ZE PAK’s employees as displayed in Table 8. Propensity score weights were calculated to adjust for non-response using a logit regression model, which, if correctly specified, can reduce the bias in the final weighted estimates (see Valliant et al., 2018). The HR data on ZE PAK employees, including indicators of gender, education level (5 categories), and age (9 categories), was pooled with the sample data (including the same indicators). The logit regression model was used to estimate the probability of being sampled based on individuals’ gender, education and age characteristics. Inverse probability weights were calculated using the probability of being sampled, estimated for each individual using the logit regression. Tables 8, 9 and 10 compare the distributions of education, gender, and age between the population, the unweighted sample, and the weighted sample. Overall, the propensity score 106 adjustments for non-response effectively corrects the sample distributions for the three key variables characterizing ZE PAK employees. The weighted sample provides similar proportions by gender, education level or age as in the full population. Table 8: Education distribution in the population, unweighted sample, and after weights adjustment for non-response Education Population Unweighted sample Weighted sample Primary or lower secondary 5.3% 1.3% 2.8% Basic vocational 31.1% 13.7% 29.6% Secondary 37.4% 27.7% 37.3% Post-secondary 0.8% 1.5% 0.6% Tertiary 25.3% 55.7% 29.6% Table 9: Gender distribution in the population, unweighted sample, and after weights adjustment for non-response Gender Population Unweighted sample Weighted sample Males 90.7% 84.7% 90.0% Females 9.3% 15.3% 10.0% Table 10: Age distribution in the population, unweighted sample, and after weights adjustment for non-response Age Population Unweighted sample Weighted sample 20/29 3.0% 3.6% 3.2% 30/34 7.1% 12.0% 7.9% 35/39 8.6% 13.4% 9.4% 40/44 8.7% 15.3% 9.7% 45/49 20.4% 22.1% 21.2% 50/54 24.6% 15.8% 24.0% 55/59 18.2% 13.7% 17.0% 60/64 8.4% 3.6% 7.1% 65/99 1.2% 0.4% 0.5% 107 Reweighting the municipalities’ sample Steps to calculate weights for the municipalities’ sample 1. Imputing missing information on gender and age using chained equations 2. Calculating base weights using the total population size divided by the sample size 3. Using raking algorithm, calibrating survey weights to known population totals by gender and age obtained from the central statistical office A random sample of 400 households from the four municipalities most affected by the ZE PAK closure was also interviewed (in proportion to the size of population in these municipalities). Within each household selected to participate in the survey (simple random selection), all working-age individuals were surveyed, i.e., people aged 16 to 64 years old. ZE PAK workers were excluded, as they were supposed to participate in the survey, through ZE PAK. The final results for the municipalities were reweighted, so that they represent an unbiased sample of the working-age population of the municipalities, excluding ZE PAK employees. To construct proper correction weights for the municipalities (from which ZE PAK employees were removed to avoid duplication), the first step was to add artificially ZE PAK employees back. The “full municipalities” sample is thus a combined sample of all ZE PAK employees and the sample of all households and working-age people in the four selected municipalities. Small number of observations with missing data on gender or age were imputed using chained equations and additional information on individuals. The base probability weights for the sample of municipality working-age people were calculated as the total population of people between 15 and 64 minus the number of ZE PAK employees from these municipalities divided by the collected sample size, which is (24169-1023)/409. The base weights for the ZE PAK sample were calculated as the number of ZE PAK employees in the four municipalities divided by the ZE PAK sample size, which is 1023/524. Thus, the sampled ZE PAK employees were treated as similar to all ZE PAK employees in the four municipalities, but their data are reweighted to reflect the real number of ZE PAK employees in the total population of working- age inhabitants of the four selected municipalities. Note that after weighting, the ZE PAK sample constitutes a relatively small part of the final population. The Central Statistical Office (GUS) population data for the four municipalities were used to adjust the sample for gender and age distribution. In this case, 10 age categories were constructed. The totals provided by GUS were used to calculate rake weights, which are the base weights adjusted to reflect population distribution of gender and age. We applied the algorithm developed by Deville and Sarndal (1992) to calibrate survey weights to known population totals as implemented in Stata (Pacifico, 2014). Tables 11 and 12 compare the distribution of education, gender, and age between the population, the unweighted sample, and the weighted sample. Overall, the weights adjustments effectively correct the sample distributions for gender and age. Weighted sample provides similar proportions by gender and age to the full population of working-age adults in the four municipalities. 108 Table 11: Gender distribution in the population, unweighted sample, and after weights adjustment for non-response Gender Population Unweighted sample Weighted sample Males 50.9% 68.2% 50.9% Females 49.1% 31.8% 49.1% Table 12: Age distribution in the population, unweighted sample, and after weights adjustment for non-response. Age Population Unweighted sample Weighted sample 15-19 7.7% 0.2% 7.7% 20-24 9.4% 3.3% 9.4% 25-29 11.4% 5.8% 11.4% 30-34 10.4% 15.9% 10.4% 35-39 11.2% 13.5% 11.2% 40-44 11.0% 17.6% 11.0% 45-49 10.7% 16.9% 10.7% 50-54 9.5% 11.9% 9.5% 55-59 9.2% 11.0% 9.2% 60-64 9.6% 3.9% 9.6% 109 Annex 5: The viable job matching tool - methodological considerations The extent of skills’ (mis)match, the availability of jobs, and workers’ preferences determine the viability of transiting to future jobs. Whether the path to an alternative job for a worker is viable depends on 1) whether the worker has the capacity to do the job, 3) whether the job attributes (salary, benefits, contractual arrangements, etc.) live up to the expectations of the worker (and compared to the value and compensation received when not working), and 3) whether the job is available locally or not. Most jobs are unlikely to score equally well on all three dimensions. The job may be available, for example, but only in another region, or only under a self-employed as opposed to a salaried contractual arrangement. The better a job scores on each of these dimensions (matching skills, attribute preferences, and local availability), the more viable is the transition. In particular, in identifying matches, the following three constraints were considered : (i) the amount of reskilling due to the transition between two occupations was minimized, (ii) only occupations in growing demand on the local labor market were considered, and finally (iii) potential candidates were further screened using the results of the preference survey described in the previous section. The technical feasibility of an alternative job in terms of skills was first assessed based on task-similarity between professions (as a proxy for skills) using artificial intelligence (AI) text mining algorithms. Job descriptions were extracted from the Polish Labor Office and Labor Offices Job Postings using Selenium web scraper and resulting in a list of 2,700 key occupations. Occupations were then compared one by one, using Latent Semantic Indexing (LSI), and a similarity score (between 0 and 1) was constructed for each pair of occupations. Latent semantic indexing (LSI) is an indexing and retrieval method that uses a mathematical technique called singular value decomposition (SVD) to identify patterns in the relationships between the terms and concepts contained in an unstructured collection of text. LSI is based on the principle that words that are used in the same contexts tend to have similar meanings. A key feature of LSI is its ability to extract the conceptual content of a body of texts by establishing associations between those terms that occur in similar contexts. Applying LSI model with 100 topics (number established using trial and errors approach), we were able to describe each job/occupation in our corpus as a mixture of topics. Using standard distance metrics, the similarity of all pairs of jobs/occupations was calculated – in other words similarity of vectors containing information about topics mixtures. We used cosine distance to create a 2.7k/2.7k (rows/columns) similarity scores matrix. The score could take values bounded between 0 and 1, going from least to most similar. To complement and refine the task-based pairing and ranking of jobs, the types of tasks required by each position were further divided into five groups: non-routine manual, non- routine interactive, routine manual, non-routine manual, routine cognitive. This classification builds on the measure of routine and non-routine task content developed by Mihaylov and Tijdens (2019), who provided a measure of the Routine and Non-Routine Task Content of 427 110 Four-Digit ISCO-08 Occupations. The indexes are created by dividing the number of tasks in a given category by the number of all tasks performed in a given occupation. The results are reported at the 4-digit level and were integrated in the dashboard, defining the top similar professions in each of five categories of occupations. Table 13: Identification of similar professions with the designed tool Second, job transitions were restricted to positions with excess demand on the local labor market. The occupational barometers developed by the powiat labor offices were used to identify jobs with excess demand, in order to restrict job transitions to occupations in growing demand, or with high potential. Finally, using the results of the preference survey it can further be determined whether jobs fall within the realm of the job attribute preferences expressed by the coal-related workers. For instance, ZE PAK workers display a strong aversion to moving regions. Job offers that are good matches, but imply moving to another part of the country, would thus be discarded if the wage increase was below the expressed welfare loss from taking up job in another location. The option was then discarded from the viable set. The proposed method of identifying possible transition paths and similar professions can be used as a tool supporting the activities of labor offices (e.g., in providing career consultations). Its ease of use allows employees to make a preliminary overview of possible career paths 111 themselves and allows employers to identify employees who may be placed in newly created positions, e.g., as part of creating new economic zones. 112 Most Recent Jobs Working Papers: 70. Towards a Just Coal Transition Labor Market Challenges and People’s Perspectives from Silesia. (2022) Luc Christiaensen, Céline Ferré, Tomasz Gajderowicz and Sylwia Wrona. 69. Towards a Just Coal Transition Labor Market Challenges and People’s Perspectives from Lower Silesia. (2022) Luc Christiaensen, Céline Ferré, Tomasz Gajderowicz and Sylwia Wrona. 68. A Tale of Two Countries: Labor Market Profiles of Youth in Urban and Rural Cameroon. (2022) Ioana Botea and Mitja Del Bono 67. Improving Smallholders’ Jobs Through Agribusiness Linkages : Findings of the Mozambique Agricultural Aggregator Pilot (MAAP). (2022) Michael Baxter,Christopher Brian Delgado, Jose Romero and David Ian Walker. 66. Rural Employment in Africa: Trends and Challenges. (2022) Luc Christiaensen and Miet Maertens 65. Understanding and Predicting Job Losses Due to Covid-19: Empirical Evidence from Middle-Income Countries. (2021) Maho Hatayama, Yiruo Li, and Theresa Osborne 64. Revisiting Labor Market Regulations in The Middle East and North Africa. (2021) Maho Hatayama. 63. Migrants, Markets and Mayors Rising above the Employment Challenge in Africa’s Secondary Cities – Key Insights. (2021) Christiaensen, Luc and Lozano Gracia, Nancy. 62. What was the Impact of Creating Better Jobs for more People in China’s Economic Transformation? What we know and Questions for Further Investigation. (2021) Merotto, Dino and Jiang, Hanchen 61. Opportunities for Youth and Women’s Participation in Ghana’s Labor-Intensive Public Works Program (2021) Dadzie, Christabel E. and Ofei-Aboagye, Esther. Click here for full Jobs Paper Series Address: 1776 G St, NW, Washington, DC 20006 Website: http://www.worldbank.org/en/topic/jobsanddevelopment Twitter: @WBG_Jobs Blog: https://blogs.worldbank.org/jobs/