Report No. 25456-TUN Republic of Tunisia Employment Strategy (In Two Volumes) Volume II: Annexes May 28, 2004 Middle East and North Africa Human Development Sector (MNSHD) Document of the World Bank PDRI Integrated rural development program (Programme de dkveloppement rural intkgrk) PDUI Integrated urban development program (Programme de dheloppement urbain intigrk) GDP Gross domestic product SME Small and medium enterprises PRD Regional development programs (Programme rkgional de dkveloppement) PRONAFOC National in-service training program (Programme national de la formation continue) SIVP 1 Youth training program for higher education graduate (Stages d`initiation 6 la vie professionnelle pour dipl6mb de 1`enseignement supdrieur) SIVP2 Youth training program for primary and secondary graduates (Stages d'initiation LEla vie professionnelle @our les jeunes ayant un niveau compris entre la 98n" annke de 1`enseignement de base et la 2"' de l'enseignement supdrieur suns succb) SMIG Industrial Minimum wage (Salaire minimum industriel garanti) TFP Vocational training tax (Taxe d la formation professionnelle) UNIDO United Nations industrial development organisation Vice President: Jean-Louis SarbiblChristiaan Poortman Country Director for Maghreb: Theodore Ahlers Sector Director: JacquesBaudouylMichal Rutkowski Task Team Leader: Setareh Razmara Acknowledgements This study was prepared with close collaboration of the Tunisian authorities and is based on findings of missions that visited Tunisia in February and November 2002 and June 2003. The report was prepared under the supervision of Regina Bendokat (Sector Manager, MNSHD). The team members were: Elizabeth Ruppert Bulmer (MNSED); Gordon Betcherman (HDNSP); Sara Johansson de Silva (MNSED); Aristomene Varoudakis (MNSED); Mohamed Salah Redjeb (Consultant); Lotfi Bouzaine (Consultant); and Ingrid Ivins (MNSED). Valuable inputs on private sector development were provided by Hamid Alavi (MNSIF), and preliminary inputs on vocational training reform were provided by Jean-Paul Peresson (Consultant). Peer reviewers were Zatiris Tzannatos (MDH), Amit Dar (HDNSP), Peter Fallon (IMF), and Martin Rama (DECRG). The task manager was Setareh Razmara (MNSHD). Valuable comments and suggestions were received from Mustapha Nabli, Pedro Alba, Tarik Youssef, Dipak Dasgupta, Farrukh Iqbal (MNSED); Jeffrey Waite, Dung-Kim Pham, Francis Steier (MNSI-ID). Mmes. Darcy Gallucio and Emma Etori were responsible for formatting the paper. Special thanks is extended to all Government offtcials, particularly Ministries of Employment, Development and International Cooperation, Finance, and Social Affairs and Solidarity for their support and active collaboration (including their related agencies). The guidance provided by Messer. M. Mongi El Ayeb (Chef de Cabinet), Ali Sanaa (Advisor), Habib Karaouli (Directeur general de 1'Agence tunisienne de l'emploi), Goubautini (ANETI), Ministry of Employment ; Abdelhamid Triki (Directeur general des Previsions), Kamel Ben Rejeb (Directeur general de la Cooperation multilaterale), Khelifa Ben Fkih (Directeur general de 1'Institut national de la statistique (INS)) , Moncef Yousbachi (Directeur general des Ressources humaines), Hedi Mamou (Directeur general de 1'Institut d'economie quantitative (IEQ)) , Habib Fourati (Chef du Departement des Statistiques demographiques et sociales (INS)), Ministry of Development and International Cooperation; Chedly Aissa, Secretaire general of Ministry of Finance ; Sayed Blel (Directeur general de la Securite sociale), Mohamed Salah Chatti (Directeur general du Travail), Abdehrazak Khelifi (Inspecteur general du Travail et de la Conciliation), Mohamed Ridha Kechrid (President-Directeur general de. CNSS), Ministry of Social Affairs and Solidarity is highly appreciated. In addition their technical team kindly provided assistance and information. Collaboration of the INS and its technicians has been particularly valuable and special thanks to their support and active collaboration. Finally, the team would like to thank the Labor Union (UGTT) for its suggestions. Table of Contents ANNEX 1 Impact of Investment Incentives on Relative Factor Prices ........................ 1 ANNEX 2 Productivity Decomposition and Wage Data ................................................. 3 ANNEX 3 Empirical Evidence on Impact of "Globalization" and "Technological Change" on the Labor Market ........................................................................ 5 ANNEX 4 Job Creation in Select OECD Countries ........................................................ 9 ANNEX 5 Forecasts of Labor Supply and Demand by Education Level.. .................. 13 ANNEX6 Active Labor Market Policy within the International Context.. ............... .21 ANNEX 7 Labor Market Regulation: International Experience ...........*.....*.*......*....a 25 ANNEX8 Assessment of the Social Security Data .....~........~...~.~~~....................~............ 31 ANNEX 9 Labor Market Monitoring .......~....~~....~.........~~.................~~~.................~...~.....35 ANNEX 10 Statistical Annex .~~..~~.....~.......~.~~~...........~...~~......~~.............~...~............~.~..........51 ANNEX 1 Impact of Investment Incentives on Relative Factor Prices' The user cost of capital was calculated for the different sectors of the Tunisian economy, on the basis of the deflator for investment goods (9); the price inflation of capital goods (Aq/q); the observed nominal lending interest rates (r); and an assumed rate of capital depreciation for each sector @)-depending on sector-specific technical obsolescence and the mix of equipment and structures. The user cost of capital (C) is derived using the formula: C=q(G-rdq/q). Over 1995- 2001, the estimated average annual change of the user cost of capital (AC/C) for the non- agricultural productive sector was 3.6 percent. Fiscal and financial incentives for investment affect the user cost of capital by reducing the cost of equipment and structures at purchase (q), or by reducing the financing costs of investment (r). Direct subsidies to investment ( "primes d `investissement "), and rebates of tariff duties and VAT rates provided by the Tunisian Investment Code affect the user cost of capital by reducing "q". Interest rate subsidies and personal and the deductibility of equity investments from (corporate and personal) taxable income affect the user cost of capital through the cost of financing (c`r"). The effect of these two categories of incentives on the user cost of capital was approximated by their estimated cost in 2001, expressed as a percentage of gross fixed investment in the domestic private non-financial corporate sector. In 2001, the estimated cost of direct investment and interest rate subsidies, and personal income tax breaks was TD 117.4 million. As the tariff duty and VAT rate rebates apply to both equipment and intermediate goods imports, which are of about equal size, only half of the estimated total cost of these incentives (TD 225.4 million) was included in the estimation. The estimated total cost of tax incentives that directly affect the user cost of capital was thus TD 230 million in 2001. This amounts to an estimated 8.8 percent of the gross fixed investment in the domestic private non-financial corporate sector (about TD 2,600 million). Financial and fiscal incentives in the investment code alone reduce the user cost of capital by an estimated 8.8 percent. These "capital-biased" incentives correspond to 34 percent of the total amount of financial and fiscal incentives provided by the investment code in 2001 (TD 668 million). Investment incentives that are targeted directly to employment-through payment of employer social security contributions-amount to only 2.7 percent of the total. The rest of fiscal and financial incentives boost corporate profits more directly, and are thus likely to have a neutral impact on the relative user cost of capital and labor. ' Source IEQ and WB staff estimates. ANNEX 2 Productivity Decomposition and Wage Data PRODUCTIVITY Overall labor productivity growth in an economy can be seen as the combined effect of productivitv growth within the different sectors on one hand, and improvements in productivity levels that occur because labor shifts from sectors with low productivity levels to sectors with higher nroductivitv levels on the other. Within sector effects are thus typically a result of technological, organization and managerial improvements which augment the productivity of labor in a particular sector. Across sector effects generally involve a reallocation from agriculture - where productivity tends to be lower than the economy-wide level - to non agricultural sectors. The decomposition of total productivity gains is given by: Ay = ciTj * AY,~+ cJj * Ae,i ,j=A,I ,S j=A,I,S Where A denotes change, ej is sector j's share in total employment, yj is productivity level in sector j, and a bar on the variable denotes sector average, and A, I, and S stand for agriculture, industry and services respectively (although the decomposition can be done at any level of sectoral detail). The first term is the within sector effect, given by holding employment shares constant and only looking at sectoral productivity gains. The second term is the across sector effect, where sector productivity gains are held constant. * Table 2.1 shows a decomposition of productivity growth in Tunisia for the period 1989-2001 and for the period 1997-2001 only. This breakdown confirms that productivity growth was largely due to productivity gains within sectors - in the period 1997-2001, productivity gains were in fact exclusively due to within sector productivity growth, while the effect of labor movements across sectors was negative. For example, slower employment growth in the services sector and faster employment growth in agriculture in the period 1997-2001, affected negatively overall productivity growth. Table 2.1. Annual Labor Productivity Growth within and between,Sectors, 1989-2000 and 1997-2000 Within Sectors Between Sectors Tota 1989-2000 Percentage Percentage Percentage Points Percent Points Percent Points Percent Amiculture 0.5 23 -0.2 -10 0.3 13 Minufacturing 0.5 22 0.1 5 0.7 28 Non-manufacturing 0.1 6 -0.1 -4 0.0 2 - L,;I I ---Construction 0.2 10 0.0 -I 0.2 9 Services 1.0 40 0.4 18 1.4 58 Total 2.2 91 0.2 9 2.4 100 Within Sect """' Tota Percentage Points Percent -0.2 -5 Manufacturing 0.8 0.9 27 Non-manufacturing 0.4 0.3 IO ---Construction 0.3 ~~1~ 0.3 IO Services 2.4 2.2 67 Total 3.4 3.3 100 Source: Staff calculations based on natiol nal accounts, employment surveys. ' For an overview of the productivity decomposition methodology and its application to Turkey, see Filiztekin (1999), "Convergence Across Turkish Provinces and Sectoral Dynamics", background paper for World Bank (2000), Turkey: Economic Reforms, Living Standards and Social Welfare Study. ANNEX2 WAGEDATA Several sources of wage data were explored for this study: national accounts, the [1997-19991 establishment surveys (that feed into nationalaccounts), the 1999 employment survey, and annual data from the social security office (CNSS). In addition, there i s data on wages in the industrial sector from UNDO'Sown establishment surveys. The trend in wage growth is fairly similar across the different sources of data (see figure 2.1). Nationalaccounts as well as CNSS data suggest that nominal wages have doubledover the 1990s, with an annual average increase of 7 percent. Inreal terms, this amounts to 20-25 percent (around 2 percent on an annual basis), depending on the sourceused. Figure2.1. Figure2.2. Average real wage trends, non-agricultural sector Level of average real wages, nonagricultural excluding public administration, 1997=100. sector excluding public administration,as share of average wage levelin nationalaccounts 1997 n -National accounts -CCNSS "ational accounts mCNSS Establishmentsurveys OEstablishmentsurveys OINS EMPL surveys 115 - I 200 110 105 150 95 100 50 0 1989 1994 1997 1999 2000 iurce: Staff calculations based on narional accounts, CSS Source: Staff calculation based on national accounts, Data and esrablishment surveys. CNSS data, employment survey 1999 and establishment survey (1997-99). There are more important differences in terms of levels, however, in particular between the average wage level, which emerges from the establishment survey sample, compared to other sources (see figure 2.2). It i s not surprisingthat national accounts data are lower than CNSS data since it includes the informal sector. However, it is not clear why the sample average wage i s so much higher. Clearly, the issue of wage data needs investigation. 4 ANNEX 3 Empirical Evidence on Impact of "Globalization" and "Technological Change" on the Labor Market3 GLOBALIZATIONANDLABORMARKET "Globalization" refers to the growing economic interdependence of countries primarily through increasing flows of goods and services, international capital, and technology. It may also include the movement of workers across borders, a potentially important aspect for Tunisia. Globalization is typically documented on the basis of increasing flows of international trade and foreign direct investment. For example, between 1990 and 1999, exports of goods and services as a share of worldwide GDP increased from 19 percent to 27 percent. The increase in gross private capital flows over this period was from 8 percent to 18 percent. Trade and investment flows remain largely between developed countries. However, some middle-income countries such as Tunisia are becoming more integrated into the global economy. Empirical evidence leads to the conclusion that globalization provides opportunities for economic growth, employment, and rising incomes. Research found that developing countries that were "globalizers" - i.e., had opened up their economies to trade since 1980 - experienced annual per capita growth rates in the 1990s of 5 percent compared to -1 percent for the others (This classification was based on the extent to which countries increased trade relative to income over the period. The top third were placed in the "globalizer" group, with the remaining two-thirds in the "less globalized" group). On the other hand, globalization has raised a number of concerns and questions. What is the effect of economic openness on employment and wages? How does it alter the distribution of income? Who wins and who loses? Despite all of the popular attention globalization has received, rigorous analysis of these sorts of questions has been limited and not always conclusive. Employment. The clearest conclusion from the research is that economic openness brings about structural changes that increase the rates of job destruction and job creation. There is more volatility and turnover in labor markets with globalization. To the extent that liberalization affects formerly protected sectors, then significant job destruction can occur. On the other hand, by stimulating improvements in productivity and output, liberalization of trade and investment can create jobs as well. As markets open up, restructuring often first leads to the destruction of jobs before new ones are created with the result that, at least in the short run, unemployment rates often rise. Over time, however, unemployment tends to decline. On a cross-sectional basis, countries that are open (in terms of trade-to-GDP) have lower unemployment rates than those that are not. W,ages. Globalization is associated with higher wages. The World Bank study found that wage growth between 1980 and 2000 in countries classified as "globalizers" was twice as high as in less globalized developing countries. However, according to data from 70 countries, the wage effects of trade and foreign direct investment seem to be different. The short-term impact of trade liberalization on wages is negative but then gradually improves to become positive after three years, on average. On the other hand, FDI provides a significant initial boost to wages but this effect diminishes over time, disappearing after 5 years, on average. On the basis of these results, then, the wage effect of globalization will differ depending on the mix of trade and investment. Distribution of wages. Many studies conclude that increasing returns to education result from trade and investment liberalization. As a consequence, wage differentials between skilled and unskilled workers widen. Much of this analysis has been undertaken in high-income countries; 3See World Bank (2002) study on the economic implications of globalization; M. Rama 2001. ANNEX3 however, this has also been observed in a number of developing countries (World Bank 2002). The fact that these changes occurred while the supply of skilled workers was growing in most countries (developed and developing) suggests that the demand for skilled labor has been increasing even faster. Unless, of course, there has been a mismatch in the types of skills offered by workers and the skill requirements of employers - a possibility that may well apply to Tunisia. However, globalization may be only partly responsible for the shifting relative fortunes of skilled and unskilled workers. Most empirical studies (generally for the developed countries and especially the U.S.) conclude that technological change is a more important factor, with international trade accounting for about 1O-20 percent of the increasing wage inequality. TECHNOLOGICALCHANGEANDLABORMARKET Technological innovations, specifically information and communication technologies (ICT), have transformed how goods and services are produced, as well as leading to new products. ICT is at the heart of current thinking about the "knowledge economy" and is seen as the key to spurring output and productivity growth, especially in high- and middle-income countries. Important organizational and managerial changes have accompanied the purely technical innovation. This includes increased contracting out, growth of small enterprises and self-employment (in some countries), and more emphasis on "numerical" and "functional" flexibility in labor deployment (OECD 1996). As noted earlier, the diffusion of new technology is closely interrelated with globalization. In fact, it is not always easy for researchers to isolate the effects of one from the other. Technological change, with the related organizational innovations, is having - and will continue to have -- major impacts on labor markets around the world. This includes where employment is located, in what industries and occupations, the nature of the employment relationship, skill requirements, wages and their distribution, the organization of work, and human resource and industrial relations practices. Like globalization, ICT offers upside employment benefits but risks as well. And, like globalization, one of the major concerns is that the diffusion of technological change is very uneven with a "digital divide" separating rich and poor countries. This divide also exists within countries with some groups (generally already advantaged) enjoying much greater access and benefits than others. Employment. Technological change affects aggregate employment in two ways - (i) eliminating jobs by replacing people with machines and (ii) creating jobs through the introduction of new products and services and by making production more efficient (and thereby reducing prices and increasing consumer demand). The ultimate effect depends on the relative weight of these two forces. The general conclusion, drawn largely from developed countries, is that technological change is necessary for long-run employment growth (OECD 1996). In manufacturing, employment growth has been correlated with level of technology: jobs have increased in high- technology science-based industries, stagnated in medium-technology sectors, and declined in low-technology manufacturing. Less research has been done on the employment impacts of technology in services. However, research suggests that employment gains in services have been largest in countries that had the greatest investments in ICT. Composition ofemployment. Technological change clearly alters the composition of employment through the processes of job creation and destruction. It is centrally linked to the changing industrial distribution of employment from primary to secondary and, ultimately, service industries. Related to this, white-collar occupations have increased their share of total employment in both industrial and developing countries. The OECD (1996) demonstrates how industrialized countries are increasingly involved in high-technology industries while shedding lower-technology ones. However, as the IL0 (2001) points out, developing countries - with their lower cost structure - can accelerate up the value chain. There are examples of this in IT where 6 ANNEX 3 some countries (e.g., India and South Africa) with a strong local skill base and infrastructure have established comparative advantages because of lower costs, including wages, compared to developed countries. Skill requirements and wages. The adoption of ICT is associated with the employment of more skilled workers. This likely reflects a number of factors: the new technologies require highly skilled workers; firms upgrade their workers to complement the technology they adopt; and highly-skilled workers are more likely to be chosen by employers to use ICT. This notion of "skill-biased technological change" - i.e., that the technologies shift labor demand to higher- skilled workers - is widely accepted as a key factor in the increasing returns to education that have been observed in many developed and developing countries. Most studies conclude that technological change, more than globalization, is responsible for widening skill differentials in many countries. ANNEX 4 Job Creation in Selected OECD Countries The experiences of OECD countries, and in particular Ireland and Portugal, point to the potential benefits from international integration in terms of both growth and employment but also emphasize the importance of adapting the economic structure to a new and more competitive environment. Successful integration requires upgrading the skill level of the work force and moving towards more flexible labor market arrangements to facilitate job creation. However, the importance of enterprise creation and economic growth more generally for employment growth implies that policy efforts must go beyond labor markets and include both a strong and stable macroeconomic framework and a considerable degree of liberalization of domestic product markets. CROSS COUNTRY EVIDENCE FROM THE OECD OECD countries differ in labor market outcomes. OECD countries in the last two decades have seen very marked differences in employment growth and unemployment trends. One set of countries, including the United States, Mexico, Ireland and Portugal, have experienced falling unemployment rates, as many more jobs were created than necessary to keep up with labor force growth. In other countries, such as France, Italy, and Finland, employment growth has been low or negative and unemployment rates have, at best, stagnated around 10 percent (see table 4.1). Table 4.1. Job Growth* and Unemployment Rates, OECD Countries Labor Market Outcomes Vary Significantly between OECD Countries Job Jnemployment Job Jnemployment European Union Srowth Rate Growth Rate 1990- Other OECD 1990- 2000 2000 2000 2000 Ireland 4.6 4.4 Iceland 2.4 2.3 Luxembourg 3.4 2.4 Norway 1.5 3.5 Germany 3.2 8.1 Switzerland 0.1 2.7 Netherlands 2.4 3.3 Spain 1.6 14.1 New Zealand 2.0 6.1 Greece 1.3 11.3 Australia 1.5 6.3 Austria `1.2 3.8 United States 1.3 4.0 Portugal 1.1 4.2 Canada 1.1 6.9 France 0.7 10.1 Denmark 0.4 4.5 Czech Rep. 1.3 8.8 United Kingdom 0.4 5.6 Poland 1.1 16.4 Belgium 0.3 6.6 Hungary -1.1 6.4 Italy 0.2 10.6 Sweden -0.6 5.9 Mexico 5.6 2.2 Finland -1.0 9.9 Turkey 3.1 6.8 Korea 2.4 4.2 *Nonagricultural sector. Source: Staff estimates based on OECD LFS ANNEX 4 Factors explaining these differences: l High economic growth rates are a prerequisite. Overall, job creation is positively associated with output growth in OECD countries: when output has expanded at a sufficiently high rate, demand for labor has increased, even as labor productivity has improved. Countries successful in generating jobs have generally recorded substantial output growth; countries with low employment growth have generally experienced sluggish output growth (see figure 4.1). l Favorable labor market institutions. Favorable labor market institutions and conditions will lead companies to meet additional demand by increasing labor input rather than capital. In general, firms in European countries - where labor market restrictions are tighter than in non-European OECD countries - appear to have accumulated capital at a much faster rate than labor, leading to relatively slower employment growth. 4 l Product market liberalization. The importance of high growth rates for generating employment implies that policies cannot focus on labor market flexibility alone but also on policies that foster firm creation (particularly SME) and innovation (see box 1). Empirical evidence indicates that a lack of competition in product markets - due to barriers to entry through e.g. barriers to trade and foreign investment or more generally due to public quasi or full monopolies - tends to reduce demand for labor and thus slow down employment creation. Evidence from countries like Australia, the UK and the US, which were successtil in creating employment and lowering unemployment rates, suggests that success is as much due to more flexible product market regulations - less state control, lower barriers to entrepreneurship, and lower barriers to trade and investment - as to more flexible labor market regulations. Although labor market policies still appear to explain the major part of cross-country differences in employment rates, differences in product market regulations also make an important difference. Figure 4.1. 6.0 + MEX 5.0 - l IRE 4.0 1 + LUX 1.0 + %#3SWE3.0 4.0 5.0 6.0 7.0 8.0 -1.0 - +~FIN + HUN I -2.0 VALUE ADDED 4See,e.g., Garibaldi and Mauro (1999) for a recent overview. 10 ANNEX 4 Box 1. Cross Market and Interactive Effects of Product and Labor Market Policies: Evidence from OECD Countries Over the past two decades, OECD countries have initiated reforms in labor markets to encourage employment, and in product markets to address the negative effects of anti-competitive regulations on productivity and consumer welfare. Recent empirical analysis by the OECD has looked at the cross effects of these policy changes, i.e. how regulatory reform may influence employment, employment security and wage inequality, and how labor market policies and institutions on the other hand may influence innovation performance and industry structure. The main conclusions from these studies are that regulatory reform in product markets can raise employment rates, while labor market reforms can enhance innovative activity and hence output growth. The degree of product markets liberalization is measured along three dimension. The level of state control relates to the size and scope of public enterprise sector, the level of control of public entities by legislative bodies, and special voting rights; as well as the degree of price controls. Barriers to entretweneurship refer to e.g. the complexity of licenses and permits systems, burdens for start-ups, and legal barriers to competition. Finally, product markets are affected by ex&cit barriers to trade and foreign investment, including ownership barriers, tariffs, and discriminatory provisions. Ill-designed entry restrictions tend to lead to inefficiencies in production as they reduce equilibrium output, shelter inefficient firms from competition, limit enterprise creation, and reduce the potential for technology spill-overs. These effects operate through many channels. Lack of competition in certain markets has increased rents, which have been captured in wage premia by insiders. In turn, this has tended to depress labor demand in favor of increased demand for capital. Output levels in monopolistic markets are below those of competitive markets, which reduces demand for overall production inputs, including labor. These effects combine to reduce and distort overall labor demand, put upward pressures on wages in certain industries, and reduce the rates of enterprise creation and survival. Importantly, although anticompetitive measures are often instituted to protect employment in a certain sector, the related productive inefficiencies are likely to spill over to the entire economy and reduce output and employment elsewhere. In addition, the budgetary costs of maintaining protection is likely to prompt a higher-tax burden, which in turn will tend to depress economic growth. ;ource: OECD (1998), Nicoletti et al (1999), Nicoletti et al (2001), Nicoletti and Scarpetta (2001). IRELAND: ANEXAMPLE OF SUCCESSFULECONOMICANDJOB CREATION The experiences of Ireland may provide particularly interesting insights for Tunisia, from two perspectives. First and foremost, it is a country which has seen: (i) high growth in output and employment, (ii) rapid economic convergence with other EU countries, (iii) a drop in unemployment rates from 17 percent in 1987 to 4 percent in 2000, (iv) a complete reversal in net migration trends, (v) a transition from a traditional to a highly modern economic structure, and all of the above while (vi) improving macroeconomic stability. And second, Ireland in the 1970s and 1980s shared many characteristics with Tunisia, in spite of differences in absolute income levels: (i) Ireland belonged to the poorest group of EU countries in 1980 together with Greece, Portugal and Spain, (ii) the sectoral employment structure was weighted towards the agricultural sector and traditional industries such as textiles and leather, and accession to the EU market posed similar challenges that Tunisia is now facing in terms of increased competition in traditional sectors, and (iii) labor force growth far exceeded the EU average, thus putting higher pressure on the need for job creation than in most EU countries (see table 4.2). 11 ANNEX 4 Table 4.2. Selected Economic Indicators: Ireland 1970-2000 and Tunisia 1997-2000 Source: WB data, OECD LFS, OECD NA The Irish growth process was largely driven by FDI and export growth. The rapid inflow of FDI, especially in high technology sectors, was fundamental in driving economic growth and job creation. Apart from direct effects, FDI had important effects on indigenous industry through demand linkages and technology spill-over. Economic and structural policy was key in setting the growth path. Several exogenous factors help explain the country's increasing attractiveness as a platform for exports of high technology goods and services, including (i) the booming IT sector in the United States, (ii) the presence of an English speaking labor force, (iii) tax breaks for foreign investors, and (iv) the benefits from structural funds from the EU. However, the growth process was probably more the result of a policy package focused on integration, education, and regulatory reform (see box 2). Box 2. Policy Package that Led to Growth and Job Creation-The Case of Ireland Overall policv stabilitv. Most importantly, Ireland has pledged to providing a stable investment climate with transparent and unchanging rules and incentives, including for foreign actors. Economic ouenina and Europehisation. Policy stability has included an unwavering commitment to the EU integration process, which has reduced uncertainty for foreign and domestic investors, and raised Ireland's attractiveness relative to the United Kingdom especially for American firms seeking a platform in Europe. Thus, Ireland joined the EU in 1973 and the European Monetary System in 1978, became part of the Single European Market in 1992 and joined the EMU in 1999. The associated reduction in tariffs and removal of exchange rate risks increased the attractiveness of the country as an exporting platform vis-a-vis Europe. Macroeconomic stabilitv. From the mid-1980s onward, Ireland moved away from an very unstable macroeconomic situation characterized by high indebtedness, fiscal difficulties, and double digit inflation rates. The Irish government imposed stringent monetary policy, cut fiscal deficits, and reduced debt, which greatly contributed to providing a stable business environment. Education and training offensive. From the 1970s onward Ireland has placed emphasis on increasing the effectiveness and efficiency of the education system, including through stronger links between the education sector and the private business sector. The result has been a highly educated work force which matches the demand for labor and which has been able to move from low productivity to higher productivity sectors. Competition and deregulation in product and labor markets. By the end of the 1990s Ireland was one of the less regulated OECD countries in terms of barriers to entry and entrepreneurship, market openness, and labor markets. Social consensus. A social pact was formed to restrain collective wage increases in order to regain competitiveness and to ensure that safety nets did not bear the brunt of fiscal retrenchment. The wage compact was compensated for with lower taxation on labor and business, and regulatory reforms were accompanied by active labor market interventions to assist in the adjustment of the unemployed. External support. As one of the poorer areas in Europe, Ireland benefited from transfers from EU's structural funds. These funds were well used and targeted to strengthening and modernizing infrastructure, including in telecommunications. Again, this upgrade was essential in attracting FDI. Source: SeeMcCarthy (2001). 12 ANNEX 5 Forecasts of Labor Supply and Demand by Education Level The objective of the model presented in this Annex is to provide an analytical framework with which to measure the gaps between employment and unemployment by category of qualifications, mostly acquired through education and skills acquisition. This statistical. model was created to establish links among the demographic data, the main parameters of the education and vocational training systems, and production. The forecasts established by this framework help identify potential imbalances and eventual gaps among the trends of these three systems: demographic, educational, and productive. The diagram below describes the structure of the statistical model. L and 2000. ) Labor supply by level of education and by skill. Outgoing pupils from the education system v and vocational training Gaps between supply Assumptions for by level of education and demand of labor by coefficients for and skill: (Projections education level and by employment from the Tenth skill. trends by level Development Plan and A of education. L cautious extrapolations from the study). Labor demand by level of education and by skill. Assumptions for the coefftcients Assumptions on - for employment f- retirees and the rate of trends by sector. looking for work by those leaving the education system and Assumptions for vocational training. Jobs in the economy by sector: (Projections from the sectoral the Tenth Development structure of Plan and cautious production trend. extrapolations from the study). Assumptions for I A 4 the trend of the overall economic growth rate. ANNEX 5 The model in this study was used to estimate projections for the period from 2002-the beginning of the Tenth Development Plan-to 2016-the end of the Twelfth Development Plan. For the period of the Tenth Development Plan, it was necessary to reconstruct as much as possible the forecasts from the different ministries and to complete them to obtain the higher levels of, disaggregation, particularly for the levels of education and skills needed for employment, and to attain the results of the projections for this period based on the available forecasts. An interpolation was also undertaken for the Ninth Development Plan to have a comparable basis between periods. Due to the lack of functional patterns, a number of assumption had to be taken with which to calculate the projections. Most assumptions use data from the employment surveys from 1997, 1999, and 2000, and from the forecasts undertaken for the Tenth Development Plan for the period 2002 to 2006. Assumptions were also proposed in some cases to assure a coherence between the different forecasts (INS, MFPE, MDE, MES) at the disaggregated level used in the projections. Finally, difficult choices had to be made to complete the necessary assumptions for projections so far in the future. The projections are thus based on the following different data sources and assumptions: l Tenth Development Plan Forecasts.developed by the Ministry of Economic Development (MDE) based on the evolution of added value factor costs) by type of activity for the period 2002-2006. These forecasts are used to illustrate the evolution of the sectoral structures of production (nomenclature/list 20) for this period. Beyond this period, the evolution of the production structure by sector is supposed to follow the trends observed or forecasts throughout the period 1996-2006. In applying the uniform rate of growth of 5 percent to the entire economy, one obtains a projection of the value added by sector for the period up until 2016. The projected sectoral growth rates are differentiated by the combined effects of the growth rates of the entire economy and from the supposed trends of the sectoral structures of production. For the Tenth Plan period, the growth rates used for the projections are the same as those used by the MDE, given that the trends are only applied to the period beyond that of this Plan. l Data from the employment surveys for the years 1997, 1999, and 2000. Combined with the added value by sector (MDE), these data allowed the labor coefficients to be calculated (labor by value added) for these same years. By applying the labor productivity growth rate by sector, one obtains a projection of the labor productivity coefficient for the period up until 2016 and an interpolation for the other earlier years for which we do not have data. This assumption turned out to be more precise for the period beyond 2000 that was based on trend of the labor coefficient by sector, which is a result of the comparison between the production data and the results of the employment survey of 1997 and 2000 with interpolations from the year 1996. It is assumed that this trend will continue throughout the period 2002-2006 with a multiplier effect by coefficient (the same for all sectors) which allows one to find overall employment provision of the Tenth Plan.' Beyond this period, we have reduced (in a uniform fashion as opposed to all of the sectors) the coefficient (0.95 instead of 1.15). In any case, we imposed on the labor coefficients that the 1996 level cannot be surpassed. All these treatments of data translate into projections for the period 2000-2016 by a strong progression 5 The forecasts of the Tenth Development Plan by the MED report a total creation for the period of about 380,000 all levels of instruction mixed in whereas the needs according to this source would be about 400,000 new jobs. The projections of active labor force from the INS (Assumption/Hypothesis Hl, used in one part of the study) shows that this is going to increase throughout the course of the period by 368,000 people. Another type of assumption from the INS (HO) foresee an increase in active labor population reach for the same period a total of 43 1,000. The forecasts for MDE would correspond thus to the average of the two hypotheses of the INS. 14 ANNEX 5 of average productivity of labor in Mining (13.9 percent per year) and the IMMCCV (8.4 percent per year). The increase in work productivity is weak in Agriculture, Finance, and Electricity. The national average is 2.4 percent a year for the different projection periods. Table 5.1 provides the grdwth rates of labor productivity by period and by major sector that result from all assumptions related to this question. Table 5.1. Labor Productivity Growth Rates (Projection Results) l The data,from the employment suweys from 1997, 1999, and 2000 are also used to calculate the employment gap according to the level of education for each sector of activity. An extrapolation has been deduced for the other years of the period 1997-2016 with the rate going up for qualified manual laborers to the detriment of those with less qualification. The extrapolation is done by applying the average annual growth rate realized between 1997 and 2000 to the rate of use of each level of instruction observed for each sector for the year 2000 for each of them. A ceiling of the improvement was imposed in order to keep account of the already high level of transformations that has transpired between 1999 and 2000.6 Table 5.2 provides the rates of use of different levels of instruction in the economy for a few key years. One will note that the table by the Minister of Vocational training and Employment shows a faster trend particularly in the rate of using the manual labor at the secondary level. The explicit objective envisages a rate of enrollment of 15 percent already in 2006 and 20 percent in2011. Table 5.2. Average Share of Instruction Levels in Total Employment (Projection Results) l By crossing the projection shares of the levels of instruction in employment with employment itself, we have established a projection of the employed population and of employment creation by sector according to the level of instruction. A matrix of conversion of post- secondary level employment by skill/specialty and by sector (see table 5.3) is used to project the jobs created for each area of specialty (economics, law, letters and humanities, and science and technology).7 6 The results of these measures have been corrected by the proportion in such a way that the sum of the rates of level of instruction equal 100%. 7 The matrix is constructed for the needs of this part of the study on the basis of the assumptions fed in part by management's opinions and by pieces of information on management in the different sectors of the economy. 15 ANNEX 5 Table 5.3. Distribution of Labor by Sectoral Qualifications (Percent) Sector Economics & Science & Law Letters & Total Management Technology Humanities 1 Agriculture and Fishing 15 1 81 1 1 1 3 1 100 1 IAA 5 1 9 1 1 1 3 1 100 IMCCV I 9 1 81 1 5 1 5 1 100 1 IME I 12 81 4 3 100 ITHC 12 100 If-U II 19 817 4? 36 rnn r Mining I 12 I 811 41 3 I 100 I Hydrocarbon 10 78 7 5 100 Electricity 10 83 2 5 100 Water 15 78 5 2 100 RTP 12 83 2 3 100 Transportation and Communications 12 83 2 3 100 Hotels, Caf&, and Restaurants 30 50 5 15 100 Finance 50 25 10 15 100 Business and Different Services 32 55 10 100 Administration, Education and Health 26 52 10 12 100 Since distribution of employment by specialty for some base years was not available, the calculations for this decomposition were limited to "flow." The elements of the matrix are chosen by assumption by issuing an opinion on the type of participation that the employees of different specialties in each sector could have. l National Institute of Statistics (INS) Projections of the active labor population until 2029. The figures in scenario Hl of the INS, which are supposed to be a continuation of the past trend of activity rates, were used to describe the trend and growth in the active population for each year. These are used in the projections like global needs, all levels of instruction mixed up, creating new types of jobs. Table 5.4 illustrates the INS projections using the assumption HI and the other assumption HO, both conceived by the INS. Table 5.4. INS Projections of Active Labor Force Trends* I Year I Assumotion HO I Assumntion Hl I I 20002001 I 3235.13322.7 I 3216.43285.6 I 2009 3990.1 2010 I 4059.8 3948.4 I 2013 4245.6 4149.4 2014 4297.8 4209.0 2015 4344.4 4264.9 2016 4385.8 4317.1 *The data for 1999 and 2000 and the projections for this part of the study used data from the World Bank and contain minor differences compared with other data in this table. Source: INS document. 16 ANNEX 5 l The Tenth Development Plan forecast for graduates with higher education level degrees, enrollment capacity in vocational training centers, and dropouts from higher education, secondary and primary levels. For those with college level and vocational training degrees, there is annual data for the period 1996 to 2001 and annual estimates for the period of the Tenth Plan (see table 5.5). For those leaving school without a diploma (primary, secondary, and higher) the documents of the MFPE show that the dropouts from secondary level school are about 230,000, those from higher education level are about 45,000 and primary are 70,000 for the entire period of the Tenth Plan. We assumed that the total number of pupils are divided equally across years of this period. Table 5.5. Higher Education and Vocational Training Degree Holders (number) Number of Pupils with Vocational Sources: * Ministry of Higher Education ** Ministry of Vocational training and Labor. l These estimates were extended beyond the period of the Tenth Development Plan, taking into account the volunte affirrnee and the measures taken recently to increase the internal efficiency of the education system.* For projections for graduates of graduate school, a lower growth rate was applied for the total number of students (2.5 percent and lower instead of fhe double digit rates foreseen for the period of the Tenth Development Plan.) Between 2000 and 2010, the number of pupils entering primary level education will go down from 200,000 to 150,000. From where those graduating from superior level school between 2012 and 2020 will be about 67,000 or more. l Those leaving the education system or the vocational training system are treated as if they were new entrants to the labor force. The other entrants, such as the returning emigrants or women who take up work again were thought to be negligible. The number who actually leave the active labor force, particularly for retirement, are calculated with the help of variable rates by level of education and by year to take into account overall trends in the structure of employment by level of education. The rates are ad hoc for post-secondary school and for secondary school levels. For the Terminale (final) years, the rates are chosen ' Throughout the period of the Ninth Development Plan, the exams to pass from one cycle to another from basic education to secondary education have been eliminated. For the Baccalaureat, taking into account a year's average has just been agreed upon and applied in 2002. The limits on the number of repeater in superior level schooling has just been relaxed. All of these measure are to increase the rate of internal efficiency in the education system and to reduce-or even eliminate-those who leave school, particularly at the primary and secondary levels. 17 ANNEX 5 so as to balance all the projections. For the primary level, the data on retirees is chosen in such a way so as to balance the new entrants and those leaving the active labor force calculated in this way with the INS projections illustrated above. In using the database from the employment surveys for the years 1999 and 2000 on the breakdown of the population by level of instruction/education, one obtains a projection of this distribution for the whole period up until 2016 and an interpolation for the other earlier years. It is assumed in these projections that the pupils who leave secondary school to join the active labor force go through the vocational training until it reaches its full capacity. Beyond this capacity, those who left secondary school add themselves directly into the active labor force. It is also assumed that a proportion, according to the ad hoc rate, of those who left the education system do not rejoin the active labor force. This analysis also includes the proportion of women who devoted themselves to the education of their children and to domestic work after having completed their education. This proportion was considered lower when the level of education attained is higher than the opportunity cost of the household work in terms of potential sacrificed revenue. l The Ministry of Higher Education's (MES's) projections of the graduates (dipomes) by skill for the period of the Tenth Development Plan are used for the period 2002 until 2007. Beyond that, it is assumed that the percentages by skill for the year 2007 remains constant (it is actually one of the MES's objectives). The relative shares of the different skills (specialties) are applied to the projections of the total number of pupils leaving higher level education to obtain the numbers for dropouts and graduates. The same shares are supposed valid for each of these two categories. The projections of the active labor force and its composition by level of instruction/education are compared to those of employment by education level and have allowed the rates of unemployment to be deduced for the entire economy and by level of education and to measure the gaps between the rate for the period from 2016, the Twelfth Development Plan. In terms of specialties or skills of those coming out of secondary level education, the rate of the new labor supply are thus established for the entire period of the projection. Based on preliminary projections for the period 2001-2016 (between the Ninth and Tenth plans), it is estimated that the educated labor force (post secondary) will increase by almost 2.5 times while the share of less educated labor will decline reduced by almost 30 percent. Based on these estimates, by 2016 almost the entire labor force will have at least a secondary diploma or a vocational training certificate (see table 5.6). Therefore in the medium- and long-terms, the labor force will continue to become more educated. 9 This change is mainly due to increasing enrollment and rising educational attainment as a result of the recent reforms in the education sector and increases in vocational training capacity, effectively eliminating all dropouts by 2016. Major changes are also projected to take place in terms of qualifications: most graduates will be in technical fields, while the share of graduates in literature and social sciences will decline (see table 5.7). 9The increased level df education of the working age population would be likely to boost participation in the labor force-otherwise, why would Tunisians keep accumulating human capital? But in that case, pressureson the labor force would increase, and in the absence of a stronger acceleration in job creation the unemployment rate could increase further-rather than smoothly decline, as shown in table 5.7. This issue needsto be further investigated. 18 ANNEX 5 Table 5.6. Projection of the Structure of the Labor Force by Level of Education in Medium and LongTerm Education level 1I N1397-2001 inth Plan Tenth Plan Eleventh Plan Twelfth Plan 2002-2006 2007-2011 201 I-2016 2016 Active population 3,136,924 3,505,ooo 3,876,220 4,205,120 4,317,100 Higher education 259,707 427,075 662,961 892,058 977,598 Secondary & vocational training (including secondary 964,590 1,148,525 1,355,491 1,547,346 1,614,294 drop-outs) Primary & others 1,912,628 Share of labor force 1OO,O% Higher education 8.3% Secondary & vocational training (including secondary 30.7% 32.8?kj 35.00/01 36.8%/ drop-outs) Primary & others 6l.Oo/al 55.0%/ 47.90/01 42.OYd 4O.OY Source: Staff projections Table 5.7. Projections of New First Job Seekers by Education Level in Medium and LongTerm Source: Staff projections During the next decade, due to lack of absorption capacity of the economy, increases in demand for skilled and educated labor will follow a slower path than supply. Based on preliminary projections on the number of graduates and sustained 5 percent annual GDP growth, unemployment among higher educated groups will triple in the medium-and long-term, while it will fall substantially for other groups, pointing to an increasing trend in unemployment for educated new entrants due to the persistent gaps between the skills demanded and the skills offered in the labor market (see table 5.8). Although the average unemployment rate could be reduced to one digit by 2016, unemployment among educated job seekers will substantially increase. Some sectors could be more dynamic (such as non-administration services) and could offer new opportunities for educated labor, but they are still subject to control, lack of competition and limited liberalization, and private sector development. Table 5.8. Projections of Unemployment Rate by Education Level in the Medium- and Long-Term Education level Ninth Plan Tenth Plan Eleventh Plan Twelfth Plan 1997-2001 2002-2006 2007-2011 a". .l-ml6 ml _-.v , 2016 I Post-secondary 9.6% 18.6% 28.3% 29.6%1 28.50/ Secondary & vocational training 16.8% 13.7% 10.0% 5.7% 3.2% Primary & others 16.7% 15.9% 9.3% 3.5% 1.1% All categories 16.1% 15.6% 12.8% 9.8% 8.1% 9m.rrn. 9+~Ct-~c+im.a+&- 19 ANNEX 6 Active Labor Market Policy within the International Context Active labor market programs (ALMPs) are important interventions for creating productive employment and managing labor market risks. ALMPs include a wide range of activities: public works, micro-credit and other forms of self-employment support; wage and employment subsidies; training and retraining; and employment services, including job placement, counseling, labor market information, etc. As globalization and rapid technological change increase the volatility in the labor market, accelerate structural change, and place a premium on the acquisition of productive skills, an effective active labor market policy assumes heightened importance. KEY POLICY ISSUES RELATED TO ALMPs ALMPs can increase the quality of labor supply, e.g., retraining; increase labor demand, e.g., direct job creation; or improve the matching of workers and jobs, e.g., job search assistance. Their objective is primarily economic - to help the unemployed find jobs, to increase the productivity and incomes of the underemployed, and to ensure that employers have a supply of appropriate workers. But ALMPs can also play an important social role by supporting productive employment for excluded or marginal workers. The international experience, however, underlines the challenges countries face in implementing an effective active labor market policy. The immediate question is whether these programs do any good. Evaluations of their impacts are mixed, with many programs assessed to have little or no impact on the employability or earnings of participants. For this reason, international organizations including the World Bank, the OECD, and the IL0 are placing great emphasis on monitoring and evaluation to identify what works, for whom, and at what cost. The overall strategy for active labor market programs involves identifying clear objectives; determining the composition of programs within the ALMP envelope; client priorities; and determining the relationship between active and passive policies (where these exist). Policymakers must confront a host of complex design and implementation issues - and incorporate the results of program evaluations - in order to maximize the probability for success. These include decisions about the complementarities of public and private roles, optimal resource allocation, targeting, and delivery. Effective and efficient active labor market programs, then, require considerable capacity. Most developing countries introducing or upgrading ALMPs must invest in knowledge, people, and technology in building this capacity. Active labor market programs can serve various objectives and policymakers need to be clear about which are the important ones. The orientation of an ALMP strategy can be to moderate cyclical downturns, reduce structural imbalances or otherwise improve the functioning of the labor market, increase productivity, support disadvantaged or at-risk workers, support at-risk employers or industries, or some combination of the above. Each of these objectives calls for different types of ALMPs and different client populations (Table 6.1). ANNEX 6 Table 6.1. Tailoring ALM Programs to Objectives Direct job creation (public works) Vulnerable groups (with least resiliency) Employment services (e.g., information, search Proximate regions, industries, or assistance, mobility assistance) occupations Training Wage subsidies Improve general labor Employment services All Training and retraining (including in-service, At risk or disadvantaged worker categories (especially for retraining) At-risk or disadvantaged worker categories iource: Betcheman ef al. (2001). ALMPs IN OECD COUNTRIES Most OECD countries have used active labor market programs for the past 40 years. During the past decade, ALMPs have assumed a higher profile in many of these countries as they have adjusted to the labor market realities of globalization and technological change. While the specifics vay depending on the national context, a number of general trends in the evolution of active labor marketpolicy can be identiJied, including: l An increasing reliance on private delivery of services - for example, for training, employment services, and public works. The Government's role in these situations has been to establish overall priorities, ensure quality, and provide financing, especially to address equity concerns. l Closer partnership with employers and communities in order to ensure that programs respond to market conditions; l More emphasis on evaluations of program impacts, including cost-efficiency. This also includes more "profiling" of clients to tailor programming to evidence on what works for whom. In some countries with unemployment benefit systems, program participation (based on profiling) is an obligatory condition to qualify for benefits. l Integrated services (one-stop window/guichet unique) so that clients can receive information, counseling, and access to services from a single source. Figure 6.1. Expenditures on ALMPs as Percent of GDP, OECD Countries, 1985-98 ANNEX 6 Average spending on ALMP has increased in the OECD. The OECD has collected statistics on expenditures by member countries on active labor market programs since the mid-1980s (see figure 6.1). The relative spending level increased in the early 1990s and continued at that new higher level through the decade. This rise likely reflects both an increasing relative preference on the part of governments for active programming and the higher unemployment in most countries in the 1990s compared to the 1980s. On average, ALMP spending by OECD countries is highest for direct job creation (31 percent of spending) and labor market training (25 percent). The remainder is spent on employment services (17 percent), disability (16 percent), and programs for youth (11 percent)." However, there are major differences in terms of how much individual countries spend on ALMPs and the distribution of their spending across the different program types. Table 6.2 presents recent expenditure figures for a sample of countries selected to offer variation by size, location, stage of development and employment policy orientation. On average Northern European countries (e.g., Belgium, Denmark, France, Germany) allocate about 1.5 percent of GDP to ALMPs, compared to 0.3 percent or less in the U.S. and Japan. Most of the European countries allocate significant resources to training for unemployed/at-risk adults but beyond that a number of different patterns are evident. Table 6.2. Expenditures on ALMPs: Selected OECD Countries (As Percent of GDP, Most Recent Years) Belgiu Denmark France Germany Hungary Japan Spain United States 2o"oo 2000 2000 2001 2001 2000-01 2001 2000-01 Public employment services and administration 0.17 0.12 0.18 0.23 0.11 0.20 0.09 0.04 Labor market training 0.24 0.85 0.25 0.34 0.07 0.03 0.14 0.04 Source: OECD (2002), Annex Table H. loThese are the five categories used by the OECD in collecting ALMP statistics. 23 ANNEX 7 Labor Market Regulation: International Experience11 International experience suggests that labor market regulation could play an important role in determining employment outcomes in a country. The challenge for any country is to find its best formula for labor market regulations that protect workers while encouraging growth of jobs and productivity. National conditions (culture, history, economy, etc.) can be very important in shaping this formula. Equity concerns and, specifically, the social protection of workers typically represent the primary motivation for labor market regulation. The main reasons for policymaker intervention in labor markets include redressing: (i) uneven market power between labor and capital; (ii) discrimination; (iii) insufficient information; and (iv) inadequate insurance against risk. Broadly speaking, policymakers can directly provide protection to workers in two ways.12 One is to use statutory regulation to establish the rules governing the employment relationship - for example, through a Labor Code. The other is to use state-sponsored employment programs to protect workers. These approaches can be illustrated using the case of protecting workers against the potentially adverse shocks of changing business conditions, technological change, etc. This is an important question for Tunisia. The first approach is based on the notion of protecting workers within the firm by establishing how labor contracts can be initiated and terminated. The second approach focuses on protecting workers outside the firm by providing income support through, for example, unemployment benefits or reemployment services through active labor market programs. Evidence suggests that different countries use different strategies in this regard (see figure 7.1). Figure 7.1. How Select Countries Provide Social Protection to Workers? INSIDEFIRM Low High High * Netherlands * Germany *France * Spain OUTSIDE FIRM * Canada * Tunisia *IX * Hungary * Japan * Korea * USA Low Source: World Bank Employment Primer, 200 1. " This sub-section is based on Betcherman, Luinstra, and Ogawa (2001) and a World Bank Employment Primer Series Note on labor market regulation (www.worldbank.org/iabormarkets/. ..). " There is a third, more indirect, way. This involves establishing a policy framework to support collective bargaining. ANNEX 7 Placing countries within this approach is (roughly) based on data about the strictness of employment protection rules ("inside the firm'? and the amount spent on active and passive labor market programs ("outside the firm `?.I3 In terms of the latter dimension, Tunisia is in the in the middle region; as we have seen in Chapter 4 of the main report, it is above average in spending on ALMPs (compared to OECD countries) but does not have an unemployment insurance program. Our focus in this section is on the indicator on employment protection rules which concerns how much scope employers have to hire and fire workers. This includes what kinds of contracts are permitted; the conditions under which workers can be terminated for "economic reasons"i4; requirements for severance and advance notice of termination; redundancy procedures; and special rules for mass layoffs. These employment protection arrangements are often characterized along a "rigidity/flexibility" or "protective/unregulated" continuum. Box 7.1 provides a stylized characterization of this continuum. Where countries fit on this continuum (which corresponds to the horizontal axis of Figure 7.1) tends to be based on a variety of factors including legal and cultural traditionsi Box 7.1. The Employment Protection Continuum RIGID OR FLEXIBLE OR PROTECTIVE UNREGULATED Fixed-term contracting restricted Unrestricted fixed-term contracting Temporary agency work restricted Unrestricted temporary agency work c. Hiring standards No hiring standards Employer dismissal rights restricted Unrestricted dismissal rights Substantial severance and advance notice required No severance or notice required Substantial administrative requirements for layoffs Simple administrative procedures Evidence suggests that employment protection makes dismissal costly to the employer. As Box 1 indicates, employment protection rules enhance job security for employees in two ways -- by restricting the ability of employers to hire workers on an explicitly non-permanent basis and/or by restricting employer freedom to terminate workers for economic reasons. Therefore, the way the labor market is regulated can affect job creation, productivity, and wages, as well as the degree of social protection afforded workers. Controversies surround employment protection regulations because of sharp differences in views about the costs and benefits of these policies. One perspective sees restrictions on non-permanent hiring and on employer dismissal rights as providing important social protection for workers. An opposing perspective emphasizes the fact that these regulations raise the cost of labor and, thus, discourage job creation and favor more privileged "insiders".16 In the end, good labor market policy - I3For OECD countries, measures of employment protection are presented in OECD (1999). For measures of active and passive programs, see OECD (2002 and earlier years) for calculations of program spending as a % of GDP. I4"Economic" reasons include business-related causesfor termination (e.g., shrinking markets, increasing competitiveness, etc.). This class of terminations stand in contrast to dismissals for "non-economic" reasons such asdiscrimination, union organizing, or job performance. I5 In countries with Anglo-Saxon heritage where common law principles prevail, statutory regulation typically plays a smaller role in directly providing employment protection than it does in countries with civil law principles (e.g., French or Spanish tradition) or in countries that have or recently had planned economies. Developing countries often follow the traditions of the countries that colonized them. l6 See Freeman 1993. 26 ANNEX 7 including employment protection, employment prpgramming, and wage-setting - must be shaped according to the economic and social outcomes.of different approaches. Contracts Table 7.1 compares a selection of other countries in terms of arrangements for fixed-term and temporary-agency work. The countries included have been selected to get regional diversity and to illustrate a range of different approaches. Countries have been ordered from most to least protective or rigid. Table 7.1. Legal Arrangements for Fixed-Term Contracts and Temporary Agency Work, Selected Countries Fixed-Term Contracts Temporary Agency Work Permitted only in the following cases: Licenses for private Specified piece work; employment agencies are Temporary replacement of absent worker; required from regional or Ethiopia Urgent work to prevent damage or disaster to national authorities, life or property; depending on scope of Work relating to the industry but performed at activities: irregular intervals; Seasonal work; Maximum duration one year after which the Chile contract becomes one of indeterminate duration. Fixed not allowed for the following: Graduates of university-level schools or No restriction specialist apprentice schools hired for work Czech that corresponds to their qualifications; Republic Adolescents; Employees under collective bargaining agreement; or Disabled persons Permitted for various reasons (e.g., specific Legal for justifiable cases Spain projects; temporary replacements; training contracts; production eventualities; special categories of workers; long-term unemployed) Japan < 1year duration without restriction Restricted to specific Up to 3 years for particular types of workers occupations Widely possible without justification General1y approved except for Germany Maximum number of 4 contracts/24 months construction (no limits in justified cases) h,>*,.P.nn+rLn-an T,,inr+*oonAnn.n.,..I,rml\ United States No restrictions No restrictions Termination Based on international evidence, in the context of economic liberalization and modernization, flexible dismissal procedure is the most important candidate issue for policy reform in the labor market regulation area. The rules and procedures discourage enterprise restructuring by imposing excessive costs and a large administrative burden on firms. There is also an element of administrative and judicial discretion that is unusual by the standards of many developed and emerging economies. International experience suggests that highly regulated dismissal procedures hinder job creation by making firms reluctant to hire workers when it would be costly to terminate them, if needed, in the future. This particularly dampens creation of permanent jobs, which are the most directly 27 ANNEX 7 affected. As a result, hiring is likely to be increasingly concentrated in fixed-term (or even unregistered) contracts that are outside these regulations. Table 7.2 provides a profile for a set of countries from most to least protective. Table 7.2. Legal Arrangements for Termination, Select Countries Justifiable reasons for Severance requirements Advance notice required economic dismissal 1 yr; :mployer; 113monthly salary for Plus? 1month if reduction of Ethiopia Uteration in work methods every additional year of workforce )r new technology to service (12 month limit) ncrease productivity. Additional 60 days pay if mass layoff. - - - Economic redundancy 20 days' wages for each 30 days notice year df service (up to 12 For mass layoffs, consultation Spain years) required for 15130days in firms with GOI> employees Employer shuts down or 2 months pay, in case of 3 months if employer shuts relocates; shut down, relocation, down or relocates and for Czech Employer ceasesto exist or transfer of employer, or redundancy; Republic merger or acquisition; redundancy. 2 months for other reasons. New technology or reorganization to increase efficiency Compelling business or No legal entitlement but Progressive increase based on operational needs often included in collective years of service (from 2 Germany agreements weeks to 7 months for >20 years of service) 1month delay required after public notice for mass layoffs Rational restructuring No legal entitlement but 30 days notice reason or unavoidable most large enterprises have Notification also to Public Japan redundancy (court voluntary plan Employment Security Office precedence, not law) in mass layoff (>30 workers) Reasonable selection criteria No restriction (except in No legal requirement but No regulation for individual United States public sector) voluntary or negotiated dismissal policies exist 60 days notice for mass layoffs Source: Betcherman, Luinstra, and Ogawa (2001). International evidence on the impact of employment protection rules shows that more flexible employment arrangements are likely to facilitate adjustment to macroeconomic shocks. According to the international evidence, table 15 indicates that the impacts of employment protection rules of restrictions on hiring and terminations on employment and unemployment levels are modest and, in the case of unemployment, often statistically insignificant. However, the 28 ANNEX 7 empirical findings are much stronger for the "dynamic" effects - on labor turnover and job tenure, job creation and destruction, and unemployment duration - and on the types of jobs created. There is a growing body of international evidence on the labor market effects of employment protection rules. Table 7.3 summarizes the research findings which are based mainly on the on the labor market impacts of job security regulations experience of developed countries. There has been less empirical analysis in developing and transition countries. This underlines the need for more research in these countries to see whether the impacts of job security rules are affected by large informal sectors, weak compliance, and other factors that distinguish labor markets in developing countries from those in developed economies. Table 7.3. Summary of Impacts of Employment Protection Regulations Impacts of Strict Limitations Regarding: 1 Fixed-Term and Temporary Agency 1 Terminating Permament Employees Work- - - - For Economic Reason; - 1 Employment ) Somewhat lower 1 Somewhat lower Labor force participation N/a Somewhat lower Unemployment Insignificant Insignificant Unemployment duration Longer Longer Non-standard employment N/a Probably higher Informal employment Higher Higher Job creation Lower Lower 1 Job destruction 1 Lower 1 Lower Labor turnover N/a Lower Job tenure N/a Longer Groups benefiting Prime-age males, skilled, Groups losing Women, youth Women, youth, unskilled Source: World Bank Employment Policy Primer Note on employment protection rules, 2001. 29 ANNEX 8 Assessment of the Social Security Data Data from the Cake Nationale de la Securite Sociale (CNSS) for the period 1994 to 2000 are assessed to consider the evolution of employment and wages in the formal sector, i.e., among those affiliated with the social security system.17 According to CNSS, employment growth in the formal sector was 12 percent between 1997 and 2000, outstripping the 8 percent employment growth observed in the economy as a whole based on national employment surveys (Table 8.1). Only about thirty percent of total employment is in the formal sector, however. Table 8.1. CNSS Summary Statistics 1994-2001" 1994 1995 1996 1997 1998 1999 2000 Total Formal Employment (CNSS) 150,363 809,979 859,959 888,716 909,794 963,107 998,964 Growth 7.9% 6.2% 3.3% 2.4% 5.9% 3.7% Share Of Total Employment 35.5 30.6 31.1 % % % Average Firm Size (Employees/Firm) 13.6 13.3 12.2 12.0 12.0 12.2 12.3 Average Annual Wage 1653,825 ,112,2 14 1,826,242 1,952,344 I 1!,075,280 !,185,535 2,243,833 Nominal Growth 7.2% 3.0% 6.9% 6.3% 5.3% 2.7% Real Growth (CPI) 0.9% 0.7% 3.2% 3.1% 2.6% 0.2% Inflation (CPI) 4.7% 6.3% 3.7% 3.7% 3.2% 2.7% 2.9% Average Wage By Firm Size ~10 Employees I,560,649 ,688,373 1,753,289 1,872,960 l,990,392 !,101,560 2,138,035 1O-49 Employees 2,018,589 !,136,833 2,100,512 2,291,898 !,464,706 !,564,691 2,761,54S 1 50-99 Employees !,3 17,606 !,346,535 2,392,558 2,519,712 !,676,911 !,736,393 2,950,282 100- 199 Employees !,485,3 15 !,514,797 2,857,185 2,835,362 2,933,58C 1,047,068 3,106,99( 200+ Employees 2,692,508 !,688,360 2,859,485 3,063,057 3,141,368 3,267,586 3,536,765 Total Number Of Firms 55,191 60,772 10,362 73,848 75,526 78,985 81,525 Firm Exits 5,883 6,54C 8,289 8,526 1,418 7,72 1 Firm Entry 11,464 16,13C 11,775 10,204 10,937 10,26: Net Firm Creation 5,581 9,59c 3,486 1,678 3,459 2,54~ As Share Of Total Firms 10.1 15.8 % % 5.0% 2.3% 4.6% 3.2% Avg. Net Job Creation (% 0 f Firm Size) 6.7% 4.3% 1.2% 1.8% 0.9% 0.3% Excludes public administration etwl oyment. Source: CNSS. Largerfivms tend to pay higher wages. According to CNSS data, between 1994 and 2000, average formal sector wages grew 5 percent annually in nominal terms, and slightly faster than inflation, implying real wage growth averaging 1.5 percent per year during this period.18 Overall, large size firms tend to pay higher wages (see table 8.1). In fact, firms with 200 or more employees pay wages that average about two-thirds higher than firms with fewer than 10 employees." i' Civil servants are not included in the CNSS data. `* Need to clarify if reported wages from CNSS are net of payroll tax and other benefits. I9 The wage figures do not account for hours worked and thus include part-time workers. ANNEX 8 Net job creation was negative in late 1990s. By tracking annual formal employment, the CNSS data provides insight into firm entry and exit as well as job creation and destruction. Because the employment figures are reported annually, the data reflect net (rather than total) job creation within each firm. The number of firms affiliated with CNSS grew steadily from 55,000 in 1994 to nearly 82,000 in 2000. Between 1997 and 2000, the rate at which firms exited the economy slowed, but the rate of firm entry slowed even more. Even though firms were being added to the economy on balance, net job creation at the firm level was negative; between 1997 and 2000, firms cut back employment levels by 0.4 percent per year on average (-1.8 percent in 1998, +0.9 percent in 1999, and -0.3 percent in 2000). These results suggest that the economy is characterized by considerable churning, in which firm creation and destruction occur simultaneously (even within the same sectors, as seen below), and at the firm level, jobs are added and destroyed. Table 8.2. Firm Size, Exit and Entry (Average 1994-2000) ~~ Source: CNSS. Smallerfirms are the most dynamic. The CNSS data, disaggregated by firm size and sector, provides some conclusions regarding the type of firms entering and exiting the Tunisian economy. For example, 93 percent of entering and exiting firms are small, with fewer than 10 employees (see table 8.2). Given that small firms account for 84 percent of total existing firms, the following can be concluded: (i) smaller firms are more flexible and incur fewer risks when launching a new venture (although we typically find that small firms have more limited access to capital); (ii) smaller firms may have lower shutdown costs compared to larger firms; and (iii) smaller firms are more vulnerable to shocks that force them out of business rather than simply requiring them to scale back operations. Looking more closely at new firms and their failure rates, the data indicate that two-thirds of the firms added from 1994 to 2000 were still operational in 2000. About 16 percent of fnms failed within the first year after launch, another 9 percent lasted only 1 year, and 5 percent survived only two years (see table 8.3). Table 8.3. Failure Rates `Numberof Firms Share New entry duration =4 years 528 1% New entry duration =3 years 1,612 2% New entry duration =2 years 3,100 5% New entry duration =l years 6,088 9% New entry duration ~1 year 10,517 16% New entry still in operation in 2000 44,922 67% Total new entrants 94-2000 66,761 100% Source: CNSS. Firm creation and destruction is mainly in construction and services. The sectoral distribution suggests that some sectors are more dynamic than others (see table 8.4). The most striking observation, however, is that these dynamic sectors have a higher share of both firm entry and exit, implying considerable churning within industries. Construction and 32 ANNEX 8 services account for the largest shares of firm creation and destruction, while the textile sector was stagnant between 1994 and 2000. *' Table 8.4. Sectoral Distribution of Firm Entry, Exit (1994-2000) ! i Entry ! Exit ' Agriculture, forestry, fishing 9% 8% Ind. agro-food 0% 1% Ind. Mat, constr., ceramic 13% 14% Ind. Mechanical, electric 9% 10% Ind. Chemical 6% 6% * ..-. _ -.....- I) -.-... _.. D> .--...-_ Other manf. mines& UL..LL.WY ~l+iti+i,=c I aI" I I I" BTP, commerce, t.-.._,~ort, telecoms M"SP 1 19% 18% Hotels. restaurants. banking. insurance. 1 I I repairs 18% 19% Other services 22% 19% Total 100% 100% Source: CNSS. " Add in table 1.10 exit/entry as a percent of the number of firms in each sector. 33 ANNEX 9 Labor Market Monitoring This Annex provides detailed information on the types of employment measuring devices currently available in Tunisia. In addition, current concerns and future strategies for improving employment data are discussed, as well as summaries of employment survey strategies in selected MNA and OECD countries. I. INFORMATION SYSTEMONLABORMARKETS The Institut national de la statistique (INS) under the Ministry of Cooperation and Economic Development is the single government entity responsible for collecting national surveys, such as the Population Census, Household Expenditure Survey, Employment Surveys (at the household level), and Enterprise Surveys. In addition, there is the Agence Tunisien de 1'Emploi (ATE), which is an employment services agency that matches job seekers to available positions. Although a useful source of employment data, ATE only collects information on those subscribed in ATE programs, or affiliates designed to match job seekers with job offers, therefore limiting the scope of information available. A third entity, the Observatoire National de 1'Emploi et des Qualifications, created in 1998, compiles and publishes employment data from other institutions, but does not conduct its own surveys. Both ATE and the Observatoire are under the management of the Ministry of Employment. 1.1. Supply Side: National Employment Survey INS collects partial employment and unemployment data as part of the population censuses conducted in 1966, 1975, 1984, and 1994. To complement this information, separate employment surveys, conducted at the household level by proxy, were carried out in 1980 and 1989. The objective of the employment surveys is to provide information on employment trends. In principle, the results of the surveys are expected to help policy makers formulate employment programs. However, given high unemployment rates, the Tunisian authorities realized that more frequent employment surveys would been needed to follow employment trends accurately. Therefore a national employment survey was launched in 1997, initially to be carried out on a biannual basis, and starting in 1999, on an annual basis. The sample framework of these new surveys is based on the census population of the 1994 census. Z.l.1. Common Characteristics of the Surveys Households are interviewed for approximately 45 minutes to one hour. The non-response rate is about l-2 percent, and interviewers cover 5-6 households per day. One visit is made to each household. The survey covers a period of 3 months (April - June) each year. The survey questionnaire begins with questions on the composition and demographic characteristics of all members of the household ages 3 and over. It also determines whether individuals are in or out of the labor force. The survey then contains a set of questions for unemployed and employed workers. In some cases, for example questions related to seasonal work, the questions are the same for both categories. Otherwise the set of questions focuses on the current place and type of work, for employed workers, and the length of unemployment, job search methods, and previous employment for the unemployed workers. The matrix below summarizes the questions contained in the 1997 survey and subsequent changes introduced in the 1999 and 2000 questionnaires. The main changes in recent surveys ANNEX 9 include more detailed disaggregations, as well as wage-related questions starting in 1999. In 2000, questions related to bonuses were added to the survey. l 1997 Employment Surwy: The sample was extracted in 2 stages, based on each gouvernorat (province). In the first stage, 1,000 districts were drawn from the breakdown of the 1994 census. In the second stage, a cluster of 25 households (on average) was drawn from each district sample, leading to a global sample of approximately 27,000 households. l 1999 Employment Survey: The 1999 Employment Survey covered a larger sample size than the 1997 Employment Survey - approximately 130,000 households. District samples were used, rather than clusters. New questions were added with respect to wage earnings, language proficiency, and types of professional degrees attained. l 2000 Employment Survey: In 2000, the Government decided to conduct the Employment Survey annually, with a large sample size to be used every 5 years (e.g., the 1999 Employment Survey), and smaller sample sizes in the intervening years. The 2000 survey used the same sampling methodology as that employed in 1997. The sample consisted of 25,000 households. New questions were added on (i) bonuses, such as the frequency of the bonus (quarterly, holidays, etc), the value of the bonus, and type of bonus (if received in-kind); (ii) details on how entrepreneurs set up new businesses (e.g., through government programs such as Fonds 21-21); and (iii) whether an unemployed worker would accept a job outside of his/her own province. l 2001 Employment Survey: The sample consisted of 45,000 households, utilizing a breakdown of districts and clusters. l 2002 Employment Survey: The 2002 Survey consisted of a sample of 130,000 households. This is the first Employment Survey with a "control" group, utilizing a sample of 46,000 households from the 2001 Employment Survey. However the results of this survey are not yet available. I.I.2. Data Missing or Not Well Defined Although the quality and level of detail of the Tunisian employment surveys is generally excellent, there are several issues that need to be clarified in order to adequately assess the employment trends. Term contracts. The introduction of term contracts is likely to be very important, but we cannot accurately measure the growth in use of term contracts, since the survey question is unclear regarding rhythm of work, which is lumped into two groups: permanent contract employees, and conjoncturels/saisonniers employees. Those holding term contracts may report themselves in either group. It would therefore be useful to have a separate question on contract length. Non-wage benefits. The returns to education and other characteristics are measured reasonably well by wages, but given that non-wage benefits are large in some sectors, the existing data do not allow an assessment of the role of non-wage benefits and the magnitude of their effect on labor supply decisions. Data on non-wage benefits, such as bonuses, was collected starting in 2000, but has not been made available yet. Informal vs. formal sector. The survey covers informal activity in all its questions, but the data does not permit separating informal employment from formal employment. Although we can make assumptions that the self-employed and employers are in the informal sector, or even with respect to the method of payment, this would be inaccurate 36 ANNEX 9 due to the considerable overlap between informal and formal working arrangements. Surveys in other countries address this issue by .asking questions on respondents' affiliation with the social security system, for example, or whether they receive particular benefits associated with formal employment. Also, additional information on the size of the firm in which they work may provide relevant information. Public enternrise sector. The survey question on "place of work" lumps together public firms with private sector firms, therefore it is not possible to separate the employment and wage trends between public and private enterprise sectors. Future surveys need to make a clear distinction between public and private sectors. Definition of unemployment. The more recent surveys have narrowed the definition of unemployment with respect to the reference period for "being unemployed" and "searching for a job". Starting in 2000, the survey included a question on job search within the past month, but the responses were not made available. It would be beneficial to add a question in which the reference period (for work and for job search) is the past week, in order to derive a definition of unemployment consistent with the IL0 and comparable to other countries. This does not require the Tunisian authorities to change their own definition, but may be useful in facilitating comparisons with other countries. Voluntarv unemnlovment. Surveys in other countries have tried to measure the extent to which unemployment is voluntary by adding questions on the lowest wage acceptable to individual job seekers. The Tunisian survey recently added a question on willingness to accept a job in another governorate, which provides some insight into the issue of voluntary unemployment, but is inconclusive. Link to poverty. Whereas employment surveys do not typically include questions on household income (level, sources, individual contributions, etc.), this information is necessary to establish the link between unemployment and poverty, which also provides an indication of the degree to which unemployment is involuntary or voluntary. There are 3 ways in which the survey questionnaires could be amended to provide information on the poverty/unemployment link: (i) the employment and household expenditure surveys could be linked using the individual identifier/household identifier to relate the responses in the two separate surveys to a particular household; (ii) a short employment module could be added to the income and expenditure survey; or (iii) income questions could be added to the employment survey (this last option is the most onerous). Other areas. Labor migration to Europe (or elsewhere).is an important phenomenon that is not accurately captured by the employment surveys in Tunisia. Adding questions to measure in a reliable way the magnitude and impact of temporary migration would be useful in informing the Tunisian authorities of the extent to which training resources are being spent on individuals who take their new skills to Europe and therefore do not contribute directly to Tunisian economic output. New questions could provide key information on (i) the age and education profiles of those migrating; (ii) how long they migrate for; (iii) incentives for migration (such as earnings in Europe); and (iv) how much is remitted back to Tunisia. This information could help rationalize the government's education and vocational training resources and curricula for future job seekers in the Tunisian labor market. 37 ANNEX 9 1.1.3. Definition of Variables For individuals age 15 and older who are active in the labor force, a detailed set of questions is asked to determine whether they are effectively employed or unemployed. The Tunisian definition of the working age population (15 years and older) is broader than the IL0 definition (15 - 64 yearsof age),), but the definition of unemployment (18-59 years) is more narrow than the IL0 definition (15-64 years). . For the purposes of this report, the IL0 definition of the labor force is used. The active population is broken down into the following categories: l Active effective: those who declare themselves as working or searching for a job. l Active marginal: those who declare themselves busy (housewives, students, retirees, etc) but performed some economic activity in the year preceding the survey. l Active potential: those who declare themselves inactive, did not perform any economic activity in the preceding year, and did not enter the labor force due to unavailability of jobs (i.e., discouraged workers). From this breakdown, a distinction is made between the employed and unemployed active population. l Active employed ("occupe"): those who worked at least one hour in the week preceding the survey. l Active unemployed ("non-occu&"): those who did not work at least 1 hour in the week preceding the survey, and declare themselves available and looking for employment. Of the category "active non-occupe", those between the ages 18 - 59 are considered as "chbmeurs", and those between the ages of 15 - 17 and 60 and older are considered as "others unemployed". The definitions have remained constant since 1994. 1.1.4. Dissemination Strategy A widespread dissemination strategy currently does not exist. While the INS is mandated to collect and produce statistics related to employment, annually and at the request of the sectoral ministries, the end-users (IEQ, l'observatoire, ATE, private sector, etc.) do not have access to the micro-data. However, with the implementation of the statistical reform in 1997, the government has begun to increase access to data. The law pertaining to the reform of the national statistical system and its implementing decrees (Law 99-32) was adopted in 1999. One key area of the reform includes special emphasis on the dissemination of data, along with a more thorough analysis of certain areas, including employment and income. In 2000, an advisory body was created, called the National Statistical Council (CNS), with one of its objectives being a more consultative process between producers and users of statistics. The CNS, in coordination with users (i.e. statistics departments in sectoral ministries) and producers (such as INS), has developed a National Statistical Plan, that will be implemented during the Tenth Economic Development Plan (2002 - 2006). A new law has also been extended to private organizations, allowing them to perform statistical operations not currently covered by government agencies, after informing the CNS. However this would necessitate their access to micro-data, which is not currently available publicly. 38 ANNEX 9 Another example of Tunisia's widening dissemination strategy is the recent adherence to the IMF's Special Data Dissemination Standard (SDDS) program.*l Tunisia is the first MENA- region country to qualify for the program, but several conditions remain to be met, including the need for higher-frequency employment statistics. The SDDS criteria normally require quarterly reporting, however as Tunisia meets most of the SDDS criteria in other statistical sectors, they are able to take a special option, called a "flexibility option", in order to qualify for the SDDS program, while efforts are made to meet the reporting criteria. 1..2. DEMAND SIDE The "Repertoire des Entreprises" is based on both fiscal files and CNSS files and is prepared by INS. Based on the Repertoire des Entreprises, 400,000 enterprises were identified, 100,000 of which are registered with the CNSS. In 2001, 12,000 enterprises out of the 400,000 were classified as having 6 or more wage earners. 1.2.1. Annual Enterprise Survey (Enqdte Nationale SUYI'Activitt! Economique) Prior to 1997, enterprise data was collected by INS, but not on a very systematic basis. However, since 1997, to estimate the National Accounts, the enterprise survey has been vastly improved and standardized. In this survey, the objective is to measure value-added for enterprises. Companies employing more than 6 wage earners were surveyed. In 1997, 11,000 enterprises out of 373,000 were identified as having more than 6 wage earners. In 2001, 12,000 enterprises were identified as having more than 6 employees. From these 12,000 enterprises surveyed, 3,000 responses were received and tabulated. These surveys are conducted annually; the 2000 survey was still being analyzed at the time of the February 2002 mission. Since the surveys are conducted to estimate production, the employment section of the questionnaire - which measures the number of employees by skill level and gender and the total size of the wage bill - is not adequately designed to assess the skills demanded by enterprises. 1.2.2. Micro-Enterprise Survey (Enqudte Nationale sur les ActivitPs Economiques) To capture employer trends in the informal sector, INS has begun to conduct micro-enterprise surveys since 1997. A sample of 11,000 enterprises out of the 363,000 identified as having less than 6 wage earners was sent questionnaires. 5,500 enterprises Responses were received from 5,500 businesses. This survey is expected to take place every 5 years, the next scheduled for September/October 2002. The purpose of the micro-enterprise survey is to capture some of the employment trends in the informal sector. Data covered includes activities breakdown by gender, average monthly salary, and by type of position. `I The Special Data Dissemination Standard (SDDS) was established by the International Monetary Fund to guide members that have, or that might seek, accessto international capital markets in the provision of their economic and financial data to the public. The SDDS is expected to enhance the availability of timely and comprehensive statistics and therefore contribute to the pursuit of sound macroeconomic policies, as well as contributing to the improved functioning of financial markets. Although subscription is voluntary, it carries a commitment by a subscribing member to observe the standard and to provide certain information to the IMF about its practices in disseminating economic and financial data. The SDDS, in taking a comprehensive view of the dissemination of economic and financial data, identifies four dimensions of data dissemination: (i) data coverage, periodicity, and timeliness; (ii) accessby the public; (iii) integrity of the disseminated data; and (iv) quality of the disseminated data. 39 ANNEX 9 I.2.3. Employment Agency Data (ATE) ATE is the national employment services agency. It makes available to job seekers a list of job openings, assistance in the job search process, vocational training, a database of other employment agencies, and information on the various government programs that provide financial and other assistance to set up micro-enterprises. Job seekers can conduct their research in person at a local ATE office, or through the ATE website. They can also leave their curriculum vitae on file for prospective employers to access. ATE offices are organized by province (69 regional offices), by skill level (10 bureaux de l'emploi des cadres) and by sector (7 sector-based offices, depending on main economic activity of the region). Local and foreign job vacancy advertisements are kept on file for reference. The job-bank database is divided into 3 categories: 1) those with higher education; 2) those with vocational training certificates; and 3) other education/skill level characteristics. II. GOVERNMENT CONCERNS AND STRATEGIES FOR MONITORING EMPLOYMENT As mentioned above, the Tunisian authorities are seeking to implement a quarterly employment survey. This survey sample size, sampling methodology, and panel aspects have yet to be determined. In order to better monitor the employment situation in Tunisia, and to meet its commitment to adhere to the SDDS, the INS is interested in upgrading the quality as well as increasing the frequency (i.e. quarterly) of the employment surveys. To assist the INS, a statistical trust fund (TFSCB, or Trust Fund for Statistical Capacity Building), through international technical assistance, will: (i) assess the current surveys and production of quarterly surveys; as well as (ii) carry out an evaluation of the national system of monitoring in the areas of employment and labor markets, along with recommendations for improvement. The request for the Trust Fund is currently under review by the TFSCB Committee. III. INTERNATIONAL EXPERIENCE In the context of Tunisia's plans to upgrade both the frequency and dissemination of the employment surveys, it may be useful to compare high-frequency strategies in selected MENA and OECD countries. Morocco Since 1999, Morocco's Statistics Department, under the Ministry of Economic Projections and Planning. has conducted a quarterly employment survey. The purpose of the survey is to provide timely data on employment for policy makers. The employment survey distinguishes between urban and rural areas. The quarterly surveys cover a sample size of 48,000 households (of which 32,000 are urban and 16,000 are rural households). Quarterly updates are available on the Statistic Department's website. Palestine The Palestinian Central Bureau of Statistics (PCBS) produces a quarterly Labor Force Survey. The program started in 1995, and 22 rounds have been conducted so far. A quarterly publication is issued detailing the main characteristics of the Palestinian labor force based on the main components (employment, unemployment, and underemployment). Main indicators are available for free on the PCBS website, and more detailed publications are available on a fee basis. 40 ANNEX9 A complete population census has not been conducted in Palestine since 1967, so a sampling frame of units was developed instead, covering the whole country. These units, called Primary Sampling Units (PSUs), are further stratified by district, place of residence (municipality, village, or refugee camp), size of locality, and by cell identification of locality. The sample size covers approximately 7,600 households. Since the quarterly surveys were introduced, the same households remain in the sample over 6 consecutive rounds. United States The Bureau of Labor Statistics (BLS) compiles two major monthly surveys on non-agricultural employment, including hours of work and earnings: the Current Population Survey (CPS) and the Current Employment Statistics (CES). Publications and time series from these surveys are available on the BLS website. On the supply side, the CPS produces a monthly survey, based on a sample survey of 50,000 households located in 792 sample areas, representing all counties and cities in the U.S., covering all 50 states and the District of Columbia. The monthly survey provides data on the labor force (employed and unemployed), broken down by demographics such as age, sex, race, occupation, etc. The surveys are conducted by trained interviewers. On the demand side, the CES monthly survey is collected from a sample of payroll reports from 390,000 businesses (employing over 47 million non-agricultural workers, who work full or part time). Data collected includes industry information on wage and salary employment, and average weekly hours and earnings. Canada Statistics Canada conducts two major monthly surveys on both the demand and supply side: (i) the Survey of Employment, Payroll and Hours, and (ii) the Labor Force Survey. The sample for the Survey of Employment, Payroll, and Hours is drawn from a census of the administrative records of Revenue Canada of firms with 100 or more employees, and a sample of firms with fewer than 100 employees. The survey measures (i) payroll employment, (ii) paid hours, and (iii) earnings in most non-agricultural industries. The survey is conducted monthly and covers a sample size of 96,000 companies, with a response rate of approximately 85 percent. The Labor Force Survey covers 53,000 households, utilizing the same households for a 6 month period. Data collected includes labor market activities and demographic characteristics of the working-age population. The response rate is approximately 95 percent. Recent results for both surveys are available free of charge on Statistics Canada's website with a 1 month delay, along with more detailed analysis available on a fee basis. Ireland The Irish Central Statistics Office conducts 3 quarterly surveys related to employment: (i) the Quarterly National Household Survey (QNHS), (")11 a survey of Industrial Employment, and (iii) a survey of Industrial Earnings and Hours Worked. The QNHS began in September 1997, replacing the annual Labor Force Survey. The survey produces quarterly labor force estimates as well as publications on special topics. Each week 3,000 households are surveyed, for a total sample of 39,000 households per quarter. The households are asked to participate in the survey for 5 consecutive quarters and are subsequently replaced by other households in the same sample area (block). 41 ANNEX 9 The Industrial Employment survey covers industrial businesses employing 3 or more persons. The survey includes (i) Total Employees by Industrial Sector and (ii) Total Persons Engaged by Industrial Sector (the latter including business owners and unpaid family members). The survey of Industrial Earnings and Hours Worked provides estimates for average earnings and hours worked for all industrial workers. The estimates are compiled by weighting the averages for the sample units using 3 size categories (lo-49 persons engaged, 50-99 persons engaged, and lOO+ persons engaged). Quarterly summary publications for all 3 surveys are available on the Irish Central Statistics Office website. Spain Spain conducts a continuous quarterly Labor Force Survey (LFS) and a Labor Costs Index Survey (LCIS) (since 1999 all weeks are covered, prior to 1999 August was not covered). The LFS collects employment-related demographic information, and measures underemployment (IL0 standard). The sample frame comprises 64,000 households, based on the last Population Census (done every 10 years). Each quarter, one-sixth of the households are changed, and a household is interviewed for 6 consecutive quarters at most. Initial interviews are done in person and follow- up is done by telephone. The LCIS encompasses wage and salary-related data, including overtime. Each quarterly sample covers 19,000 units, excluding the agriculture and public administration sectors, sourced from Social Security files. All businesses with 500+ employees are included in the sample, while for those with less than 500 employees, a probability sample methodology is used. Data is collected monthly by mailed questionnaires. Portugal Since 1998, Portugal's National Statistics Institute has conducted a quarterly Labor Force Survey in line with the harmonized EU methodology. The current sample covers 21,000 households drawn from the 1991 population census. Publications are available on the National Statistics Institute website covering employment and unemployment by region, age, economic classification, and first-time job seekers. The National Statistics Institute also conducts a quarterly survey on wages and earnings. Businesses with 10 or more full-time employees are surveyed, initially in person and subsequently by mail. The samples are renewed every 5 years, currently covering approximately 2,500 businesses. France Quarterly employment data is compiled by France's National Institute of Statistics and Economic Studies (INSEE). The following sources are used in the computations: 1) the population census; 2) quarterly establishment survey; 3) quarterly statistics from UNEDIC (organization in charge of collecting employer social contributions); and 4) quarterly statistics from URSSAF (another organization which collects employer social contributions). The data is seasonally adjusted and consistent with IL0 methodology. Quarterly publications on all employment related topics are available on INSEE's website. Quarterly unemployment data is also compiled by INSEE, utilizing the annual Labor Force Survey. The data is updated on a monthly basis using an econometric equation using monthly data on the number of month-end job seekers at the National Employment Agency (ANPE). The data is also seasonally adjusted. 42 ANNEX 9 INSEE calculates 2- wage indices on a quarterly basis: 1) Hourly Base Wage Rate index for "blue collar" workers; and 2) Monthly Base Wage Index for all employees (mostly private sector, excluding agriculture and social sectors). These data are collected from a quarterly survey carried out by the Ministry of Labor on industries with 10 or more employees. The data is not seasonally adjusted, and currently covers 45,000 local units since the survey was renewed in 1999. IV. RECOMMENDATIONSFORMONITORINGTHELABORMARKET(TOBECOMPLETEDWITH GOVERNMENT) A. Supply Sidtimployment Surveys at the Household Level (0 Improving the quality of Employment Surveys by better measuring some information and adding additional disaggregation in the questionnaires, such as (1) improving the liability of employment creation data in agriculture sector; (2) seasonal and part time jobs employment; (3) information on term contracts; (4) employment information for separating informal and formal sectors, public/private and offshore/onshore enterprises; (5) labor migration to outside the country; etc. ther Non- Other indicator - administratives 26 21 Liov't. gov't method used `rogram program for setting up own business 2000: Reason 1Finished IQuitschool Finished Finished dilitary Laid off Temporaq :signed Job Employed health Family Returned Othc vocational intership ervice firm ended part time easons reasons from ~~~~ployment/lchool / training cutbacks abroad disaggregation 2000: Regular/ pcrm. Conjoncture Seasonal Didn't `irst Rhythm of work last mejob 1work - further year earth Yes, anywhere Yes, nearby Yes, if NO gouvcrtl. Tunis district ANNEX 10 Statistical Annexz2 I. Trends 1. Labor Market Trends 1997 - 200 1 2. Agriculture and Nonagriculture Trends by Region and Urban/Rural II. Labor Force Participation 3. Working Age Population - Urban/Rural, Gender, District 4. Education/Gender 5. Education/Gender - Total Working Age Population, Active Working Age Population, And Inactive Working Age Population 6. Education - Active/Inactive 7. Potential/Marginal/Total - By Region, Gender, Marital Status, Education 8. Education and Employed/Unemployed 9. Age and Gender 10. Education, Gender, and Employed/Unemployed III. Unemployed 11. Total, Region, Gender, Main Cities 12. By Occupation 13. Youth Unemployment - By Age and Education 14. By Age and Education (Further Breakdown of Age) 15, Unemployment Duration in Months - New Entrants/Previously Employed - By Age Breakdown 16. Distribution (Share and Rate), by Gender, Education, Vocational Training, Age, Urban, Occupation 17. Profiles - New Entrants/Previously Employed - Average Age, Duration, Gender, Education 18. Age and Gender IV. Employment Status - Full Time, Part Time, Seasonal 19. Gender, Average Hours, Average Days 20. Gender, Education, Age, Urban, Sector V. Characteristics - Average Age, Gender, Education, Occupation, Vocational Level 21. Public Administration, Textile Worker, Average Private Employee 22. Unemployed Workers Using Age 23. Self-Employed Workers Using Micro-Finance Programs 24. Workers Employed in Public Works 22 Unless otherwise noted, "labor force" is defined as including those aged 15 - 64, for the purpose of calculating participation rates. However for "employed" or "unemployed," the age group encompasses those aged 15 or over in order to capture all who are unemployed, unless otherwise noted. ANNEX 10 1. Labor Market Trends 1997- 2001 Levels Growth Rates 1997 1999 2000 2001 1997- 1999- 2000- 1997- 1999 2000 2001 2001 Labor Force (15 - 64) 2,978,334 3,143,880 3,215,698 3,292,736 6% 2% 2% 11% Male 2,255,734 2,370,018 2,419,174 2,468,386 5% 2% 2% 9% Female 722,600 773,862 796,524 824,350 7% 3% 3% 14% Urban 1,912,800 1,997,617 2,078,150 2,156,459 4% 4% 4% 13% Large Cities 908,965 939,552 961,398 999,766 3% 2% 4% 10% Other Urban 1 1,003,8351 1,058,065/ 1,116,752/ 1,156,6931 5%/ 6%1 4%) 15%1 Rural 1,065,534 1,146,263 1,137,548 1,136,277 8% -1% 0% 7% Employment (>=15) 2,503,572 2,634,965 2,704,928 2,788,780 0% 3% 3% 11% Male 1,906,400 1,992,078 2,039,462 2,095,431 4% 2% 3% 10% Female 1 597,172 1 642,887 1 665,466j 693,349 1 8%( 4%1 4%/ 16%/ Part Time 496,662 438,756 595,410 452,211 -12% 36% -24% -9% Seasonal 343,590 405,063 547,113 342,652 18% 35% -37% 0% Sectors Agriculture 546,166 595,937 593,023 609,793 9% 0% 3% 12% Manufacturing Excl. Textiles 247,289 239,303 249,575 291,759 -3% 4% 17% 18% Textiles (Incl. Clothing, Leather) 259,233 242,009 268,189 276,909 -7% 11% 3% 7% Nonmanufacturing Industry 32,865 32,704 33,195 33,451 0% 2% 1% 2% Construction 304,845 371,943 339,967 337,039 22% -9% -1% 11% Services (Excl. Civil Service, Health, Education) 672,944 696,247 724,25 1 768,361 3% 4% 6% 14% Public Admin, Educ, Health Services (Public, Private) 413,230 443,492 473,163 452,673 7% 7% -4% 10% O/W Public Administration, Public Health & Educ', * 406,874 na 461,056 442,098 -4% 9% O/W General Public Administration 193,866 na 230,602 213,938 -7% 10% Outside of Tunisia 2,174 2,000 1,663 2,120 -8% -17% 21% -2% Unemployment (>=15) 474,762 508,915 5 10,770 503,956 7% 0% -1% 6% Male 349,334 377,940 379,112 372,955 8% 0% -2% 1% Female 125,428 130,975 131,058 131,001 4% 0% 0% 4% First Time Job Seekers 168,153 183,396 159,323 152,401 9% -13% -4% -9% Previously Employed 306,609 325,519 351,447 351,555 6% 8% 0% 15% =15) IYYI I a-m Agricultural Nonagr. Total rl Total Labor Not in Employment Employment Employed b nempll`eyed Force Labor Force Total Tunis 36,047 1 544,621 1 580,668 ( 114,212 694,880 681,916 1,376,796 Northeast 109.099 I 262.636 1 371.735 1 63.612 1 435.347 I 419.866 1 855.213 1 khwest I 112.779 t 192.056 1 304.835 1 81.846 1 386.681 1 451.185 1 837.866 1 ICentral West I 132.426 1 192.601 1 325.027 1 72.765 1 397.792 1 427,481 1 825.273 1 ICentral East I 95,585 1 512.477 1 608,062 1 81,681 I 689,743 1 630,528 1 1.320.271 1 Southwest 33,170 1 93,929 1 127,099 [ 26,955 1 154,054 1 200,099 1 354,153 Southeast 27,060 1 158,972 1 186,032 1 33,691 1 219,723 1 335,103 1 554,826 I Total I 546,166 1 1,957,292 12,503,458 1 474,762 1 2,978,220 1 3,146,178 1 6,124,398 1 1999 Agricultural Non-agr Total Unemp,-A/TotalLaborI P employment employment Employed 1 85.585 1 409.050 1 471.991 1 881.041 1 rCentra1 West 131.276 1 207,764 1 339.040 1 74.121 1 413.161 1 464.432 1 877.593 1 ICentral East II 127,628 1 495,339 1 622,967 1 96,641 1 719,608 1 689,781 1 1.409.389 1 Southwest 30,493 1 96,075 1 126,568 ) 33,527 1 160,095 ) 215,284 1 375,379 Southeast 41,737 I 169,690 1 211,427 1 36,780 1 248,207 1 340,456 1 588,663 I Total I 595,937 1 2,027,698 12,623,635 1 508,915 1 3,132,550 1 3,378,074 ) 6,510,624 1 t 2000 Agricultural Non-Agr Total ~nemp,-.4ITotalLaborI I Employment Employment Employed rCenhdWest 210.892 1 352.485 1 60.435 1 412.920 1 471.756 1 884.676 1 I Central East II 141.593 I 131,809 1 536,024 1 667,833 1 83,627 1 751.460 1 700,116 1 1.451.576 1 ISouthwest I 23,167 1 92,970 I 116,137 I 32,169 I 148,306 1 239,990 1 388,296 1 Southeast 26,396 1 167,615 1 194,011 1 50,443 1 244,454 1 371,500 ) 615,954 Total 593,023 1 2,090,003 12,683,026 1 510,770 1 3,193,796 ( 3,486,762 1 6,680,558 t 2001 Agricultural Non-Agr Total ~nemp,n.dITotalLaborI Employment Employment Employed 652,412 427.351 340.910 kentral West 173.777 I 185.399 1 359.176 1 59.869 1 419.045 I 468.355 1 887.400 1 I Central East II 91,391 I 557.482 1 648.873 1 99.834 1 748.707 1 749,861 1 1.498.568 1 I Southwest I 33,640 1 106,495 1 140.135 1 26,862 1 166.997 1 232,780 1 399.777I I Southeast I 26.124 1 177,124 1 203,248 1 46,796 1 250,044 1 368,418 1 618,462 1 I Total I 609,793 1 2,162,312 I2,772,105 1 503,956 1 3,276,061 1 3,581,726 1 63857,787 1 Source: Employment Surveys, INS. 53 ANNEX 10 Employment and Labor Force Trends by Urban/Rural (All Are >=15) 2000 Agricultural Non-Agr Employment Employment Total I I I I I I I Large Cities 16,784 789,78 1 [ 806,565 1 150,158 956,723 1 998,126 1 1,954,849 Urban Total 98,819 1,716,614 1,815,433 33 1,878 2,147,311 2,330,396 4,477,707 Rural 5 10,974 445,698 956,672 172,078 1,128,750 1,25 1,330 2,380,OSO Total 609,793 2,162,312 2,772,105 503,956 3,276,061 3,58 1,726 6,857,787 Source: Employment Surveys, INS. 54 ANNEX IO Working Age Population - Urban/Rural, Gender, District Source: Employment Surveys, INS. 55 ANNEX 10 4. Education/Gender (15-64) - 2001 r Labor Force Participation Rates 1997 (Percent) I rEducation/Gender IMalelFemalelTotall Total 77.2 25.5 51.4 None 80.5 16.0 33.9 Primary Complete 89.2 27.9 61.9 I I I Secondary 163.21 31.1 150.2 Post-Secondary 173.3 1 58.5 168.0 Labor Force Participation Rates 1999 (Percent) Education/Gender Male Female Total Total 77.0 25.6 51.2 I I , None 179.9) 17.4 135.4 Primary Incomplete (KouttebiLiterate) 74.5 18.8 63.2 Primary Complete 89.1 28.1 61.6 Secondary 64.7 28.1 49.0 I Post-Secondary 171.0 1 56.1 165.4 Labor Force Participation Rates 2000 (Percent) Education/Gender Male FemaleTotal Total 76.5 25.7 51.0 I I 1 None I81.1 1 16.7 134.1 I Primary Incomplete (KouttebiLiterate) 173.01 20.4158.7) Primary Complete 89.0 27.7 60.9 Secondary 64.9 28.0 49.1 Post-Secondary 71.6 54.8 65.3 Labor Force Participation Rates 2001 (Percent) I I Education/Gender Total 76.0 25.8 50.9 None 79.0 17.7 33.6 Primary Incomplete (KouttebiLiterate) 74.1 22.3 59.9 Imary Complete 188.61 27.5 )60.6 ) 1 Secondarv 164.01 27.2 148.4 1 Post-Secondary 173.01 54.9 )65.7 1 Source: Employment Surveys, INS. 56 ANNEX 10 5. Education/Gender - Total Working Age Population, Active Working Age Population, and Inactive Working Age Population Distribution Source: Employment Surveys, INS. 57 ANNEX 10 Secondary 69.5 74.6 46.4 22.9 17.0 12.0 8.1 5.7 34.0 Post-Secondary 0.1 1.2 15.7 12.5 1.5 0.8 1.0 1.6 4.5 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Employment Surveys, INS. 58 ANNEX 10 Source: Employment Surveys, INS. 59 ANNEX IO Education Distribution/Age Group 2001 115-17118-19 ) 20-24 125-29 1SO-39 ) 40-49 150-59 I >=60 I Total Post-Secondary 0.1 0.6 17.7 14.0 1.8 0.8 1.3 1.2 4.5% 1 Total 1100.01 100.0~ 100.0~ 100.0~ 100.01 100.0~ 100.0~ 100.0~100.0%~ Source: Employment Surveys, INS. 60 ANNEX 10 Education - Active/Inactive Labor Force Participation, Active/Inactive by Education Distribution 1997 (15-64) 1 Active 1 Inactive 1 Active 1 Inactive 1 None 1 489,1191 955,594 1 17%1 35%1 Primary Complete 1,297,354 797,355 45% 29% Secondary 883,084 876,554 31% 32% Post-Secondary 210,257 98,830 7% 4% I I I I Total 1 2,879,814 1 2,728,333 1 loo%/ 100% I Labor Force Participation, Active/Inactive by Education Distribution 2001 (15-64) I Active Inactive Active Inactive None 456,649 900,634 14% 29% Primary Incomplete 40,385 27,070 1% 1% I Primary Complete 1 1,330,300 I 866,108 1 28%) Secondary 1,054,819 1,122,860 33% 37% Post-Secondary 297,057 155,272 9% 5% I I I I Total 1 3,179,210 1 3,071,944 1 lOO%l 100% Source: Employment Surveys, INS 61 ANNEX 10 Potential/Marginal/Total - by Region, Gender, Marital Status, Education Characteristics bv Labor Force Attachment 1997 Marital Status Single 9,355 I 27,074 1 5,603,469 1 73% 1 32% 1 61% Married 2,491 1 52,888 1 3,243,328 1 19%1 63%) 35% Widowed 362 1 2.522 1 293.531 1 3%1 3%1 3% I I I I Divorced 623 1 `9101 55,792 1 ' 5%1 l%l 1% 1 Education I I I I I I I 2,875 44,179 2,426,033 22% 53% 29% Complete 6,949 28,364 3,718,555 54% 34% 45% I Secondarv I 2,791 I 9,592 1,890,775 22% 12% 23% 1 Post-Secondary 2161 1,259 1 311,759I 2% 2% 4% Characteristics by Labor Force Attachment 2001 Potential Marginal Total' Potential Marginal Total' Region Urban 8,344 16,625 4,121,516 67% 35% 66% Large Cities 3,477 6,996 1,885,685 28% 15% 30% Other Urban 4,867 9,629 2,235,831 39% 20% 36% I Rural I 4.020 1 30.390 1 2.130.064 1 33% I 65% 1 34% I Widowed 276 1 1.209 1 123.935 1 r-divorced I 1921 502 1 49,598 1 2%1 l%l l%l Education None 2,133 24,491 1,357,283 17% 52% 22% Primary Incomplete 255 670 67,455 2% 1% 1% Primarv ComiAete 5.375 12,596 2,196,408 43% 27% 35% Secondary 4,050 5,165 2,177,679 33%1 11%1 35% Post-Secondary 551 4,093 452,329 4%/ 9%1 7% a Total working-age population. Source: Employment Surveys, INS. 62 ANNEX IO Education, Also Employed/Unemployed (15-64, Labor Force, >=15 Empl/Unempl) Distribution of Educational Attainment for the Labor Force 1999 63 ANNEX 10 Source: Employment Surveys, INS. 64 ANNEX IO 9. Age and Gender - 1997 and 2001 Labor Force Participation Male Female Total 1997 2001 ,m poJ 1997 g&l 15-19 187258 173988 102783 81950 29004 1 255938 20-29 641070 687553 293753 321567 934823 1009120 30-39 629788 672378 175763 226644 80555 1 899022 40-49 408726 525367 94348 123339 503074 648706 50-59 230085 247392 37470 45756 267555 293 148 60-64 69022 62835 9748 10791 78770 73626 Total 2165949 2369513 713865 810047 28798 14 3179560 Labor Force Participation Rate Male Female Total 1997 2001 1997 2001 1997 2001 15-19 37.1% 31.8% 21.1% 15.6% 29.3% 23.9% 20-29 77.2% 75.2% 35.5% 35.4% 56.4% 55.4% 30-39 97.1% 97.0% 26.9% 31.1% 61.8% 63.3% 40-49 96.5% 96.6% 22.2% 23.1% 59.3% 60.2% 50-59 84.6% 84.0% 13.3% 14.9% 48.3% 48.8% 60-64 54.1% 50.0% 7.7% 8.2% 30.9% 28.6% Total 77.2% 76.0% 25.5% 25.8% 51.4% 50.9% Source: Employment Surveys, INS 65 ANNEX 10 10. Education,Gender,andEmployedKJnemployed- 1997 and2001 1997: Labor Force Participation, Employment, and Unemployment Education Participation Employment Unemployment All 2879814 2407411 472403 Primary 1297354 1043848 253506 2nd l e r Cycle 387401 314633 72768 2nd 2em Cycle 495683 422147 73536 Superior 210257 193135 17122 Nothing 489119 433648 55471 Male 2165949 1818784 347165 Primary 1036760 843270 193490 2nd ler Cycle 301930 246891 55039 2nd 2emCycle 358921 307960 50961 Superior 146110 135390 10720 Nothing 322228 285273 36955 Female 713865 588627 125238 Primary 260594 200578 60016 2nd l e r Cycle 85471 67742 17729 2nd 2emCycle 136762 114187 22575 Superior 64147 57745 6402 Nothing 166891 148375 18516 Education Participation Rate EmploymentRate Lnemployment Rate All 5 1.4% 42.9% 16.4% Primary 61.9% 49.8% 19.5% 2nd ler Cycle 47.7% 38.8% 18.8% 2nd 2em Cycle 52.3% 44.5% 14.8% Superior 68.0% 62.5% 8.1% Nothing 33.9% 30.0% 11.3% Male 77.2% 64.8% 16.0% Primary 89.2% 72.6% 18.7% 2nd ler Cycle 63.0% 5 1.6% 18.2% 2nd 2em Cycle 63.3% 54.3% 14.2% Superior 73.3% 67.9% 7.3% Nothing 80.5% 71.3% 11.5% Female 25.5% 21.O% 17.5% Primary 27.9% 21.5% 23.0% 2nd ler Cycle 25.7% 20.4% 20.7% 2nd 2em Cycle 35.9% 30.0% 16.5% Superior 58.5% 52.7% 10.0% Nothing 16.0% 14.2% 11.1% Age group: 15 - 64, "Employment Rate" is defined here ns number of employed over the total population. 66 ANNEX 10 2001: Labor Force Participation, Employment, and Unemployment Education Participation Employment Unemployment All 3179210 2681267 497943 None 456649 409296 47353 Primary Incomplete 40385 35945 4440 Primary Complete 1330300 1096150 234150 Secondary 1054819 874739 180080 Post-Secondary 297057 265137 31920 Male 2369289 2000986 368303 None 278692 247403 31289 Primary Incomplete 36261 32401 3860 Primary Complete 1052871 869318 183553 Secondary 804908 673197 131711 Post-Secondary 196557 178667 17890 Female 80992 1 68028 1 129640 None 177957 161893 16064 Primary Incomplete 4124 3544 580 Primary Complete 277429 226832 50597 Secondary 249911 201542 48369 Post-Secondary 100500 86470 14030 Education Participation Rate Employment Rate Unemployment Rate All 50.9% 42.9% 15.7% None 33.6% 30.2% 10.4% Primary Incomplete 59.9% 53.3% 11.0% Primary Complete 60.6% 49.9% 17.6% Secondary 48.4% 40.2% 17.1% Post-Secondary 65.7% 58.6% 10.7% Male 76.0% 64.2% 15.5% None 79.0% 70.1% 11.2% Primary Incomplete 74.1% 66.2% 10.6% Primary Complete 88.6% 73.1% 17.4% Secondary 64.0% 53.5% 16.4% Post-Secondary 73.0% 66.4% 9.1% Female 25.8% 21.7% 16.0% None 17.7% 16.1% 9.0% Primary Incomplete 22.3% 19.2% 14.1% Primary Complete 27.5% 22.5% 18.2% Secondary 27.2% 21.9% 19.4% Post-Secondary 54.9% 47.2% 14.0% Age group: 15 - 64, "Employment rate" is dczfined here as #@`employed over the total population. Source: Employment Surveys, INS. 67 ANNEX IO 11. Total, Region, Gender, and Main Cities - 1997- 2001 Unemployment Rates (Percent) Female 17.4 16.9 16.5 15.9 I Urban Female 1 16.71 16.71 17.21 16.71 Rural Female 18.7 17.4 16.4 14.3 Main Cities Tunis 18.2 17.0 16.5 16.7 Ariana 14.6 12.7 20.6 12.6 Benarous 16.1 16.0 14.0 12.7 Bizerte 14.2 15.4 19.1 16.6 Kairouan 16.0 17.5 10.8 12.7 Sousse 10.4 14.4 16.9 16.5 Sfax 11.8 8.8 8.7 9.7 Gabes 13.0 16.8 17.0 16.9 Source: Employment Surveys, INS. 68 ANNEX 10 12. Occupation - 1997-2001 r Occupation I Percent Unemployment Rates (>=15) I I 1997 I 1999 I 2000 I 2001 I Administrator/Manager 1.5 1.1 2.6 3.5 Tech.SpecialistlProfessional 1.5 8.6 6.2 17.0 Int. Professional 4.0 5.5 4.7 6.5 Office Clerk, Cashier 7.1 9.6 8.9 15.4 Sales/Service Worker 8.8 1 9.7 1 11.0 1 13.9 Farmer/Skilled Agriculture 5.4 4.5 5.1 4.5 Trade, Craft, Artisan 14.5 19.6 17.7 18.7 Plant/Machine Operator 9.3 10.2 9.9 10.6 Ekilled Labor Occupation Percent Labor Force 1997 I 1999 I 2000 I 2001 mnistrator/Manager 2.1 1 Tech. Specialist/Professional 3.9 4.1 4.5 5.0 Int. Professional 7.8 7.0 7.4 6.8 Office Clerk, Cashier 4.8 5.5 6.3 5.5 E/Service Worker Farmer/Skilled Agriculture 16.1 13.1 15.2 14.3 Trade, Craft, Artisan 23.4 16.3 16.9 17.2 Plant/Machine Operator 6.5 10.7 12.1 9.8 bkilled Labor 23.1 1 22.2 1 Source: Employment Surveys, INS. 69 ANNEX 10 Youth Unemployment - By Age and Education (Percent) I Youth Unemployment Rates and Distribution by Education 1997 U Rate Distribution of Unemployed 15-19 Years 20-29 Years 15-19 Years 20-29 Years None 29.6 18.3 6.5 6.0 lpriiary Complete I 33.6 1 25.0 1 74.1 1 45.3 1 I Secondary I 40.4 1 27.4 1 19.4 ) 43.2 1 Post-Secondary 19.3 0.0 5.5 Total 34.4 25.0 100.0 100.0 Source: Employment Surveys, INS. 70 ANNEX 10 Age and Education (Further Breakdown of Age) :mDlovment bv Age And El 46.848 1 13,281 1 50-59 13,814 3,162 838 187 >=60 5,666 988 23 0 Total 474.762 253.693 146327 17.122 Age All Levels Primary Secondary Post-Secondary 15-17 35.8% 35.3% 40.5% 18-19 32.9% 31.4% 40.3% 20-24 30.1% 27.1% 36.6% 30.5% 125-29 I 20.5% 1 22.1% I 20.7% 1 17.1% I 30-39 11.1% 14.2% 7.6% 4.5% 40-49 6.6% 9.0% 3.4% 1.1% 50-59 I 5.2% 1 4.1% 1 2.0% 1 1.2% >=60 3.2% 1 2.7% 1 0.4% I 0.0% Total 15.9% 1 19.3% ) 16.5% 1 8.1% Unemployment by Age and Education 2001 Aee All Levels 1 PrimarvI , I Secondarv" IIPost-SecondarvY 15-17 43,517 I 18-19 43,667 26,095 15,360 1 20-24 135,768 59,472 65,310 1 5,517 25-2' 119.523 I 46.343 1 50.455 I 16,690 1 30-39 93,668 41,139 31,162 8,723 40-49 39,944 22,400 7,066 752 SO-59 16.294 4.583 2.33 1 154 1 11.575 I 72 1 Total 503,956 234,603 180,150 31,920 Age All Levels Primary Secondary Post-Secondary 15-17 37.0% I9 31.6%/ 30.0% I 36.5% 1 20-24 29.2% 24.9% 35.5% 40.5% 25-29 21.9% 21.1% 23.1% 23.4% 30-39 10.4% 11.9% 9.5% 7.6% riG40-49 I 6.2% 1 8.4% 1 3.6% 1 l.l%l 50-59 5.6% 6.2% 3.6% 0.6% >=60 6.2% 5.3% 9.4% 3.1% Total 15.3% 17.5% 17.0% 10.7% Source: Employment Surveys, INS. 71 ANNEX 10 Unemployment Duration In Months - New Entrants/Previously Employed By Age Breakdown Unemployment Duration (Months, 1997) 1 New Entrants 1 Previouslv EmDloved o-12 30% 79% 13-24 26% 11% 25-36 18% 5% 37-48 12% 3% 49-60 7% 1% 61-72 5% 1% >72 3% 0% Unemployment Duration (Months, 1999) New Entrants Previously Employed o-12 31% 76% 13-24 31% 16% 1 25-36 I 15%) 4%1 37-48 9% 2% 49-60 7% 1% 61-72 3% 1% >72 4% 0% Unemployment Duratibn (Months, 2000) New Entrants Previously Employed o-12 41%l 71% 13-24 32% 20% 25-36 14% 5% 1 37-48 I 8%1 2%1 49-60 3% 1% 61-72 2% 0% >72 1% 0% Unemployment Duration (Months, 2001) New Entrants Previously Employed o-12 49% 75% 13-24 26% 17% 1 25-36 I 13%l 5%l 37-48 6% 2% 49-60 2% 1% 61-72 0% 0% 1 >72 I l%l O%I Note: Excludes potentials Source: Employment Surveys, INS. 72 ANNEX IO 16. Distribution (share and rate), by Gender, Education, Vocational Training, Age, Urban, Occupation (>=15) 73 ANNEX 10 Unemployment Distribution 2001 (Percent) I 1 Share ) Rate I Gender Male 75.7 15.1 Female 24.3 14.7 Eation I I I None 10.11 9.2 Primary Incomplete 1.0 8.9 Primary Complete 46.6 17.2 I I Secondary 35.8 1 16.7 r~ Post-Secondary 10.5 1 Vocational Training Apprent. Prof 17.7 22.7 Diplome Formation Prof. 10.0 17.8 Cap 12.1 21.9 Btp 8.9 15.2 Bts 1.1 10.1 Other Vocational Training 17.4 19.9 Other (Incl. Non-Vocational) 32.8 9.1 Age 15-19 17.4 33.6 20-29 50.7 24.8 30-39 18.6 1 10.2 40-49 7.9 6.0 50 Or Over 5.4 5.6 rILocation I I I Large City Other Urban 37.2 15.9 Rural 34.2 14.8 Occupation I Administrator/Manager 1.1) 3.4 Tech. Specialist/Professional 6.5 1 16.7 Int. Professional 3.4 6.3 Office Clerk, Cashier 6.4 14.9 I t Sales/ServiceWorker 14.0 I 13.6 Farmer/Skilled Agriculture 5.2 1 4.5 Trade, Craft, Artisan 24.7 18.2 Plant/Machine Operator 8.0 10.3 Unskilled Labor 30.7 16.0 Source: Employment Surveys, INS. 74 ANNEX 10 17. Profiles - New Entrants/Previously Employed Average Age, Duration, Gender, Education (>=15) Post-Secondary Source: Employment Surveys, INS. 75 ANNEX 10 18. Age and Gender Male Female Total Male Female Total 65071 22,113 87,184 37.4% 27.0% 34.1% I 180,150 1 75,141 I 255,291 1 1 26.2% 1 23.4% 1 25.3% 1 30-39 70,702 22,966 93,668 10.5% 10.1% 10.4% 40-49 33,298 6,646 39,944 6.3% 5.4% 6.2% 50+ 23,734 4,135 27,869 5.8% 5.8% 5.8% Total 372,955 131,001 503,956 15.1% 15.9% 15.3% Source: Employment Surveys, INS. 76 ANNEX IO Gender, Average Hours, Average Days - 1997-2001 Employment Status - Full time/part time/seasonal (>=15) Shares 1 1997 I 1999 I 2000 I 2001 Total Employment Male 73.6% 75.4% 75.4% 75.1% Female 26.4% 24.6% 24.6% 24.9% Full Time I Male 1 78.0%1 77.9% 1 76.7% 1 Female 22.0% 22.1% 23.3% Part Time Male 62.7% 66.4% 66.9% I I I I Female 1 37.3%) 33.6% 1 33.1% Permanent Male 73.9% 73.7% Female 26.1% 26.3% SeasonalIConjoncturel Male 73.9% 85.7% Female 26.1% 14.3% SeasonaKonjoncturel I Full Time 1 64.8%/ 66.9%1 59.8% 1 67.5% 1 1 Part Time 1 35.2%/ 33.1% t 40.2%1 32.5% 1 Permanent Full Time 83.5% 85.9% 84.1% 86.1% I Part Time 1 16.5%1 14.1%1 15.9% I 13.9% I Average Hours Worked Last Week 44.5 45.1 43.8 46.6 Full Time 50.3 49.8 49.9 51.2 Part Time 21.3 22.3 21.4 22.3 77 ANNEX IO 20. Gender, Education, Age, Urban, Sector - 2001 I I Source: Employment Surveys, INS. 78 ANNEX 10 21. Public Administration, Textile Worker, Average Private Employee - 2001 ' Those reporting place of work as public administration (variable 326). ' Wage earners employed in regular permanent jobs. Source: Employment Surveys, MS. 79 ANNEX 10 22. Unemployed Workers Using ATE - 2001 WJ 55.6 16.0 Bts 55.7 9.9 Other Formation Prof. 40.3 7.4 Age 15-19 24.6 4.2 20-29 36.8 7.0 30-39 30.1 6.0 40-49 18.4 3.4 50 Or Over 18.8 5.3 Location Large City 32.2 4.3 Other Urban 37.1 8.6 Rural 23.3 4.5 Reason For Unemployment First-Time Job Seeker 39.4 9.1 Occupation Administrator/Manager 34.0 7.5 TechSpecialistiProfessional 51.1 13.7 Int.Professional 53.2 15.0 Office Clerk, Cashier 52.4 14.2 Sales/Service Worker 31.0 5.3 Farmer/Skilled Agriculture 17.9 3.3 Trade, Craft, Artisan 30.0 4.4 Plant/Machine Operator 29.0 4.8 Unskilled Labor 22.% 4.1 All Workers 31.0 6.0 Source: Employment Surveys, INS. 80 - - - - - - - - - - - - - - - - - - - - - - - -2b. s - - - - - - - - - - - - -stIc) a! - - - - - - - - - - - - z.-g L- E ae 76 -u P5& c8s- -z9 ANNEX 10 24. Workers Employed in Public Works (%) - 2001 Characteristics of Workers Employed in Public Works 2001 (>=15) _ - Gender 100.0 Male 98.6 Female 1.4 Education 100.0 -None I 25.5 1 1 Primary Incomplete I 2.2 1 Primary Complete 54.2 Secondary 17.7 Post-Secondary 0.4 hocactional Training I 100.0 I Apprent. Prof 24.6 Diplome Formation Prof. 16.8 Cap 11.8 1 Btp I 5.4 1 Bts 0.0 Other Vocational Training 17.7 Other (Incl. Non Vocational) 23.7 I 100.0 I 15-19 4.7 20-29 27.0 I 30-39 29.6 I 40-49 I 23.6 1 50 Or Over 15.1 Location 100.0 Large City 17.4 Other Urban 33.4 Rural 49.2 Sector 100.0 Agriculture 9.0 Industry Manf. (Excl. Textiles) 1.0 Textiles 0.0 Industry Non-Manf. 0.2 Construction 86.9 Services: Health, Education, Public Admin 1.5 Other Services 1.4 Source: Employment Surveys, INS. 82