Policy Research Working Paper 10894 Can Public Credit Schemes Improve Access to Finance for Small Businesses? Evidence from Indonesia Salman Alibhai Hillary C. Johnson Cecile Thioro Niang Francesco Strobbe Finance, Competitiveness and Innovation A verified reproducibility package for this paper is Global Practice & available at http://reproducibility.worldbank.org, East Asia and the Pacific Region click here for direct access. September 2024 Policy Research Working Paper 10894 Abstract Examining one of the world’s largest public business sup- subsidized credit does not alleviate longer-term credit port programs, this paper studies how subsidized credit constraints by serving as a stepping stone to unsubsidized and partial credit guarantees shape access to finance for commercial credit in this context. The results highlight the micro and small businesses in Indonesia. The analysis uses challenge of reaching borrowers without collateral, even in administrative data on more than 8.4 million borrowers programs that explicitly target them using instruments such and unique quantitative and qualitative data to show that as partial credit guarantees. The paper sheds light on how subsidized credit can enable firms to access formal credit public credit schemes for small businesses can be designed for the first time and boost financial inclusion. However, to optimize inclusiveness and additionality. This paper is a product of the Finance, Competitiveness and Innovation Global Practice and the Gender Innovation Lab, East Asia and Pacific Region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at at aalibhai@worldbank.org, hjohnson1@ worldbank.org, cniang@worldbank.org, and fstrobbe@worldbank.org. A verified reproducibility package for this paper is available at http://reproducibility.worldbank.org, click here for direct access. RESEA CY LI R CH PO TRANSPARENT ANALYSIS S W R R E O KI P NG PA The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Policy Research Working Paper Can Public Credit Schemes Improve Access to Finance for Small Businesses? Evidence from Indonesia Salman Alibhai Hillary C. Johnson Cecile Thioro Niang Francesco Strobbe JEL Classification: G21, H81, E43, O16 Keywords: Government credit programs, subsidized credit, partial credit guarantees, micro-small and medium enterprises, financial inclusion Acknowledgements The authors would like to thank to the Coordinating Ministry for Economic Affairs and the Ministry of Finance of Indonesia for their partnership in providing access to the KUR program data and providing helpful inputs throughout the study. The authors are also grateful to Alexis Diamond, Tatiana Didier, Aufa Doarest, Michael Fuchs, Calvin Koenig, Clarita Kusharto, Aaditya Mattoo, Lalita Moorty, Vina Noor, Akaravuit Pancharoen, Habib Rab, and Iqbal Wibisono, who provided guidance and support throughout the design and implementation of this study. The authors also would like to thank Ana Maria Aviles, Pietro Calice, Ana Carvajal, Francesca de Nicola, Ivor Istuk, Harish Natarajan, Kiyotaka Tanaka, and Timothy Ogden for useful comments on the draft and the data collection team at REDI for their ability to deliver quality data under difficult circumstances. The team gratefully acknowledges funding from the Swiss State Secretariat for Economic Affairs (SECO) under the Indonesia Financial Sector Strengthening Program and from the East Asia Pacific Gender Innovation Lab (EAPGIL). EAPGIL is supported by the Umbrella Facility for Gender Equality (UFGE) in partnership with the Australian Department of Foreign Affairs and Trade. UFGE has received generous contributions from Australia, Canada, Denmark, Finland, Germany, Iceland, Ireland, the Netherlands, Norway, Spain, Sweden, Switzerland, the United Kingdom, the United States, the Bill and Melinda Gates Foundation (BMGF), and the Wellspring Philanthropic Fund (WPF). Finally, the authors would like to thank the entrepreneurs who volunteered their time to share their experiences with us. 2 1. Introduction Micro, small and medium-sized enterprises (MSMEs) comprise an estimated 95% of all businesses worldwide, accounting for a staggering 70% of global employment and 50% of global economic output. However, many MSMEs struggle to access sufficient finance to help their businesses grow and thrive.1 Economic literature highlights the role of interrelated market failures in limiting MSME access to finance— notably higher perceived credit risk, higher transaction costs, and lack of assets that can be offered as collateral. Governments around the world have pursued a wide range of measures to address the constraints that prevent MSMEs from accessing finance, including subsidized lending and partial credit guarantees (Carvajal and Didier, 2024). Subsidized lending programs have potential to increase investment, sales, and profits (Banerjee and Duflo 2014), and exports (Zia 2008; Kapoor, Ranjan, and Raychaudhuri 2017). Some evidence also suggests that partial credit guarantees can increase firm output (Arraíz et al. 2014) and firm sales, employment, wage levels, and survival (Oh et al. 2009). Improving access to finance through targeted programs thus has the potential to boost growth, jobs, and improve livelihoods. Although existing evidence suggests that subsidized lending can have positive impacts on firms, findings are mixed about the extent to which these programs reach the most credit constrained firms or provide additionality. More information is needed to assess the implementation conditions that can enable these goals. In this paper, we study one of the world’s largest public support programs for MSMEs, the People’s Business Credit (Kredit Usaha Rakyat, KUR) in Indonesia, to provide policy lessons on the potential and pitfalls of government-supported financing schemes for MSMEs.2 The KUR program has disbursed over 50 million loans totaling over US$ 100 billion in lending to small businesses since its inception in 2007. In its current form, KUR provides subsidies to Indonesian banks to lend to MSMEs at interest rates capped at 6%, roughly 10 percentage points below market rates. These loans are also backed by a publicly funded partial credit guarantee. Examining the KUR program and its beneficiaries, we aim to answer three research questions. First, to what extent do public support schemes provide additionality — reaching new borrowers or crowding in rather than crowding out private funds? Second, can subsidized credit serve as a stepping-stone 1 See for example Beck et al. (2005, 2008, 2011), Caprio and Demirguc-Kunt (1998), Demirgic-Kunt and Maksimovic (1998, 1999, 2002), Booth et al. (2001), Giannetti (2003), Bancel and Mittoo (2004), Lerner and Schoar (2005), Antoniou et al. (2008), de Jong et al. (2008), among others. 2 The KUR regulations do not provide a definition of MSME for the program; however, Indonesia’s SME Law 20/2008 defines microbusinesses in Indonesia as ones with less than 300 million IDR (US$20,000) in annual sales, small businesses as ones with between 300 million and 2.5 billion IDR (US$20,000 to US$ 166,667), and medium-sized businesses as having between 2.5 billion and 50 billion IDR in annual sales (US$166,667 to US$ 3.3 million) (IFC 2016). Data for this study show that the vast majority of KUR borrowers are microbusinesses when defined using approximate annual sales or an alternative definition of fewer than 5 salaried employees. See Appendix A for details. 3 to inclusion in commercial lending? Finally, what are the implementation conditions needed to promote additionality and effective targeting of MSME finance operations? To answer these questions, we use administrative loan-level data from the KUR program itself, which includes information on over 8 million debtors between 2015 and 2020. Using the program database as a sampling frame, we complement the administrative data with surveys of a representative sample of 1,402 KUR borrowers as well as qualitative interviews with separate samples of 100 KUR borrowers and 100 non- borrowers. In sum, these data sources provide a rich and novel set of insights into the profiles, perspectives, and experiences of KUR borrowers since 2015 and shed light on the ways in which the publicly supported scheme plays out in supporting MSMEs on the ground. Our main and novel finding is that the subsidized credit provided through KUR enhances financial inclusion by reaching previously excluded micro and small businesses; however, it does not serve as a stepping- stone to unsubsidized commercial lending. Prior to receiving a subsidized KUR loan for the first time, only 11 percent of borrowers had ever received a commercial loan from a bank. Although successful at reaching first-time borrowers, the longer-term impacts on their financial inclusion are uncertain because KUR borrowers do not move into unsubsidized commercial lending. Approximately half of KUR borrowers went on to receive KUR loans multiple times within a 5-year period, and only 3 percent of firms who received KUR financing graduated to other forms of commercial lending. Until recent regulatory changes, firms had little incentive to choose commercial sources of financing over KUR’s below-market interest rates, and banks preferred to encourage repeat lending to minimize risk. We also identify key design features of the KUR scheme which pose challenges for inclusivity and consumer protection and hold important lessons for the design of public MSME finance support schemes. First, we find that although the KUR program targets firms with insufficient collateral, in practice, 96% of KUR borrowers submit collateral for loans, with collateral worth an average of more than four times the amount of their loan. De facto collateral requirements impede the program from reaching the target audience and hinder women’s access to KUR. Second, we find that the program’s implementation model through financial institutions has advantages in connecting borrowers to the financial sector, but also leads to heterogeneity in program communication and implementation, creating confusion for borrowers about their loan terms and conditions. Although the study provides several policy-relevant findings, it relies on descriptive analysis complemented by qualitative insights. Due to data limitations, it was not possible to construct a valid counterfactual group 4 to assess the program’s impacts or to demonstrate causal impacts of the program.3 Notably, we cannot detect what would have happened to MSMEs’ access to finance in absence of the KUR program. Moreover, while we see that borrowers access KUR loans repeatedly, we cannot determine whether repeated use of KUR loans is due to their favorable conditions or whether firms struggle to find unsubsidized commercial lending that can meet their needs. Despite these limitations, the study illuminates important trends based on a large, comprehensive administrative database of all KUR borrowers between 2015 and 2020 and a nationally representative survey of KUR borrowers. Following the initial release of findings of this study in 2022, several key policy changes were introduced to the KUR program. In January 2023, a new regulation introduced a pathway to graduation from KUR loans, by reducing subsidies and increasing interest rates for each repeat loan from a single borrower.4 The regulation also introduced penalties for participating banks that contravened program guidelines by requesting collateral from borrowers when it was not required and introduced other policy reforms to optimize the impact of the program and its outreach to first-time borrowers. Both lenders and firms now have greater incentives to access KUR as a time-bound subsidy, that can help unbanked firms climb the ladder of financial inclusion. The Government of Indonesia’s administrative data from the KUR program for the year 2023 shows that 72% of KUR borrowers were first-time borrowers, 49% were female, and 53% were graduating to larger KUR loans or commercial lending.5 Our paper confirms some findings from existing literature on KUR while nuancing others. It also complements the literature by using large-scale, nationally representative data on KUR borrowers. Existing studies of the KUR program rely on internal program monitoring data, interviews with government officials, financial institutions, or small samples of KUR borrowers. Our findings on KUR’s large collateral requirements echo those in other studies (De Braw et al. 2020; Martokoesoemo et al., 2019; ILO 2019; OECD 2018; Dea 2019). We also confirm that KUR’s de facto collateral requirements pose a constraint for women’s access to the program (Martokoesoemo et al., 2019; Dea 2019; Farida 2015). Existing literature has cast doubt on KUR’s ability to reach first-time borrowers (Martokoesoemo et. Al. 2019; ILO 2019; OECD 2018). Nevertheless, using large-scale, nationally representative quantitative data, we find that the challenge is more nuanced. First-time borrowers are able to access KUR; however, the main challenge lies in the incentives to graduate from KUR to commercial lending. 3 The initial study design used genetic matching to identify potential borrowers in the program data who did not receive KUR. However, interviews revealed that almost all borrowers in the database had received KUR at some point, suggesting either a very low rejection rate or that the monitoring and evaluation data does not accurately capture individuals who apply for KUR loans and do not receive them. 4 Regulation Number 1 of 2023 of the Coordinating Minister of Economic Affairs, Concerning Guidelines for the Implementation of KUR. January 25, 2023. 5 KUR Administrative Data, provided by Coordinating Ministry of Economic Affairs (CMEA), March 2024. 5 Beyond the context of Indonesia, our paper contributes to the literature that studies government support schemes to small businesses and government driven lending. A large body of empirical work examines at an aggregate level whether public lending schemes promote financial development and growth (King and Levine, 1993; Demirguc-Kunt and Maksimovic, 1998; Rajan and Zingales, 1998). A case study approach is used to examine the practical ways in which large-scale government programs play out, exploring both their intended and unintended impacts, and focusing on design lessons for similar schemes. Rather than exploring whether public MSME finance schemes have an impact, the paper focuses on how these schemes can be strengthened, acknowledging that targeted efforts to unlock financing from MSMEs are and will remain a cornerstone of public policy in emerging markets (McKenzie, 2023). Our paper brings nuance to the existing evidence about the extent to which subsidized loans reach firms that need them the most and provide additionality. First, unlike findings from Brazil that earmarked credit favors larger, established firms (Ornelas et al. 2019), our study shows that KUR successfully targets microbusinesses that are first-time borrowers. However, consistent with the notion that less risky borrowers with collateral are better able to access subsidized loans, we find that KUR also reflects this trend. Second, while many papers focus on the extent to which subsidized credit reaches firms that are credit constrained (Banerjee and Duflo 2014; Ornelas et al 2019; Zia 2008), we investigate whether subsidized loans can pave the way to commercial credit, but we find no evidence to support this, suggesting that microenterprises remain relatively more costly to serve than other populations, driving lenders upmarket in absence of subsidies (Klein and Ogden 2023). Third, we uncover the consumer protection challenges posed by decentralized implementation, a topic that has not been emphasized in the literature. Finally, unlike other programs that use government capital for loans (Ornelas et al 2019 for Brazil and Zia 2008 for Pakistan), KUR relies on the bank’s own reserves without expanding their lending capacity. As such, our findings that KUR enables banks to reach new borrowers under these conditions are particularly notable. The rest of the paper proceeds as follows. The next section describes the context of MSME finance in Indonesia, the KUR program, and the data. Section 3 highlights the main analytical results. Section 4 discusses the policy implications of the findings and concludes. 6 2. Context and Data 2.1 MSME Finance in Indonesia The financial sector in Indonesia is shallower than in other low- and middle-income countries or other countries in the region (Figure 1). Domestic credit to the private sector has averaged 37% of GDP in Indonesia over the past decade, compared to an average of 99% in low- and middle-income countries and 144% in developing East Asia and the Pacific. While domestic credit availability has been steadily increasing elsewhere in the region, in Indonesia the availability of domestic credit has remained stagnant. Figure 1: Indonesia’s financial sector is shallow compared to peers Domestic credit Domestic creditto toprivate sectoras private sector as aa share share ofof GDPGDP 200 Domestic credit to private sector (% of GDP) 175 150 125 100 75 50 25 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 East Asia & Pacific (excluding high income) Low & middle income Indonesia Source: World Development Indicators Over the past decade, MSME clients have comprised an increasing share of banks’ portfolios in terms of number of lending accounts but not in terms of the share of total lending. Lending accounts to MSMEs have increased from 7 million in January 2012 to over 40 million by the end of 2022 (Figure 2). Aligned with an increase in lending accounts, the total volume of credit to MSMEs has also tripled over this time period from approximately US$30 billion in 2012 to US$90 billion at the end of 2022 (Figure 3). However, although the share of lending accounts to MSMEs has increased from approximately one-fifth to one-third of banks’ total lending accounts, the overall share of lending volumes to MSMEs has remained relatively stagnant at about 19% of banks’ total lending volumes. 7 Figure 2: MSME clients make up an increasing Figure 3: Lending volumes have increased over share of banks’ loan accounts time but not the share of lending to MSME Number Number of lending oflending andand accounts accounts share share Lending volumes and share of lending accounts issued to MSME Lending volumes and share of lending to MSME of lending accounts issued to MSME of lending to MSME 7 Mio 30% Total lending (in IDR Billion) % of Loans for MSME 140 45% Number of lending accounts (in millions) % of lending accounts for MSME 120 40% 6 Mio 25% 35% 100 5 Mio 30% 20% 80 25% 4 Mio 15% 60 20% 3 Mio 15% 10% 40 2 Mio 10% 20 1 Mio 5% 5% 0 0% 0 0% Jul-17 Jan-12 Jun-18 May-19 Mar-21 Dec-12 Feb-22 Oct-14 Apr-20 Sep-15 Nov-13 Aug-16 Aug-16 Sep-15 Feb-22 Apr-20 Nov-13 Dec-12 Oct-14 Jul-17 Mar-21 Jan-12 Jun-18 May-19 % of Accounts are for MSMEs % of Loans for MSME Number of Account of Total Credit Total Loans of Commercial Banks Number of Account of Total MSMEs Credit Total MSME Loans of Commercial Banks Source: BI Source: OJK Improving access to finance for MSMEs is part of the target outcomes under the economic transformation pillar of Indonesia’s Medium-term Development Plan 2020-2024. Indonesia’s central bank, Bank Indonesia, mandates banks to lend to the MSME sector through a minimum “Macroprudential Inclusive Financing Ratio (RPIM)” which stipulates that, as of June 2024, 30% of bank lending portfolios should be dedicated to MSMEs. In addition to regulations, the Government of Indonesia has developed programs to support financial inclusion of MSMEs, which receive approximately 90% of the budget allocated to innovation and business support instruments (World Bank 2021). Most of this funding is channeled through the KUR program. 2.2 Description of the KUR program In 2007, the Government of Indonesia launched the “People’s Business Loan” (Kredit Usaha Rakyat, KUR) program as a flagship public program to enhance MSMEs’ access to finance. KUR subsidizes credit for working capital and/or investments for individuals, business entities, and/or business groups that are productive and feasible but do not have sufficient collateral (Permenko Regulation No. 8 of 2019). Governed by an inter-ministerial Policy Committee led by the Coordinating Ministry for Economic Affairs (CMEA), the program began as a partial credit guarantee aimed to enhance bank lending to MSMEs by addressing the higher perceived credit risks associated with lending to small and previously unbanked firms. In 2015, the program was re-designed with the addition of interest rate subsidies to make credit more affordable for 8 MSMEs. Table 1 provides an overview of key policy changes in KUR’s implementation. Table 1: Overview of key changes to KUR’s implementation Area 2007 2015 2023 Overview of Partial credit Interest rate subsidy and Interest rate subsidy and partial program guarantee partial credit guarantee credit guarantee Types of KUR Micro KUR; Micro KUR; Ultra-micro KUR;* offered Small KUR Small KUR; Micro KUR; Migrant KUR Small KUR; Migrant KUR; Special KUR* Credit Credit guarantee Banks receive fee as part of Banks receive fee as part of subsidy guarantee fee company (CGC) subsidy and pay to CGC and pay to CGC receives fee directly from government Amount of Set by Government: Based on agreement Based on agreement between CGC guarantee fee 1.5% in 2007 between CGC and KUR and KUR Distributor 3.25% in 2010 Distributor Interest rate Declined from Declined from Ultra-micro KUR: 3% 24% for micro KUR 12% for all KUR windows in 1st Micro/Small KUR: 6% and 16% for small 2015 to 2nd Micro/Small KUR: 7% KUR in 2007 to 6% for all KUR windows in 3rd Micro/Small KUR: 8%; 22% for micro KUR 2020 4th Micro/Small KUR: 9% and 14% for small Migrant KUR:6% KUR in 2010 Special KUR: 6% Restrictions Debtor has not No restrictions on previous Debtor has not received on previous received loans lending commercial working capital credit lending from any banks or investment financing, except for or government consumption credit, ultra micro programs. credit, or fintech loans Limits to KUR is restricted to Repeated Micro KUR Micro KUR limited to 4 times in repeat KUR first-time borrowers. loans allowed with a total agricultural, plantation, fisheries, borrowing accumulated ceiling of 75 livestock production sectors and million IDR; 2 times in other sectors with a Repeated Small KUR lifetime maximum of 100 million loans allowed with a total IDR; accumulated ceiling of 500 Repeated Small KUR loans allowed million IDR with a total accumulated ceiling of 500 million IDR. * Special KUR was introduced in 2017, and Ultra-micro KUR was introduced in 2020 Since its inception, KUR has grown into one of the world’s largest public support programs for MSMEs. Since 2007, nearly IDR 1,565 trillion (US$ 100 billion) have been disbursed through more than 50 million 9 loans. Currently, there are three stated objectives of KUR: (i) increasing and expanding access to finance for productive businesses; (ii) increasing the capacity and competitiveness of MSMEs; and (iii) encouraging economic growth and development (Regulation No. 1 2022, Section 2). In its current form, KUR provides public financing to subsidize interest rates on loans for MSMEs and fund a partial credit guarantee for these loans. The government provides interest subsidies directly to participating banks, allowing them to lend to MSMEs at capped interest rates.6 The interest rate subsidy also covers a guarantee fee which banks pay to selected Credit Guarantee Companies (CGCs), but the loan capital comes from the banks themselves. Each year, the banks and financial institutions which participate in KUR (‘distributors’) submit an estimate of the amount of KUR loans they expect to be able to disburse in the coming year. Based on estimates and past performance, the government formulates annual lending quotas and performance targets for each bank, as well as an aggregate annual target for the KUR program. All participating banks must offer loans according to conditions outlined in the KUR Regulations in order to receive the subsidy. All banks participating in KUR report into a central Credit Program Information System (SIKP), which registers and tracks each KUR borrower. To become a KUR distributor, financial institutions or cooperatives must meet certain criteria: (i) being healthy and having good performance; (ii) cooperating with a KUR guarantor company in distributing KUR, and (iii) having access to the online KUR data system (SIKP). Financial institutions that meet these criteria can apply to become a distributor by submitting documentation and entering into a financing cooperation agreement (Permenko No 8 of 2019, Article 6). In 2022, there were 44 KUR distributors. However, the majority of KUR loans are disbursed by one distributor, Bank Rakyat Indonesia (BRI), which distributed approximately 70 percent of KUR loans in 2022.7 The fiscal cost of the KUR subsidy program amounted to approximately 0.1% of GDP in 2018-21, rising to 0.2% in 2022-23. Total spending on KUR in 2023 amounted to approximately US$2.76 billion, with the majority of the expenditure program derived from the cost of the interest subsidies and a small portion derived from the cost of the credit guarantees. The total fiscal cost of the KUR program since the introduction of interest rate subsidies in 2015 totals approximately US$12 billion. Earlier estimates found the cost of interest subsidies to be approximately ten times the cost of the credit guarantee.8 The program claims approximately 90% 6 For Islamic financial institutions the proper term is margin subsidy. Throughout this paper, for simplicity, “interest rate subsidy” is used to refer to the interest rate subsidy given to commercial banks and to the equivalent “margin subsidy” given to Islamic financial institutions. 7 Data drawn from the SIKP as reported on the CMEA’s website: http://sikp.kemenkeu.go.id/dashboard. 8 Indonesia Economic Quarterly, March 2017. “The New Kredit Usaha Rakyat (KUR) Program.” 10 of the public budget allocated to direct expenditures on innovation and business support programs in Indonesia (World Bank 2021).9 Outstanding KUR loans as of March 2024 total approximately US$36 billion (IDR 545 trillion), roughly 40% of the total US$90 billion (IDR 1,362 trillion) of outstanding MSME loans in the country. During the period of the study, the cumulative non-performing loan ratio for KUR loans was 1.2%, with a slightly higher NPL ratio for micro-KUR (1.66%) than small KUR (0.75%).10 The NPL for KUR loans compares favorably to NPL of broader MSME loans in the country: during the same time period, the average monthly NPL rate was 4.22% for MSME loans in Indonesia.11 There are five types of KUR loans; however, this study focuses on the two most common categories that existed between 2015 and 2020, i.e. the period of our data: micro KUR and small KUR loans.12 Table 2 compares the conditions of micro and small KUR loans for the year 2019.13 Table 2: Comparison of terms and conditions for micro KUR and small KUR loans for the year 2019 Term/condition Micro KUR Small KUR Interest rate 6% 6% (paid by borrowers) Interest rate subsidy 10.5% 5.5% (paid to banks) Term/condition Micro KUR Small KUR Loan amount Up to 50 million IDR 50 – 500 million IDR (US$ 3,500) (US$ 3,500 to US$ 35,000) Loan duration  Max. 3 years (working capital)  Max. 4 years (working capital)  Max. 5 years (investment)  Max. 5 years (investment) Lifetime limit  No limit for the production sector 14  500 million IDR (US$ 35,000)  200 million IDR (US$ 14,000) for the for both production and non- non-production sector production sectors Collateral requirements  Principal collateral is the business or  Principal collateral is the object financed by KUR business or object financed by  Additional collateral is not required KUR and without contract or agreement15  Additional collateral in accordance with the assessment of the distributor 9 Tax incentives play an important role in Indonesia’s policy mix to support businesses; however, the value of these incentives is not included in this calculation due to data constraints. 10 Data from the Indonesia Coordinating Ministry of Economic Affairs (CMEA), available here. 11 Data from Indonesia Financial Services Authority (OJK). 12 The five types of KUR loans include: (i) micro KUR loans; (ii) small KUR loans; (iii) KUR loans for migrant workers (TKI); (iv) ultra-micro KUR loans; (v) special KUR. Ultra-micro KUR loans were introduced in August 2020, after the period of the data used in this study. 13 Some terms and conditions, including the interest rate and the subsidy amounts, have changed over time. The conditions that were in place in 2019 reflect the conditions that were most common during the study period of this report. 14 Production Sector in the regulations refers to “the economic sector that produces goods and/or services”. 15 Without contract or agreement means that it is not notarized and cannot be confiscated in the case of default. 11 2.3 Description of the KUR borrowers Most KUR borrowers are micro entrepreneurs: 97 percent have fewer than five full-time salaried workers and only 44 percent have any paid workers. They operate in a diverse range of sectors of activity, with the most common ones including trade (42 percent of borrowers), crop and animal production (22 percent of borrowers), and manufacturing (14 percent of borrowers). On average, businesses have been in operation for seven years; however, there is considerable variation, with some firms that were established shortly before receiving their loan and others that have been in operation for over 30 years. According to novel data collection for this study described in Section 2.4, the average firm had profits of 3.6 million IDR (US$ 240) in the month preceding the survey (December 2020), and the median firm had profits of 1.5 million IDR (US$ 100). Although most KUR borrowers have some form of formal documentation for their businesses, most have limited financial management practices, and many do not use bank accounts for their businesses. Eighty- five (85) percent of KUR borrowers have some form of formal documentation for their businesses. Most commonly, businesses only have the Business Certificate (SKU) or the Certificate of Business Domicile (SKDU), which is issued by local authorities and recognizes that the business exists and operates in the locality. Nevertheless, fewer than half of KUR borrowers separate their business and personal finances, and only 28 percent have any written business records. Less than a quarter of firms record every purchase, can easily determine how much cash they have on hand, or can use their records to prove to a loan officer that they would have sufficient funds to repay a loan. Despite limited financial record-keeping, three-quarters of borrowers can calculate the costs of the products and services they offer. KUR borrowers range in age from the late teens to the elderly, with an average entrepreneur age of 40 years old. Thirty-seven percent of borrowers are female, and 63 percent are male. Borrowers have education levels roughly on par with the working age population in Indonesia; however, individuals with no formal education are under-represented and those with upper-secondary levels of education are over-represented among KUR borrowers. Most do not have formal business training and have instead developed business skills through learning by doing or through the support of friends and family with businesses. A more detailed description of KUR borrowers is provided at Appendix A. 2.4 Data We use four types of data to explore how KUR works from the perspective of MSME borrowers who received KUR between 2015 and 2020. First, we use administrative data from the KUR program database to explore 12 basic characteristics of borrowers. This dataset covers all KUR loans issued between December 2015 and March 2020 and includes information on over 8 million debtors. Second, we use a unique, quantitative survey dataset that is the largest and only nationally representative data on KUR borrowers to date. Third, qualitative interviews with KUR borrowers enable a deeper exploration of KUR borrowers’ perspectives and experiences. Finally, qualitative interviews with individuals who did not receive KUR reveal some challenges related to implementation and access to KUR. The unique quantitative survey was collected for this study and includes a nationally representative sample of 1,402 KUR borrowers. To ensure a representative sample and enable subgroup analysis, we used weighted stratified sampling to select firms to interview from the national program database.16 Strata including less than 1% of KUR beneficiaries were oversampled to ensure that each subgroup of interest would have sufficient representation in the sample to allow precise estimates at the subgroup level, and all analysis in this report incorporates design weights to account for sampling design. Due to the COVID-19 pandemic, all interviews were carried out over the phone in January and February 2021. All firms were asked modules on basic business information, business practices, workers, revenue, financial history prior to receiving KUR for the first time, and financial history after receiving KUR for the first time. In addition, firms were asked one of two of the following modules: experiences with the KUR program or impact of COVID-19 on the business.17 To better understand how KUR borrowers with different characteristics perceived and experienced the KUR program, we conducted qualitative interviews with a separate sample of 100 KUR borrowers, drawn from the program database. The qualitative sample was not drawn to be representative and includes KUR borrowers with a variety of different characteristics. These qualitative interviews covered the following topics: perception and information received about KUR, KUR loan application process, use and benefits of KUR loans, business productivity and access to financial services, alternative financing sources available, the impact of COVID-19 on the business, and suggestions for improving the KUR program. Due to the COVID-19 pandemic, the qualitative interviews were conducted using phone interviews. To avoid challenges of survey length and respondent fatigue, the depth of probing on each topic varied by respondent. Qualitative interviews were also conducted with 100 individuals who did not receive KUR but who had been identified as potential debtors in the program’s database.18 Among them, some firms had applied 16 Strata included gender of KUR recipient, size of KUR loan, financial institution that issued the KUR loan, and geographic region. 17 These modules were not asked of all respondents due to concerns about survey length and respondent fatigue. The modules to be asked were randomly assigned and balanced across sampling strata to ensure all modules included nationally representative information. 18 97 percent of individuals in the SIKP who were successfully contacted for quantitative interviews with KUR borrowers had received KUR, so the sampling for these interviews was a convenience sample of entrepreneurs who were screened but not eligible for the surveys of borrowers. 13 for KUR but had been rejected, others had applied for KUR but had withdrawn their application before receiving KUR, and others did not apply for KUR. These interviews covered the following topics: perception and information received about KUR, KUR loan application process, reasons and notification of KUR loan rejection or withdrawal, alternative financing sources available, adjustments to the business plan or scale after rejection or withdrawal, the impact of COVID-19 on the business, and suggestions for improving the KUR program. As with the qualitative data collection of KUR borrowers, interviews were conducted over the phone, and interviewers had the flexibility for which topics to explore in more depth with each respondent. 3. Results 3.1 KUR’s subsidized credit reaches many first-time borrowers, helping them enter the formal financial sector and invest in their firms KUR primarily targets borrowers who are excluded from the formal financial system: most KUR borrowers (61%) had never received any type of loan, including loans from friends or family, prior to accessing KUR. These borrowers typically relied on personal savings and reinvested profits to meet both their investment and working capital needs before accessing KUR (Figure 4). In fact, only a small percentage of borrowers relied on formal financing arrangements, such as commercial banks, microfinance institutions, and input suppliers. KUR also finances business plans that may not be implemented at the same scale otherwise. Qualitative interviews with rejected KUR applicants reveal that less than half were able to find alternative financing for their projects, and those who did typically reduced the scope of their investment. Figure 4: Most KUR borrowers relied on savings to finance business investments and working capital prior to receiving KUR Main source Main ofof source capital capital prior to prior to receiving receiving KUR KUR Personal savings 38% 35% Savings from business/reinvested profits 21% 37% Friends or family other than spouse (loans or gifts) 8% 6% Spouse (loans or gifts) 6% 7% Commercial banks 4% 4% Microfinance inst itutions 3% 3% Input supplier 1% 4% 0% 10% 20% 30% 40% % firms citing as main source Main source of capital investments Main source of working capital Source: Author calculations using survey of KUR borrowers Although most KUR borrowers were previously unbanked, there are some borrowers who previously accessed unsubsidized commercial lending and switched to KUR. Between business creation and receipt of 14 their first KUR loan, 11% of KUR borrowers had received at least one commercial loan from a bank. As shown in Table 1, eligibility criteria for KUR loans issued between 2015 and 2023 did not restrict access to firms without a lending history. In the five years before receiving KUR, 9 percent of borrowers received loans from a commercial bank, 10 percent from microfinance institutions, and 2 percent from regional development banks (Figure 5). After receiving KUR, the usage of these funding sources significantly decreased, indicating that KUR may be displacing other sources of formal lending for some borrowers. The finding of a possible crowding-out effect for some borrowers is supported by qualitative interviews in which some firms reported attempting to switch their non-KUR loans to KUR loans due to their lower interest rates. Figure 5: After receiving KUR, fewer firms use alternative sources of formal and informal finance for their businesses Formal Formal Financing Financing UsedBefore Used and After Before and AfterReceiving ReceivingKUR KUR 12% 10% 10% 9% 8% 6% 4% 3% 2% 2% 2% 0% 0% 0% 0% Non-KUR Loans Micro Finance Institutions Digital Lender Regional Development Banks Prior to receiving KUR for the first time After receiving KUR for the first time Source: Author calculations using survey of KUR borrowers While some studies suggest that microbusinesses may divert business loans to household consumption (Casano et al 2013), KUR loans are primarily used for business purposes. The primary use of KUR loans is working capital for the primary business (55.3% of loan value), followed by investments in physical capital for the primary business (27.2% of loan value, Table 3). Qualitative interviews shed light on the various ways in which these loans are utilized for business expansion and improvement. Many firms choose to expand their operations by hiring additional workers, opening a new branch location, or increasing their stock.19 Others focus on enhancing the attractiveness of their businesses by moving to a better business location or renovating the store. Businesses also aim to improve productivity by investing in new production equipment or additional productivity-enhancing inputs.20 19 For instance, in the agriculture sector, firms utilize KUR loans to increase their purchases of fertilizer to maximize their yields and increase the area of agricultural land cultivated by either renting or purchasing new land or by purchasing inputs to expand production area of idle land. 20 While the majority of KUR loans are used for business purposes, funds are also fungible, and not all of the credit is allocated for productive purposes. Among the study population, 10% of the loan value was allocated towards non-business expenses, and 24% of KUR borrowers used at least part of their KUR loans for non-business expenses. 15 Beyond helping many entrepreneurs get access to credit for the first time, qualitative evidence suggests that KUR can also help entrepreneurs position themselves to receive formal, commercial loans in the future. Borrowers report that KUR motivated them to keep better business records and to formalize their businesses. It also helped them develop a good relationship with the bank and increased their awareness of the conditions and application process for formal loans. Entrepreneurs also felt that receiving KUR increased their self-confidence and boosted the credibility of their businesses. As a result, most KUR borrowers receive larger loan sizes with each successive KUR loan, indicating an increased level of credibility and trust from lenders.21 In this way, subsidized credit has the potential to support first time borrowers’ entry into market-based financing. Table 3: Share of KUR loans used for different purposes Purpose Average share of KUR loans used Working capital 55.3% Capital investment 27.2% Other business expense 0.2% Expenses for respondent’s other business 4.6% Expenses for business not owned by respondent 1.2% Non-business purposes 10.4% Have not yet used 1.2% Note: Borrowers may use KUR loans for more than one purpose, and they were asked the percentage of their loans used for each purpose. The table notes the average across all firms, regardless of whether they spent any of their funds on the purpose. 3.2 KUR borrowers rely on repeated subsidized loans, rather than graduating to unsubsidized commercial lending Although KUR helps businesses become more bankable by encouraging formalization, record-keeping, and developing relations and familiarity with banks, borrowers do not typically transition to unsubsidized lending. Indeed, only 3 percent of firms accessing KUR between 2015 and 2020 graduated to other forms of commercial lending during this period (Figure 6). Available data do not enable us to assess the extent to which firms who would like to transition to unsubsidized commercial loans are able to access them. However, there is evidence that prior to recent policy changes borrowers did not have incentives to make this transition even if they could. Between 2015 and 2023, borrowers were permitted to receive KUR loans multiple times. In fact, approximately half of the firms that received KUR loans between December 2015 and March 2020 received 21 Seventy percent of repeat borrowers received a larger loan amount for their second loan, and on average the second loan is 80% larger than the first. Similarly, the third KUR loan is larger than the second for 63% of KUR borrowers with three or more loans, and on average the third KUR loan is 62% larger than the second one. 16 them multiple times during this time period (Figure 7). This likely underestimates the true likelihood of turning to repeated KUR loans, as borrowers who received loans for the first time toward the end of the study period would not have time to repay the first loan and take another loan during the study period.22 Indeed, almost all borrowers who received KUR for the first time in 2015 had received multiple KUR loans during the 5-year study period (Figure 8). Regulations in effect at the time of the survey would allow a firm to continuously fund business using subsidized KUR loans for up to 48 years, assuming average loan sizes and repayment according to the loan periods allowed in the regulations.23 Figure 6: Three percent of KUR borrowers graduate to commercial lending Do FirmsDoWho Receive Firms Who KUR Receive KURGraduate to Other Graduate to Other Types Types of Loans? of Loans? 81.5 % 80% 60% 40% 20% 10.0% 8.5% 6.5% 2.9% 0.5% 0% Only had Only had Graduated Graduated Graduated Graduated KUR micro KUR small micro KUR micro KUR small KUR (any type) to small KUR to non-KUR to non-KUR loans loans Source: Author calculations using survey of KUR borrowers Figure 7: Approximately half of KUR borrowers have received multiple KUR loans Number of KUR loans borrowers received Number of KUR loans borrowers received 60% 50.8% 50% 40% 30.7% 30% 20% 12.5 % 10% 6.0 % 0% 1 2 3 4 or more Source: Author calculations using survey of KUR borrowers 22 The most common type of KUR loan, a micro KUR loan for working capital, can have a tenor of up to 3 years. 23 According to Regulation 8 of 2019, micro KUR loans in the non-production sector were limited to a lifetime maximum of 200 million IDR and small KUR loans were subject to a lifetime maximum of 500 million IDR. With an average micro KUR loan size of 18.5 million IDR and an average small KUR loan of 74.6 million IDR, a firm in the non-production sector could receive an average-sized KUR loan 16 times – 10 micro KUR loans and 6 small KUR loans. The calculations assume the maximum term allowed for working capital loans—the most common use of KUR loans—which is 3 years. 17 Figure 8: Almost all firms who received KUR in 2015 received multiple KUR loans within a 5-year period all firms Figure 8: Almost Number ofwho KUR received in 20152015 KURbetween loans received received multiple and 2020 KUR by year loans of first KURwithin loan a 5-year period Number of KUR Number of Loans Received KUR loans receivedBetween 2015 between 2015 and and 2020 2020 by Year by year of of first First KUR KUR Loan loan Number of KUR Loans Received Between 2015 and 2020 by Year of First KUR Loan 100% 0.75 100% 2.86 1.95 11.66 2.86 8.65 0.75 1.95 10.47 90% 11.66 8.65 11.13 10.47 90% 8.20 11.13 21.17 21.84 8.20 21.17 80% 21.84 80% 70% 35.15 70% 42.09 35.15 42.09 60% 40.95 60% 33.27 40.95 33.27 50% 50% 89.53 40% 89.53 40% 76.88 76.88 30% 30% 52.98 44.38 52.98 20% 44.38 42.03 42.2 20% 42.03 42.2 10% 10% 0% 1.88 0% 1.88 2015 2016 2017 2018 2019 2020 2015 2016 2017 2018 2019 2020 1 2 3 4 or more 1 2 3 4 or more Soorce: Author calculations using survey or KUR borrowers The low interest rates make KUR an attractive option for borrowers, leading to a lack of incentives for businesses to graduate to unsubsidized lending. Indeed, borrowers pay approximately half the interest rate of an unsubsidized commercial loan of a similar size.24 The declining interest rates of KUR loans over time even led to some borrowers applying for new KUR loans at different banks to reimburse previous KUR loans and lower their overall interest payments. In some cases, KUR distributors encouraged borrowers to take repeated KUR loans after repaying their first loans and made promotional visits to attract repeat borrowers. Graduation from the KUR credit subsidies, an often-stated goal of the program, has been inhibited by the ability to access repeat KUR loans and a lack of incentives to use more costly unsubsidized loans. Acknowledging the challenges with repeat borrowing, in 2023 the Government of Indonesia amended the KUR regulations to introduce a graduation policy.25 This policy limits the number of times borrowers in the productive sectors (agriculture, plantation, fishery, and animal husbandry) can receive micro KUR to four, up to a maximum of IDR 100 million per planting season or production cycle, and limits the number of micro KUR loans to 2 in other sectors.26 The policy also increases the interest rate by one percentage point for each successive micro and small KUR loan (i.e. from 6 percent for the first KUR loan to 9 percent for the fourth KUR loan), helping borrowers prepare to graduate to non-subsidized loans. As they are implemented, 24 The prime lending rates for micro credit loans offered by two large banks were 11.25 and 14 percent for Bank Mandiri and BRI, respectively, in October 2021. In contrast, KUR interest rates stood 6 percent in 2020 (Permenko No. 8 of 2015, Article 13 paragraph 2, Article 17 paragraph 2; Permenko No. 15 of 2020, Article 8, paragraph 3, letter c). 25 Regulation Number 1 of 2023 of the Coordinating Minister of Economic Affairs, Concerning Guidelines for the Implementation of KUR. January 25, 2023. 26 There are no strict limits on the number of times borrowers can receive small KUR; however, the total amount of cumulative lending cannot exceed 500 million IDR. 18 these reforms are likely to hold important lessons for the ways in which public financing schemes can use targeted subsidies to help borrowers climb the ladder of financial inclusion and transition to commercial lending over time. 3.3 Stronger regulations, guidelines, and oversight are needed to ensure inclusiveness and consumer protection. KUR is implemented by financial institutions, which capitalizes on expertise of the financial sector to screen potential borrowers for creditworthiness and promotes engagement of KUR borrowers with formal financial institutions. While implementation through financial institutions has several benefits, our study reveals some unintended challenges of a decentralized implementation model and suggests a need for enhanced regulations, guidelines, and oversight of the program. First, despite targeting borrowers without sufficient collateral, borrowers were required by distributors to provide substantial collateral for KUR loans, limiting the program’s stated inclusion goals. Second, loan terms and conditions vary by KUR distributor leading to confusion among borrowers. Clear communication is essential to ensure borrowers understand loan terms and conditions and their rights and obligations. Finally, improvements to data collection on the program can enable improved program oversight and evaluation. Program regulations define KUR as credit or financing for individuals or business entities that do not have sufficient collateral.27 The regulations enable borrowers with limited collateral to access KUR loans through three provisions. First, the interest rate subsidy includes a subsidy to cover the fees for a partial credit guarantee provided through KUR guarantors. This partial credit guarantee protects KUR distributors from the risk of default, making it easier to grant loans to individuals or business entities with limited physical assets to offer as collateral. Second, the basic collateral required for micro KUR loans and small KUR loans is the business or object to be financed by KUR. Therefore, all KUR borrowers possess the basic collateral required for KUR loans. Finally, while additional collateral may be required for small KUR loans, it is not required for micro KUR loans.28 In practice, however, both micro KUR and small KUR borrowers who received loans during the study period were required to present substantial additional collateral to receive KUR. In contrast to the regulations that state additional collateral is not required for micro KUR loans, 96% of borrowers were required to present collateral for micro KUR loans received between 2015 and 2020 (Figure 9). On average, KUR borrowers 27 Permenko No 8 of 2020 defines KUR as “credit or financing for working capital and/or investment to productive and feasible individual debtors, business entities, and/or business groups that do not have additional collateral or insufficient additional collateral.” 28 More recent changes to regulations in January 2023 have emphasized that additional collateral is not required for micro KUR loans, and sanctions are imposed for KUR distributors that do not adhere to this policy. 19 provided collateral worth 4.4 times the value of their loan, much more than the required 1.9 collateral value to loan ratio required for other bank loans of a similar size (Figure 10). This high ratio can in part be explained by the fact that firms were required to submit land titles for loans as small as IDR 15 million (US$ 1000). Nevertheless, KUR’s large collateral requirements are not just a function of a few extreme cases: for the median borrower, collateral requirements for KUR are still larger than for other loans of a similar size (1.7 compared to 1.4). Figure 9: Almost all KUR borrowers were required to submit collateral to receive KUR %% firms who firms were who were required required to submit to submit collateral, collateral, by type by type of loan of loan 95.7% 100% 100% 92.5% 80% 60% 40% 20% 0% Micro KUR Small KUR Non-KUR Commercial Loan Source: Author calculations using survey of KUR borrowers Figure 10: KUR borrowers present collateral that is valued at multiple times the amount of the loan Ratio of Ratio collateral of value collateral value toto loan loan amount amount 5 4.5 4.4 4 3.5 3 2.5 1.9 2 1.7 1.4 1.5 1 0.5 0 KUR Other loans KUR Other loans Median Average Source: For KUR loans, survey of KUR borrowers; For other loans, 2023 World Bank Enterprise Surveys This requirement for substantial collateral limited the inclusivity of KUR loans and in particular hindered women’s access, since female entrepreneurs in Indonesia are less likely than their male counterparts to own buildings or land that can serve as collateral (World Bank 2016). Qualitative interviews revealed that lack of collateral was a principal challenge women faced in accessing KUR,29 and quantitative data show 29 Another barrier to women’s access to KUR may be the assessment of their husband’s credit history. Some women report their KUR application was denied because their husbands had outstanding loans or had previously defaulted on a loan. 20 that female entrepreneurs, in particular lower educated ones and those running smaller businesses, are under-represented among KUR borrowers. If female entrepreneurs had similar access to KUR as male entrepreneurs, one would expect to see similar trends among KUR borrowers as in the population of Indonesian microentrepreneurs. However, female entrepreneurs are less prevalent among KUR borrowers than in the population of Indonesian microentrepreneurs (Figure 11). Whereas on average female entrepreneurs in Indonesia have lower education levels and smaller businesses than male entrepreneurs (World Bank 2022), female KUR debtors have higher education levels and seek larger loans than their male counterparts (Figure 12). Figure 11: Female entrepreneurs are under-represented among KUR borrowers Share of KUR Share borrowers of KUR and borrowers andmicroentrepreneurs in Indonesia microentrepreneurs in Indonesia who who are are female female 49.4% 40% 36.7% 20% 0% KUR borrowers Microentrepreneurs Source: KUR borrowers: SIKP database; Microenterprises: World Bank 2022 Figure 12: Female KUR borrowers have higher education levels and larger loans than male KUR borrowers Characteristics Characteristics of of male and and maleKUR female female KUR borrowers borrowers 11.7 Type of Small KUR 8.3 KUR 88.3 Micro KUR 91.7 Diploma/Bachelor 5.6 3.7 Education Senior High 30.1 26.7 14.3 Junior High 14.2 46.3 Elementary 51.5 18.3 Marital Not married Status 12.4 81.7 Married 87.6 0 10 20 30 40 50 60 70 80 90 100 % of KUR borrowers Women Men Source: Author calculations using SIKP database Note: Gender differences are statistically significant at the 1% level 21 The reliance of KUR distributors on collateral also suggests the credit guarantee mechanism may not be working as intended. Credit guarantee schemes are designed to reduce the amount of risk that banks assume when issuing a loan, thus reducing the need for collateral that also serves the same purpose (Levitsky 1997). As such, collateral requirements for KUR loans suggest that banks do not fully trust the guarantee mechanism to mitigate their risk. Global evidence suggests that when banks do not have complete trust in credit guarantee companies or find the delays or procedures required to receive a claim too burdensome, partial credit guarantees can be ineffective (Levitsky 1997). The regulations in place during the study period did not include a mechanism to track or enforce the provision of micro KUR loans without additional collateral. Moreover, banks had incentives to shift collateralized commercial loans to KUR loans in order to meet the government’s ambitious targets for KUR distribution (World Bank 2017). The government issued a reform in January 2023 that aims to address these concerns. The revised regulations emphasize that additional collateral is not required for micro KUR loans and impose sanctions on KUR distributors that still require collateral.30 The reform aims to ensure that micro KUR loans are provided without additional collateral and to hold distributors accountable for non-compliance. However, a deeper exploration of challenges with the partial credit guarantee and steps to address them are warranted to ensure that financial institutions can comply with these requirements without unreasonable exposure to risk. Loan terms and conditions vary across borrowers and are not well-understood by KUR borrowers, raising consumer protection concerns. Almost half of KUR borrowers (48%) paid administrative fees for loan processing, including notary fees, stamp fees, and insurance premiums, and the amount paid varied considerably among borrowers and was a source of confusion for them. Among those who had to pay fees, the median firm paid 40,000 IDR (US$2.50), and reflecting heterogeneity in fee requirements, the average fee amount was 347,487 IDR (US$23). Some borrowers reported that mandatory deposits were deducted from the total loan amount, and they had been unaware of these requirements prior to loan disbursement. Communication about collateral requirements was also unclear for borrowers. Some learned from mass media campaigns or bank officers that no collateral was required for KUR loans but were required to submit collateral at the time of disbursement, causing frustration. 30 Supervisory bodies of the KUR program can trigger the sanction if they find the distributor requested additional collateral on a micro KUR loan. In these cases, the KUR interest subsidy is not paid on the concerned loans, or if disbursement has already taken place, the distributor must reimburse the State Treasury (Permenko No. 1 of 2023, Article 14). 22 While the KUR program data captures a wealth of information, several shortcomings limit its ability to track whether the program is reaching targeted businesses in an inclusive way. For example, the data system includes an open question on collateral; however, the formatting of the variable makes it difficult to assess the type or value of collateral presented or to cross-check this collateral with any standardized registries. In addition, the data system does not seem to accurately track applicants who are rejected from KUR. Tracking rejection is important to understand the program’s inclusiveness and possible ways to enhance inclusion.31 Clearer guidelines on how to enter certain data fields, data entry requirements, and training for bank officers could improve the quality and scope of the program’s monitoring and evaluation data. 4. Lessons Our paper shows that subsidized credit can enable entrepreneurs to enter the formal financial system for the first time. There are two main mechanisms that explain KUR’s impact on financial inclusion. First, the Government of Indonesia sets ambitious targets for the program, which can push financial institutions to focus on lending to MSMEs.32 On average, the KUR program has disbursed over 90% of its annual targets since 2015, and borrowers describe active outreach from bank officers in qualitative interviews. Second, entrepreneurs who are usually deterred from applying for formal loans due to perceived high interest rates feel more comfortable applying for subsidized loans. Indeed, qualitative interviews with borrowers reveal that in the absence of KUR many would not apply for other types of loans due to the high interest rate. Some report concerns about their ability to repay higher-cost loans, while others simply prefer to use their existing means than to pay more in interest. By supporting MSMEs in accessing the formal financial system for the first time, subsidized credit programs could have potential longer-term benefits. Our qualitative work shows that the KUR program helps MSMEs formalize their businesses, keep better financial records, and become familiar with bank practices and loan procedures. These changes and the credit history built through KUR loans can improve the bankability of MSMEs. In addition, experience receiving and repaying loans may boost entrepreneurs’ confidence in their ability to manage credit and demonstrate the potential benefits of using commercial finance to expand their business activities. This may increase entrepreneurs’ demand for future credit. The program may also enable banks to learn more about lending to the MSME segment, which could facilitate positive changes in the supply of MSME credit. 31 For example, knowing whether women are less likely to apply to KUR than men or are more likely to face rejection from KUR could provide guidance for how to promote women’s access to KUR. 32 The program is largely implemented by state-owned banks, which may be particularly sensitive to policy priorities. 23 Nevertheless, we find that these intermediate benefits are not translating into helping borrowers transition to commercial lending. Only 3% of KUR borrowers between 2015 and 2020 graduate to unsubsidized commercial lending. Rather, we find that firms are likely to continue using subsidized KUR loans. Approximately half of KUR borrowers received subsidized KUR loans multiple times during the 5-year period studied, and this is an underestimate of the likelihood to return to subsidized lending given many borrowers would not yet have reached the end of their first KUR loan tenor during the study period. Regulations in effect at the time of the study enabled repeat use of KUR loans, and neither borrowers nor banks had sufficient incentives to transition to unsubsidized loans. For borrowers, KUR loans offer much lower interest rates than unsubsidized loans, and qualitative work reveals that some borrowers change their borrowing patterns to take advantage of the lower interest rates. For banks, promoting repeated lending can enable them to meet the government’s large distribution targets without increasing their portfolio risk as much as targeting new borrowers. Following discussion of the preliminary results of this study, the Government of Indonesia introduced new regulations to promote graduation by imposing stricter rules on repeat borrowing and gradually increasing the interest rate for subsequent loans.33 These reforms orient the KUR program towards incentivizing graduation to commercial lending. Because existing evidence cannot detect whether borrowers could access unsubsidized loans after KUR if they wanted to, it will be important to assess whether these policy changes lead to the intended effect of greater graduation from KUR to unsubsidized commercial lending or whether KUR borrowers cease accessing commercial loans altogether. Our results demonstrate challenges in ensuring that subsidized credit reaches the target population. Although the KUR program had multiple design elements to target MSMEs without sufficient collateral, in practice, KUR borrowers were required to present substantial collateral. Indeed, the amount of collateral required for KUR loans during the study period exceeded that of other loans of a similar size. The de facto collateral requirements likely explain women’s lower access to KUR loans during the study period.34 Highly collateralized KUR loans are a sign that the partial credit guarantee embedded in KUR is not working as intended. Following discussion of the preliminary results of this study, the Government of Indonesia introduced new regulations that emphasize that collateral is not required for micro KUR loans and impose sanctions on financial institutions that still require collateral for these loans.35 Further research could help illuminate challenges with the partial credit guarantee mechanism. 33 Regulation Number 1 of 2023 of the Coordinating Minister of Economic Affairs, Concerning Guidelines for the Implementation of KUR. January 25, 2023. 34 After the period included in this study, the government introduced ultra-micro KUR, which offers smaller loan sizes, emphasizes that no collateral is required, and specifically targets women. Further research is needed to assess whether the additional KUR window facilitates women’s access to KUR. 35 Regulation Number 1 of 2023 of the Coordinating Minister of Economic Affairs, Concerning Guidelines for the Implementation of KUR. January 25, 2023. 24 Our results also highlight the importance of sufficient guidelines and training in programs that are implemented through multiple institutions at scale. Heterogeneity in implementation can arise when multiple entities implement a program, and without effective communication, this heterogeneity can create confusion and frustration for program recipients and increase challenges in oversight. Our findings reveal that KUR borrowers did not always have a clear understanding of loan terms and conditions and their rights and responsibilities as borrowers, raising consumer protection concerns. Standardized templates that harmonize the presentation of loan terms and conditions in simple language could help mitigate these consumer protection risks. These types of templates could also be adopted for other commercial lending products to enhance transparency and ensure borrowers have sufficient information to make decisions about loan product adoption.36 To facilitate oversight of the program and ensure that the program is reaching the targeted borrowers, improvements to the program’s monitoring and evaluation system can ensure data are reported in a harmonized and systematic way. Our results also identify several areas of future research to better understand the optimal design of public credit schemes for small businesses. While this study is informed by one of the largest databases of beneficiaries of public credit schemes in the world, the analysis relies on descriptive statistics and inferences and cannot conclusively answer questions about the additionality of access to subsidized credit. Importantly, we cannot detect what would have happened in absence of the program. Experimental studies with a valid counterfactual exploring the impact of subsidized credit schemes are needed to assess the additionality of these types of programs. Moreover, rigorous cost-benefit analyses of public credit schemes such as KUR could help determine optimal budget allocations and the fiscal costs of helping unbanked firms enter the financial system. Additionally, further research is needed to understand how policy changes that limit repeated access can influence graduation from subsidized to unsubsidized commercial credit over time. More evidence is also needed about how partial credit guarantee mechanisms can be optimized to ensure they reach borrowers with limited collateral. Finally, further research on how public credit schemes can effectively target firms in green segments of the economy could help connect programs like KUR with broader global initiatives around green and sustainable finance. 36 Harmonization in the way both subsidized and unsubsidized loan product information is presented could also facilitate graduation, as borrowers will already be familiar with interpreting the disclosed information. 25 References Antoniou, A., Guney, Y. and Paudyal, K., 2008. The determinants of capital structure: capital market- oriented versus bank-oriented institutions. Journal of financial and quantitative analysis, 43(1), pp.59-92. Arráiz, I., Meléndez, M., & Stucchi, R. 2014. Partial Credit Guarantees and Firm Performance: Evidence from Colombia. Small Business Economics, 43(3), 711–724. http://dx.doi.org/10.1007/s11187-014- 9558-4 Badan Pusat Statistik. 2021. “Keadaan Angkatan Kerja di Indonesia Februari 2021.” https://www.bps. go.id/publication/2021/06/08/b547a5642aeb04d071cb83d4/keadaan-angkatan-kerja-di-indonesia- februari-2021.html Bancel, F. and Mittoo, U.R., 2004. Cross-country determinants of capital structure choice: a survey of European firms. Financial management, pp.103-132. Banerjee, A. and Duflo, E. 2014. Do Firms Want to Borrow More? Testing Credit Constraints Using a Directed Lending Program. Review of Economic Studies, 81, 572-607. https://economics.mit.edu/ sites/default/files/publications/2014.04%20Do%20Firms%20want%20to%20borrow%20more.pdf Beck, T., Demirguc-Kunt, A. and Levine, R., 2005. SMEs, growth, and poverty: Cross-country evidence. Journal of economic growth, 10, pp.199-229. Beck, T., Demirgüç-Kunt, A. and Maksimovic, V., 2008. Financing patterns around the world: Are small firms different?. Journal of financial economics, 89(3), pp.467-487. Beck, T., Demirgüç-Kunt, A., & Pería, M. S. M. 2011. Bank financing for SMEs: Evidence across countries and bank ownership types. Journal of Financial Services Research, 39(1), 35-54. Bloom, Nicholas, Raffaella Sadun, and John Van Reenen. 2016. Management as a Technology? NBER Working Papers 22327, National Bureau of Economic Research, Inc. Booth, L., Aivazian, V., Demirguc‐Kunt, A. and Maksimovic, V., 2001. Capital structures in developing countries. The journal of finance, 56(1), pp.87-130. Carvajal A. and T. Didier. 2024. Boosting SME Finance for Growth: The Case for more Effective Policies. World Bank, Washington, DC. 26 Caprio, Gerard, and Asli Demirgüç-Kunt. 1998. “The Role of Long-Term Finance: Theory and Evidence.” The World Bank Research Observer, vol. 13, no. 2, pp. 171–89. Coordinating Ministry (Permenko) Regulation no 1/ 2023 Coordinating Ministry (Permenko) Regulation no 15/ 2020 Coordinating Ministry (Permenko) Regulation no 8/ 2015 Coordinating Ministry (Permenko) Regulation no 8/ 2019 Coordinating Ministry (Permenko) Regulation no 8/ 2020 de Brauw, A., Sayaka, B., Pasaribu, S., Ambler, K., Herskowitz, S. and Suryani, E., 2020. Agricultural value chain finance in Indonesia: Executive summary. Intl Food Policy Res Inst. De Jong, A., Kabir, R. and Nguyen, T.T., 2008. Capital structure around the world: The roles of firm-and country-specific determinants. Journal of banking & Finance, 32(9), pp.1954-1969. De Mel, Suresh, David McKenzie, and Christopher Woodruff. 2008. Returns to Capital in Microenterprises: Evidence from a Field Experiment. The Quarterly Journal of Economics 123, no. 4, 1329–72. Dea, V. 2019. An Overview of Support for Women Entrepreneurs in Indonesia and Canada: Focus on SMEs and Start-Ups. https://www.angin.id/wp-content/uploads/2019/03/An-Overview-of-Support-for- Women-Entrepreneurs-in-Indonesia-and-Canada.pdf Demirguc-Kunt and Maksimovic. 1998. Law, Finance and Firm Growth. Journal of Finance, 53 (6), 2107- 2137. https://doi.org/10.1111/0022-1082.00084 Demirgüç-Kunt, A. and Maksimovic, V., 1999. Institutions, financial markets, and firm debt maturity. Journal of financial economics, 54(3), pp.295-336. Demirgüç-Kunt, A. and Maksimovic, V., 2002. Funding growth in bank-based and market-based financial systems: evidence from firm-level data. Journal of financial Economics, 65(3), pp.337-363. Farida, et al. 2015. Micro Enterprises’ Access to People Business Credit Program in Indonesia: Credit Rationed or Non-Credit Rationed? International Journal of Economic Perspective, 9 (2), 57-70 Giannetti, M., 2003. Do better institutions mitigate agency problems? Evidence from corporate finance choices. Journal of financial and quantitative analysis, 38(1), pp.185-212. 27 International Labour Organization. 2019. Financing Small Businesses in Indonesia – Challenges and Opportunities https://www.ilo.org/jakarta/whatwedo/publications/WCMS_695134/lang--en/index. htm International Finance Corporation. 2016. Women-owned SMEs in Indonesia: A Golden Opportunity for Local Financial Institutions: Market Research Study. International Monetary Fund. 2021. International Financial Statistics and data files, and World Bank and OECD GDP estimates. https://data.worldbank.org/indicator/FS.AST.PRVT.GD.ZS?locations=Z4 Kapoor, M., Ranjan, P., Raychaudhuri. 2017. The impact of credit constraints on exporting firms: Evidence from the provision and subsequent removal of subsidised credit. The World Economy, 40 (12), 2854- 2874, https://onlinelibrary.wiley.com/doi/abs/10.1111/twec.12498 Kersten, R., Harms, J., Liket, K. and Maas, K., 2017. Small Firms, Large Impact? A systematic review of the SME Finance Literature. World development, 97, pp.330-348. King, R. G., & Levine, R. 1993. “Finance and Growth: Schumpeter Might be Right”, The Quarterly Journal of Economics, 108(3), 717–737. https://doi.org/10.2307/2118406 Klein, J. and Ogden, T., 2023. Lessons for Global Microfinance from… the United States? Financial Access Initiative Working Paper 2023-01 Lerner, J. and Schoar, A., 2005. Does legal enforcement affect financial transactions? The contractual channel in private equity. The Quarterly Journal of Economics, 120(1), pp.223-246. Levitsky, J. 1997. Credit guarantee schemes for SMEs – an international review. Enterprise Development & Microfinance, 8 (2), https://doi.org/10.3362/0957-1329.1997.013 Levitsky, J. 1997. SME Guarantee Schemes: A Summary. The Financier, vol 4 (1) Martokoesoemoa, D., B. M. Sinaga, N. Kusnadi, and Y. Syaukat. 2020. Business Vulnerability and Credit Access for Agriculture-Based Micro and Small Women Entrepreneurs. Economics and Finance in Indonesia, 66 (2): 97-111. McKenzie, David. 2023. Is there still a role for direct government support to firms in Developing Countries? Policy Research Working Paper; No. 10628. World Bank, Washington, DC. 28 OECD. 2018. SME and Entrepreneurship Policy in Indonesia 2018” OECD Studies on SMEs and Entrepreneurship https://www.oecd.org/publications/sme-and-entrepreneurship-policy-in- indonesia-2018-9789264306264-en.htm Oh, I., Lee, J.D., Heshmati, A. and Choi, G.G., 2009. Evaluation of credit guarantee policy using propensity score matching. Small Business Economics, 33, pp.335-351. OJK, Indonesia Banking Statistic. 2021. Volume 19, No. 9. https://www.ojk.go.id/ Ornelas, J.R.H., Pedraza, A., Ruiz Ortega, C. and Silva, T., 2019. Winners and losers when private banks distribute government loans: evidence from earmarked credit in Brazil. World Bank Policy Research Working Paper, (8952). Purwono, R., Nugroho, R., and Mubin, M. 2019. Response on New Credit Program in Indonesia: An Asymmetric Information Perspective. Journal of Asian Finance, Economics and Business. 6 (2): 33- 44. Rajan, R. G., and Zingales, L. 1998. Financial Dependence and Growth. The American Economic Review, 88(3), 559–586. World Bank. 2017. Indonesia Economic Quarterly, March 2017. https://elibrary.worldbank.org/doi/ abs/10.1596/30840 World Bank. 2021. Policy Effectiveness Review of Innovation and Business Support Programs (iPER). World Bank: Washington, DC World Bank. 2022. Equitable Entrepreneurship: Overcoming barriers for women in business in Southeast Asia. World Bank: Washington, DC Zia, B.H., 2008. Export incentives, financial constraints, and the (mis) allocation of credit: Micro-level evidence from subsidized export loans. Journal of financial economics, 87(2), pp.498-527. 29 Appendix A: Detailed Description of KUR Borrowers The vast majority of KUR debtors are microenterprises with fewer than five employees. Of the KUR debtors, 97 percent have fewer than five full-time salaried workers, a commonly used definition of microenterprises. Even considering broader definitions of what constitutes a microenterprise, 87 percent of KUR debtors have fewer than five paid workers, and 84 percent have fewer than five paid or unpaid workers (Figure A.1: Panel A). In fact, most KUR debtors do not have any paid workers at all. Only 20 percent of KUR borrowers have any full-time salaried workers, and 44 percent have any paid workers. Almost a quarter of firms have no paid or unpaid workers (Figure A.1: Panel B). Even among businesses that have workers, most KUR debtors classify as microbusinesses. The average KUR debtor that has any paid or unpaid workers has 2.9 workers, and the average firm employing full-time salaried workers similarly has 3 full-time salaried workers on average (Figure A.1: Panel C). Figure A.1: Most KUR debtors are microenterprises Panel A: A: Panel Panel %% of firms % A:of firms of firms with with less less with lessthan than than 5 5 workers 5 workers workers Panel B: Panel Panel % B: %of B: %firms of firms of firms with with with atat atleast least one least one one worker worker worker Panel A: % of firms with less than 5 workers Panel B: % of firms with at least one worker 100%100% 96.7% 96.7% 100%100% 96.7% 87.2% 87.2% 84.3% 84.3% 100% 100% 78%78% 80% 80% 87.2% 84.3% 80% 80% 80% 80% 78% 60% 60% 60% 60% 60% 60% 44%44% 40% 40% 40% 40% 44% 40% 20% 20% 40% 20% 20% 20% 20% 20% 20% 20% 0% 0% 0% 0% Full- 0% Full -time time salary salary Paid workers workers Paid All workers All workers Full- 0% Full salary time -time salaryPaid workers Paid workers All workers All workers Full-time salary Paid and (full(full andworkers part part -time -time ) (paid ) (paid Alland and workers unpaid unpaid ) ) Full-time salary Paid (full(full and part andworkers -time -time part ) (paid workers Alland and ) (paid unpaid unpaid) ) (full and part-time) (paid and unpaid) (full and part-time) (paid and unpaid) Panel Panel C: Average C: Average number number of workersfor of workersfor firms firms with at least with one at least worker one worker Panel C: Average Panel number C: Average of number workersfor of workersforfirms with firms with atat least least oneone worker worker 6 6 6 4.8 4.8 4 4 4.8 4 3 3 2.9 2.9 3 2.9 2 2 2 0 0 0 Full Full- -time time salary salary Paid workers Paid (full(full workers andand -time partpart ) All -time ) workers (paid All workers andand (paid unpaid unpaid) ) Full-time salary Paid workers(full and part-time) All workers (paid and unpaid) Source: Source: Author Author calculations calculations SIKP database using using SIKP database Source: Author calculations using SIKP database Source: Author calculations using survey of KUR borrowers The average turnover of KUR beneficiaries also qualifies most borrowers as microbusinesses. The SME Law 20/2008 defines microbusinesses in Indonesia as ones with less than 300 million IDR in annual sales (IFC 2016). In the month preceding the survey, the average KUR borrower had sales of 12.9 million IDR and the median firm had sales of 4 million IDR. Although the previous month’s sales may not be reflective of each month’s sales, if the average sales last month are multiplied by 12 months, the resulting 154.8 million IDR in average annual sales of KUR borrowers is well below the 300 million ceiling to be considered 30 a microenterprise. Indeed, using this approximation of annual sales, 88 percent of firms are considered microenterprises, and 12 percent are SMEs. Most KUR borrowers are the sole owners and managers of their businesses, though interesting gender patterns emerge. As shown in Figure A.2, 89 percent of KUR borrowers are the sole owners of their businesses, and 71 percent are the only managers. Female KUR borrowers are more likely to co-own their businesses than male KUR borrowers. The majority of female KUR borrowers who co-own their business (71 percent) co-own their business with their spouse, and 2 percent co-own with individuals who are not family members (Figure A.2: Panel C). However, the relationship to co-owners is more diverse for men. Among men who co-own their businesses, 35 percent co-own with their spouse, 37 percent co-own with another family member, 25 percent co-own with friends or neighbors, and 3 percent co-own with other individuals. Although women are more likely to co-own their businesses than men, they are just as likely to co-manage their businesses (Figure A.2: Panel B). Moreover, the majority of both male (71 percent) and female (70 percent) KUR borrowers who co-manage their businesses do so with their spouse (Figure A.2: Panel D). Figure A.2: Most KUR borrowers are the sole owners and managers of their businesses, but gender differences in ownership and management structure exist Panel A: Number of firm owners, Panel B: Number of firm managers, Panel byPanel Panel A: A: Number gender Panel A: A: Number ofof Number Number of of KUR main of KUR KUR firmfirm owner KUR firm owners owners firm owners owners Panel Panel Panel B: B: Number B: Number by Panel gender B: Number of of Number of of KUR main of KUR KUR firm KUR owner firm firmfirm managers managers managers managers with with withwith gender gender gender gender differences differences differences differences with withwith gender with gender gender differences gender differences differences differences 100% 100% 92.9 92.9 80% 80% 100% 88.8 88.8 100% 70.8 71.4 71.4 70.8 88.888.892.992.9 81.9 81.9 80% 80% 69.7 69.7 71.471.4 70.870.8 69.769.7 80% 80% 80% 80% 81.981.9 60% 60% 60% 60% 60% 60% 60% 60% 40% 40% 40% 40% 40% 40% 27.6 27.6 27.7 27.7 27.6 27.6 40% 40% 27.627.6 27.627.6 27.727.7 20% 20% 17 17 20% 20% 20% 20% 10.5 10.5 6.7 17 17 6.7 20% 20% 10.510.56.7 6.7 .6 .6 .4 .4 1.6 1.61 2.7 2.7 1.1 1.1 1 0% 0% 0% 0% .6 .6 .4 .41.1 1.1 0% 0% 0% 0% 1.6 1.6 1 1 2.7 2.7 1 11 Owner 1 Owner Owner Owner 2 22 2 Owne Owne OwneOwne rs rs rs rs Over Over Over 3 3 Owne Over 33 Owne Owne Owne rs rs rs rs 1 11 1 Mana ManaMana ger Mana ger ger ger 2 22 2 Mana ManaMana gers Mana gers Over gersgers Over Over 3 33 3 Mana Over Mana gers Mana Mana gers gersgers Ov Ov Overall Ov erall erall erall Men Men Men Men W W Women W omen omen omen Ov Ov Overall Ov erall erall erall Men Men Men Men W W Women W omen omen omen Panel Panel Panel Panel C: C: Relationship C: Relationship C: Relationship Relationship to to to co-owner to co-owner co-owner co-owner for for for for businesses businesses businesses businesses Panel Panel Panel D: D: Relationship Panel D: D: Relationship Relationship to Relationship to to co-manager to co-manager co-manager for co-manager for businesses businesses for for businesses businesses Panel C: Relationship with withwith to multiple with multiple co-owner multiple owners, multiple owners, owners, by for by by gender gender owners, by gender genderbusinesses Panel D: Relationship with with withwith multiple multiple to multiple co-manager managers, managers, multiple managers, by by for businesses by gender gender managers, by gender gender with multiple owners, by respondent gender with multiple managers, by respondent gender 80% 80% 80% 80% 71.5 71.5 80% 80% 70.7 70.7 70.9 80% 80% 70.9 70.1 70.1 71.571.5 70.770.7 70.970.9 70.170.1 60% 60%56.1 56.1 60% 60% 60% 60%56.156.1 60% 60% 40% 40% 34.5 34.5 37.1 37.1 40% 40% 40% 40% 34.534.5 31.3 37.1 31.337.1 40% 40% 31.6 31.6 31.331.3 27.4 27.4 27.427.4 25.3 25.3 27.2 27.2 24.431.631.6 24.4 hor calculations culations using using SIKP database 20% 20% SIKP database 25.325.3 20% 20% 27.227.2 24.424.4 culations or using using calculations SIKP database 20% 20% SIKP database 10.6 10.6 20% 20% 5.4 5.4 10.610.6 2 3.1 3.11.3 2 3.1 3.4 3.4 5.4 5.4 0% 0% .6 .6 .6 .6 2 1.3 2 3.11.3 1.3 0% 0% 3.4 3.4 .1 .1 .1 .1 .8 .4 .8 .8 .4 .8 1.4 1.4.4 1.4 .4 1.4 0% 0% 0% 0% Spouse Spouse Spouse Spouse Another Another Friends Another Another Friends Friends Friends or or oror Third Third Third Third Spouse Spouse Spouse Spouse Another Another Friends Another Another Friends or or ThirdThird Friends Friends or or Third Third F Family Family Family amily Member Member Member Member Neighbo Neighbo Neighbo Neighbo rs rs rs rs Parties Parties Parties Parties F Family amily Family Member Member Family Member Member Neighbo Neighbo Neighbo Neighbo rs rs rs rs Parties Parties Parties Parties Ov Ov Overall Ov erall erall erall Men Men Men Men Women W W omen W omen omen Ov Overall Overall Overall erall Men Men Men Men W W W omen Women omen omen Source: Author calculations using survey of KUR borrowers 31 Businesses from different sectors of activity receive KUR. The most common sectors of activity among KUR borrowers are trade (42 percent of borrowers), crop and animal production (22 percent of borrowers), and manufacturing (14 percent of borrowers) (Figure A.3). KUR regulations prioritize production sectors, including agriculture, hunting, and forestry sectors; marine and fisheries sector; processing industry sector; construction sector; salt mining sector; tourism sector; production service sector; and/or other production sectors that increase the number of goods or services (Permenko No. 8 of 2019). As shown in Figure A.3, a small share of KUR borrowers operate in some prioritized sectors, including construction, whereas large shares of KUR borrowers operate in some non-prioritized sectors, including wholesale and retail trade. Figure A.3: KUR borrowers operate in a variety of sectors of activity Share of KUR borrowers in different sectors of activity in different sectors of activity Share of KUR borrowers Wholesale and retail trade 41.9 Crop and animal production 21.5 Manufacturing 14.3 Food and beverage services 12.4 Other services 4 Other primary activities 2.7 Transport and accomodation 2.5 Construction .7 0 10 20 30 40 50 Source: Author calculations using survey of KUR borrowers Most KUR borrowers (85 percent) have some form of formal documentation for their business (Figure A.4). The most common form of documentation is the Business Certificate (SKU) or the Certificate of Business Domicile (SKDU), which is issued by local authorities and recognizes that the business exists and operates in the locality. Of KUR borrowers, 79 percent have this form of documentation for their business. Prior to 2019, other forms of business licenses, including the Trade License (SIUP) and the Micro and Small Business License (IUMK) were issued by the central government, and 12 percent of KUR borrowers had these forms of formal documentation. A business with a business license can apply for a tax identification number (TIN), which 10 percent of KUR borrowers have. Figure A.4: Most KUR borrowers have formal documentation for their business documentati Types ofof on of business documentation of business Types 78.5 80% 60% 40% 20% 11.8 14.7 9.9 0% Tax Number (TIN) Trade License (SIUP) Business Certificate (SKU) Do Not Have Any Micro and Small Business Certificate of Business License (IUMK) Domicile (SKDU) Source: Author calculations using survey of KUR borrowers 32 Businesses that receive KUR turn a profit, and most have aspirations to grow. The average firm had profits of 3.6 million IDR in the month preceding the survey (December 2020), and the median firm had profits of 1.5 million IDR. Prior to receiving KUR for the first time, 83 percent of firms reinvested their profits to fund their business needs, suggesting that firms are profitable enough to enable investments. In qualitative interviews, most KUR borrowers reported that their businesses enable them to meet their daily needs and that they have been able to grow their businesses. Most debtors also aspire to grow their businesses, usually by expanding into another sector or product line. However, some do not aspire to grow, including those who are still consolidating their current activities, recovering from an unexpected shock, or are aging. Figure A.5: Most KUR borrowers have secondary sources of incomes % of KUR borrowers in households with other income sources % of than the KUR borrowers of interest business with in households other income sources than the business of interest 100% 87.6 80% 70.4 60% 41.9 40% 20% 18.5 0% Respondent has Respondent has Any household Income from any salaried job secondary earnings member has salary job household member (any job) Source: Author calculations using survey of KUR borrowers While businesses that received KUR are profitable, most KUR borrowers have secondary sources of income. As shown in Figure A.5, 70 percent of KUR borrowers have secondary earnings from another job, and 19 percent of borrowers have a salaried job in addition to their business. Most households (88 percent) have income sources beyond the business that received the KUR loan, and 42 percent of KUR borrowers are in households where at least one household member has a salaried job. 33