THE ORIGINS OF SELF-EMPLOYMENT: Experiences of a Transition Economy

1. THE ORIGINS OF SELF-EMPLOYMENT: Experiences of a Transition Economy Leora F. Klapper(World Bank) joint with: ASLI DEMIRGUC-KUNT & GEORGE PANOS NBER Entrepreneurship Working Group MeetingMarch 9, 2007

2. Motivation • A flexible, well-functioning, and entrepreneurial labor market can contribute to economic growth through the efficient (re)allocation of labor and increased competitiveness. A better understanding of the determinants of entrepreneurship - the environment that motivates and supports the creation of self-employment – is essential for understanding the microeconomic foundations of economic growth. • In transition countries, the SME sector has been the largest creator of new jobs and the vast majority of these new enterprises are microbusinesses (Ayyagari, Beck & Demirguc-Kunt, 2004; Klapper, Sarria-Allende & Sulla, 2004). • Notable amount of interest on issues of mobility and transition into self-employment in developing countries, stressing financial, institutional, psychological & sociological factors (Djankov et al., 2005; 2006; Paulson and Townsend, 2004; 2005;) as well as labor market characteristics (Earle and Sakova, 2000; 2001; Dutz et al., 2001). • We examine the nature of the entrepreneurial decision for the transition to self-employment in Bosnia & Herzegovina (BiH) and its viability, using a rich panel survey for the years 2001-2004. • BiH offers an interesting setting to examine the aftermath of reforms on the dynamism of latent entrepreneurship.

3. Labor and Business Environment in Bosnia & Herzegovina Pre-1990s: • Part of the former Yugoslavia, a centrally-planned economy. • Although the environment was hostile to self-employment and entrepreneurship, BiH had a large entrepreneurial middle class (Earle and Sakova, 2000). 1990s: War & Transition • Establishment after a 4-year war in early 90’s - 6% of 4.4m population killed or registered missing - 60% forced to relocate & almost 1m left the country • 75% drop in GDPpc ($2,400  $600) • Informal sector accounting for as much as 50% of GDP. • The destruction of the stock of productive capital combined with the dislocation of private social safety nets and social capital decreased living standards and increased the vulnerability of the population to further economic shocks (World Bank, 2002).

4. More recent years: • Facing massive unemployment and a deficient social welfare system, the promotion of self-employment and microenterprise became a political priority. • New laws were passed and initiatives were taken to encourage SMEs and new firms. But BiH still faces major challenges…. • High & Rigid Wages in the Formal Sector. • High Taxation of Wages and Profits – High Employer Contributions. • Large & Growing Informal Sector (No Social Insurance). • Persistent Unemployment. • Slow Privatization – Public Sector dominates Formal Sector. • Difficult Access & High Cost of Credit. • Lack of Systemic Trust in the Regulatory & Financial Environment. • World Bank Doing Business Data (2005): Particularly low ranking, with respect to ease of starting a business, dealing with licenses, registering property and trading across borders.

5. Data • World Bank Living Standards Measurement Survey. • 4 Panel Waves, 2001-2004. • Sample designed to be representative at the country level, the entity level, and for urban, rural and mixed municipalities. • The individual- and household-level questionnaire included modules covering demographics, housing, education, labor, migration, health, credit and social assistance. • Waves 1 and 4 also contain modules on consumption • Sample: Labor force population, aged 16-65. • Final sample includes 5,370 individuals and 2,273 households

6. Labor Force Composition • Self-Employed: • Owner/co-owner of: (a) enterprise with workers (employers), (b) enterprise without workers (own account). • Carrying out independent (non-agricultural) activity. • Validation: Declared work-related characteristics such as earnings and paid pension and health contribution. (The remainders are considered informal sector). • Formal Sector Employees: • Employed in: private sector, public enterprises, and international organizations (with paid pension contributions). • No distinction between full-time/part-time (flexible employment not common). • Informal Sector: • Workers in private sector for whom pension contributions and taxes are not paid. • Unpaid supporting family members, farmers on own farm, workers engaged in other activity, such as sale of agricultural products. • Inactive, but reporting earnings or hours of work. • Common rule-of-thumb practice (World Bank, 2002). • Involuntarily Unemployed: • Individuals without a job, actively searching for one (able to take it if offered) • Inactive: • Students, pensioners and housewives (with no earnings-hours). • Individuals inactive in all four waves of the panel are dropped.

8. New Entries into S.E. in 2002-2004, by individuals not S.E. in 2001The data allows us to directly identify individuals that switched during the sample to self-employment, as well as the ex-post performance of new entrepreneurs in terms of their survival in the early period in business.

9. Predicted Determinants of Entry into S.E. :1- Individual Characteristics: • MALE: Positive • AGE & AGESQ: Positive and Concave (Job Shopping: Miller, 1984) • EDUCATION: Mixed (Le, 1999; Bates, 1995; Parker, 2004) • MARRIAGE & CHILDREN: Positive (Borjas, 1986) • HEALTH & DISABILITY: Mixed – The disabled have access to both targeted grants/microcredit schemes (to address discrimination) and government social benefits. • REGIONAL VARIATION: Mixed (Urban vs. Rural, FBiH vs. RS), (Parker 2004)

10. 2)Psychological & Sociological Traits: • The self-employed are more likely to be risk-takers than the rest of the population, since they are faced with more uncertain future prospects and lifetime earnings profiles. They are also more likely to depend on social networks. • Cantillon, 1755 (Uncertainty Resolution); Say, 1828 (Coordination Ability); Knight, 1921 (Alertness); Schumpeter, 1934; 1939 (Innovation, Instinct, Leadership); Leibenstein, 1968 (Risk-Taking), Gomez & Santor, 2001; Ravallion & Lokshin, 2006 (social capital/risk sharing) SOCIAL CAPITAL: “Is there someone you can count on to listen or help in a time of crisis?” • Entrepreneurial decisions are much likely to be influenced by attitudes, emotional predispositions (“optimism”) and cognitive biases. • Arabsheibani et al., 2000; Puri & Robinson, 2005, Heaton and Lucas, 2000; Moskowitz and Vissing-Jorgensen, 2000; Hamilton, 2000; Gentry and Hubbard, 2001; Parker, 2006; Fraser and Greene, 2006 OPTIMISM: Index from 8 GHQ questions about mental health and anticipatory feelings (i.e. “Felt hopeless in terms of the future?”)

11. 3) Labor Market Experience: • UNEMPLOYED, INFORMALLY EMPLOYED, (Formally Employed) • Unemployment : “Pushing” • Past formal employment experience: “Pulling” • Ability Learning (Jovanovic’s, 1982), Human capital acquisition (Lucas,1978), Small Firm experience (Boden, 1996) • Informal sector experience: “Pushing” or “Pulling”? • Could informal sector experience - despite all its negative consequences – be valuable for economic development in transition? (Kaufmann & Kaliberda, 1996; Johnson, Kaufmann and Shleifer, 1997; Blau, 1985) • Informal work experience develops skills and abilities, which could later foster healthy transitions to formal entrepreneurship when circumstances allow it (Entrepreneurs as “jacks of all trades”, Lazear, 2004). • In a weak regulatory environment, the informal sector could be a subsistence shelter for unmatched employees and latent entrepreneurs (Harris and Todaro, 1970).

12. 4) Wealth, Access to Finance & Institutions: • Access to capital has been extensively analyzed as an important constraint in new business establishment, usually indicating a positive relationship between wealth, instrumented or not, and entrepreneurial activity. • Empirical evidence on the role of financial institutions is relatively scarce (Paulson and Townsend, 2004). • The effect of remittance receivership from abroad on local entrepreneurial activity is ambiguous. (Funkhouser, 1992; Ahlburg, 1995; Rodriguez and Tiongson, 2001; Muço et al., 2004; Amuedo-Dorantes and Pozo, 2006). • Financial Indicators: • LHHCONS (Wealth): Equivalized per capita household consumption, deflated by the regional poverty line. • Customer Affiliation with Financial Institutions: INFORMLOAN, MICROLOAN, BANKLOAN • Remittances: REMITDOM, REMITABROAD • Grants: GRANT, SOCIALSERV

13. Summary Statistics

14. Probit Results (Tables 6 & 7): • The profile of the newly self-employed is more likely to be male, aged 43, residing in an urban area, likely to be married and have some formal education, in good health, and with past self-employment experience. • Optimism and social capital have a significantly positive effect, both when included together and separately. • The inclusion of the consumption variable indicates a significant positive effect of past wealth on current self-employment. • The affiliation with informal sources of finance is not likely to foster entrepreneurial activities. Remittances from abroad exhibit a significantly negative effect on the probability of becoming self-employed, in accordance with the “disincentive effects” of remittances*. • Grants from institutional sources, such as international initiatives, have a positive impact on the probability to become self-employed, although the overall use is low. • Results are fairly robust when household heads and sub-samples of non-farmers and individuals coming from employment are examined separately.

15. Multinomial Probit Results (Table 8): • Multinomial Probit regressions distinguishing between (a) S.E. Employers (2%), (b) Own-Account S.E. (3%) and (c) Not S.E. • SE of either type are more likely to be male, older, and less formally educated. All SE individuals are also more likely to have previous SE experience and greater personal wealth. • Individuals creating employment for others are also more optimistic and have greater social networks. They are also less likely to be previously unemployed, receive informal sources of finance or remittances from abroad, and importantly, (weakly) more likely to make use of bank financing. • Individuals creating own account businesses are not more optimistic, but very likely to have strong social networks and other family members that are SE. They are also more likely to have informal sector experience.

16. Self-Employment Performance: • We examine the individual-specific determinants of survival in self-employment for at least 2 consecutive years (given the fact that our panel only covers the period 2001-2004) • As show in Table 2, 47.8% of individuals (53.1% of household heads) becoming self-employed during the years 2002-2003 in BiH quit their new entrepreneurial venture during their first year of activity. • It is well-known that new entrepreneurs bear the highest risk of failure during their first few years of activity (Parker, 2004). • Although firm and sector-specific determinants are also of vital importance, the nature of the database only enables person-specific analysis. • We employ a two-stage Heckman probit model with sample selection (van de Ven and van Praag, 1981), including optimism and receipt of pensions in the first stage.

17. Results: • Men and urban area residents are more likely to survive the difficult first year in self-employment. Education increases the probability of survival. • Prior wealth has a persistently positive effect in the survival equation, indicating that higher potential of self-financing is an essential component of self-employment activity and longevity in BiH. • The bank loan variable exerts a (weakly) significantly positive effect in the survival equation. This could be attributed to the good screening mechanisms of the financial institutions with respect to the entrepreneurial prospects of the individuals they choose to finance. Hence, while the decision to become an entrepreneur is not related to financing from banks, the ability to survive is significantly increased by an existing relationship with a bank. • Charities established to promote entrepreneurship appear to be unsuccessful in promoting success. • Informal sector employees are more likely to become self-employed than formal sector employees, and further more likely to make it through their first year. • Previous self-employment experience and the existence of another self-employed member in the household increase the likelihood of survival.

18. Main Conclusions • Informal sector workers are more likely to transition to formal self-employment and to be successful as entrepreneurs. These results support the perception of the informal sector as an incubator for self-employment in the formal sector in the early years of transition, through which individuals acquire skills that can facilitate their future entrepreneurial activities. • Financing constraints play an important role in the decision to enter into self-employment and its success; informal and formal sources of external finance do not appear important. • Overseas – and in some case domestic – remittances significantly decrease the likelihood of becoming an entrepreneur. • Behavioral profiles matter – “optimism” is important to entrepreneurs with employees, while “social networks” support own account entrepreneurs.