The stated objective of investments by DFIs and impact investors to improve the supply of finance to SMEs is often job creation. Exactly what that means is typically left undefined. This blog examines job creation in more detail, and also expresses a little frustration with the available academic evidence on this question.
We recently published a BII Insights paper “How and why we finance SMEs” that describes our approach towards SME financing, why we think the SME financing gap exists, and how we can help close it. It also gives our view on the impact case for SME financing. To inform ourselves, we commissioned an evidence review from the International Growth Centre: Why Do SMEs matter?
The evidence leaves us with unanswered questions about job creation by SMEs. There are many estimates of job creation by SMEs but the statistics are often of limited use, and while many studies look at the effect of SME financing on job creation, they rarely tell us whether workers are really benefiting. However, one reason why research is of limited use to practitioners could be that practitioners are not vocal enough about the questions that matter. Politicians and investors often promise job creation, but they are not often very precise about what they mean. We must be clearer about what an objective of job creation really entails, and hence what evidence we need.
The first thing to get straight is the distinction between gross and net job creation. If a restaurant takes out a loan to expand and hires new staff, that is gross job creation. The effect on overall employment is unclear because, for example, a competing neighbourhood restaurant may lose business and shed jobs as a result. Net job creation is the sum of gross job creation and destruction. Some studies, such as Bazzi et al (2023), find that SME credit programs have no effect on the overall level of employment, because job creation at borrowing firms is offset by destruction elsewhere.[1]
Does that mean SME financing does no good for job creation? Not necessarily. In developing countries with rapidly growing populations and youth unemployment problems, net job creation would seem to be the obvious policy priority. Yet as Feng et al. (2023) show, unemployment tends to be very low in poorer countries, especially for lower-skilled workers, because in the absence of a comprehensive welfare state people must find some sort of work to survive. Donovan et al (2023) show that gross labour market flows – movements into and out of employment – are much higher in poorer countries, and that is not because people frequently switch jobs as they climb the jobs ladder to higher quality jobs, it is because of frequent movements on and off the bottom rung, into and out of low-quality jobs. Labour market churn – job creation that replaces short-lived bad jobs with equally bad jobs – is not the goal. The real development priority is the net creation of decent jobs, that are better than the precarious and badly paying occupations people would otherwise find, as the population grows. The process of replacing bad jobs with better jobs is driven by the gross job creation when new jobs are better than those they replace, and that can happen without an increase in overall employment.[2]
Bandiera et al. (2022) show that development is usually accompanied by an increase in the proportion of people with salaried jobs, but this process has stalled in Africa, which the authors attribute to a lack of larger formal sector firms. Formal sector firms tend to offer better jobs, but there can be comparatively decent jobs in the informal sector too (UNU-WIDER, 2022). There are various ways of looking at job quality – Hovhannisyan et al (2022) propose an index based on sufficient income, access to employment benefits, job stability, and adequate working conditions.[3] Research by the Small Firm Diaries programme shows that job stability can be very low at SMEs, with some employees only being engaged for weeks or months at a time.
Without knowing something about whether newly created jobs are pushing up job quality across society, it can be hard to know what statistics on SME job creation really tell us about their contribution to development objectives. Evidence about how many jobs SMEs create, relative to firms of other sizes, is not a great deal of use.[4] We cannot move from facts about where most jobs are created to a conclusion about where finance should be directed to create jobs. We need more. We need to know whether SMEs contribute to job creation in a way that is positive for development, as opposed to creating a large number of unstable low-quality jobs. There are two versions of that question: whether SMEs do that, and whether SME financing interventions get us more of it. Unfortunately, the empirical evidence we found does not put us in a position to answer those questions with much confidence.
Why we still think SME financing is good for job creation
At BII we do see the creation of better jobs as an important motivation for SME financing.[5] If we do not have quite as much direct empirical evidence about how workers benefit as we would like, what do we base that belief on? A combination of theory and less direct empirical evidence.
We are also informed by first-hand accounts from workers at the companies we have invested in. We sometimes conduct worker surveys, and workers often tell us their jobs are significantly better than their previous occupation – as the workers of Dayntee Farms told us, for example. We conduct these surveys when opportunity and need coincide, our sample is not representative of all the SMEs we indirectly finance, and worker surveys cannot tell us about net effects, but knowing that firms are creating jobs that their workers value, in contexts where good jobs are rare, cannot help but inform our thinking. We would certainly worry if worker surveys did not find that.
Theory also suggests that removing financing constraints on firm growth will result in better paid jobs (as the papers referenced in the introduction to this paper explain). We know that job quality tends to be higher in larger firms. The theoretical model in Bazzi et al. (2023) is very clear that better access to finance should push up wages by helping more productive firms enter – their empirical work did find that new entrants are more productive, but unfortunately they did not have data on wages. A similar paper by Cai and Szeidl (2023), finds positive effects of SME lending on jobs in borrowing firms in China, but also that their job creation is offset on net by job losses at non-borrowers, so there is no overall change in employment. They do collect wage data and find the positive impact on wages in borrowing firms is somewhat larger than the negative effect on wages at competitors, but the difference is not statistically signficant. That paper finds the main impact of SME financing was lower prices for consumers, and estimates a 63% social return on lending. In contrast to these two papers, Jaume et al (2021) find that an increase in SME credit supply in Mexico did raise overall formal employment in local labour markets, and they find the main effect was to create more higher wage jobs. Other than these three papers, most research only looks at the impact of lending on borrowing firms.
Beyond studies of specific interventions to increase lending to SMEs, there are numerous studies of bank branch expansion that find positive effects on local employment and wages, which we can infer occurs because small firms are getting better access to credit.[6] We also know that stability is an important aspect of job quality, and that access to finance helps firms weather volatility.
We have good reasons to believe better access to finance will result in better jobs at SME. We would like more empirical evidence about the overall economy-wide effects, telling us how the impact is shared between proprietors, workers, and customers; how large those impacts are in comparison to other interventions, such as financing larger firms; and what contextual factors predict when better access to credit for SMEs is likely to have a larger effect on job quality across society.
One challenge could be that the impact of firm growth depends on local labour market conditions – the owners of growing SMEs have little reason to raise wages if they can find workers easily. It could be that a sustained improvement in SME financing will only lift job quality across society over time, through the cumulative effect of new jobs being somewhat better than those being destroyed, which could be difficult for impact evaluations of a single intervention to identify. Or perhaps researchers will find something if they look harder.
[1] The fact that the overall effect of SME financing incorporates the impact on non-borrowers means that it cannot be observed in the impact metrics reported by SME lenders, such as employment growth at borrowing firms.
[2] We have made this point before. See Carter and Sedlacek (2019). Gross job creation – directly at firms receiving investment and indirectly through backwards and forwards linkages – is something investors can hope to measure or estimate.
[3] Job quality is a challenging thing for DFIs to report information about. It is hard to obtain the data, and hard to interpret. For some metrics, such as job turnover rates or unionisation rates, it is hard to say whether higher numbers are better or worse. There are arguments for an absolute defintion of a decent job, but also arguments that definitions should be context specific. Job quality varies a great deal within and across countries. It is not clear whether DFIs should be judged on levels – say for example the percentage of jobs in their portfolios that are decent -– or on changes (improvements in job quality, from different levels). There are difficult trade-offs here – targetting the level of decent jobs would tend to take DFIs away from lower income countries, where job quality is lower, and away from sectors within countries that are more likely to employ lower income people. All these problems mean that reporting on job quality by DFIs remains a work in progress and is recognised as an area for improvement.
[4] Many papers study job creation rates by SMEs, but they often only give a partial picture. How much SMEs contribute towards overall job creation results from the combination of employment growth by established SMEs and jobs created and destroyed as firms enter and exit. We found that many studies only report job creation rates by established firms.
[5] Other motivations include improving the quality of existing jobs, increasing the supply of a greater variety of goods and services at lower prices, and, if they initially have low incomes, raising the incomes of proprietors.
[6] More evidence about the impact of financial sector development is summarised in the blog “Why the financial sector matters for development.”
