Business support for small and medium enterprises in low- and middle-income countries: a systematic review

Publication Details

 Piza, C, Cravo, T, Taylor, L, Gonzalez, L, Musse, I, Furtado, I, Sierra, AC and Abdelnour, S, 2016. Business support for small and medium enterprises in low- and middle-income countries: a systematic review, 3ie Systematic Review 25. London: International Initiative for Impact Evaluation (3ie).

Link to Source
Caio Piza, Tulio Antonio Cravo, Linnet Taylor, Lauro Gonzalez, Isabel Musse, Isabela Furtado, Ana C. Sierra, Samer Abdelnour
East Asia and Pacific (includes South East Asia), South Asia, Middle East and North Africa, Sub-Saharan Africa, Latin America and the Caribbean
Private Sector Development
Business Environment, Small Scale Enterprise
Equity Focus
None specified
Review Type
Effectiveness review


This review by Piza and colleagues is the first to systematically look at the evidence in the effect of programmes that support small businesses (small and medium enterprises, SMEs) on their performance, ability to create jobs, labour activity, abilty to invest, and level of innovation.

Main findings

Headline Findings: a summary statement

Whilst these finidings should be interpreted with caution, as there are significant limitations on the evidence base, the evidence suggest that, on average, SME-support interventions improve indicators of the firm's performance, employment generation, labour productivity, exports, and investments.

Evidence Base

This review includes a total of 40 studies covering the period between 2003 and 2010, of which 36 studies were used for the meta-analysis and meta-regressions. Of these, twenty-six were located in Latin America, six in Asia, six in Africa, and two in Europe.  The intervention categories evaluated are matching grants (eight studies), export promotion (eight studies), innovation programmes (seven studies), tax simplification (six studies), training (six studies), access to credit (four studies), local productive systems (three studies), formalisation (three studies), and cluster interventions (two studies).  15 studies focus on the manufacturing sector, while 13 included all sectors and the remaining 12 focus on either a different sector or combination of them. The studies yield 72 effect sizes (ES), of which 27.8 per cent evaluate firm performance, 20.1 per cent employment, 15.3 per cent exports, 11.1 per cent labour productivity, 8.3 per cent investment, and 8.3 per cent innovation outcomes. The outcome for firm performance is an amalgmation of measures for sales (growth), profits, production, value added, assets, and total factor productivity.

Implications for policy and practice

  • Meta-analysis indicates that business support to SMEs improves firms’ performance (average effect size (ES) of 0.13 standard deviations (SD)), helps create jobs (average ES of 0.15 SD), has a positive effect on labour productivity (average ES of 0.11 SD), leads to a small increase in exports (average ES of 0.04 SD), as well as an increase in the firms’ investment (average ES of 0.13 SD).
  • Overall there is no effect found on innovation.
  • When the analysis is disaggregated by type of intervention, they find that matching grants continue to improve firm performance (average ES of 0.15 SD) and employment generation (average ES of 0.14 SD).
  • The meta-regression shows that the interventions have a siginificantly larger impact on bigger firms and, more importantly, that once the authors control for risk of bias and only include randomised control trials (RCTs) then all effect sizes appear much smaller.
  • Additionally, funnel plots and Egger’s tests suggest that there may be a publication bias towards studies that report a positive impact on employment and labour productivity.

Implications for further research

The authors call for further research assessing the impact of SME support programmes in L&MICs. Only a few intervention types have been evaluated in more than two places, limiting the generalisability of the results. Replications of programmes across different contexts are therefore needed. The review also reveals a lack of, and need for, evaluations in the African context. Moreover, cost-effectiveness should complement evaluations to inform policy makers about the resources needed to achieve a given target.  The authors also emphasise that, considering the lack of evidence assessing the impact of SME-support programmes, researchers should use all available studies but report carefully their limitations in order to ensure an appropriate interpretation of the results.  To this end, researchers should also present a better qualitative discussion of the implementation processes of the interventions wherever possible.


Governments, development agencies and organisations around the world have sponsored many assistance programs targeted to small and medium enterprises (SMEs) and aimed at spurring firms’ performance regarding innovation, productivity, exports and employment generation. The support can be indirect through institutional reforms that address constraints that prevent SMEs from getting access to credit, or directly through training or value chain support, for instance. Despite being a widespread practice, there is limited evidence on the impact of SME support in the literature to date. Though some of the evidence indicates a positive effect of SME support programs on selected outcomes, there is a need to systematically review and synthesise the evidence to provide an unbiased account about the impact of business support for SMEs on firm performance. As the evidence appears to be predominantly from Latin America, a focus on its applicability to the African context is also warranted.

Research objectives

To systematically review and synthesise existing evidence on the impact of business support services for small and medium enterprises in low- and middle-income countries by comparing the relative effectiveness of different direct and indirect business support services on firm-level outcomes. The review particularly seeks to understand how the available evidence applies to African contexts, and the mechanisms that contribute to the effectiveness of an intervention.


The authors included studies using randomised designs, quasi-experimental designs, and and other studies that control for selection bias to evaluate the impact of direct or indirect business support services on firm performance, employment, productivity, labour productivity (primary outcomes), innovation, export, investment, access to credit, formalisation and management practices (intermediate outcomes) of small and medium sized enterprises in low- and middle income countries. The authors included published and unpublished literature covering the period between 2000 and 2014. The authors electronically searched a large number of relevant academic databases, journal articles and portals of relevant institutions (e.g. UNDP), in addition to citation tracking and contacting experts int the field for relevant data. The review covered studies published in English, Spanish and Portuguese. The authors conducted double-independent screening of full texts. They also conducted double-independent risk of bias assessesment, for which they used the 3ie risk of bias tool.  The authors used standaridised mean differences (SMD) to code continuous variables and risk ratios (RRs) to code binary variables. The authors furthermore tried to account for heterogeneity within and between studies by estimating random effects models and using meta-regression to test for the impact of moderating variables. Sensitivity analysis was also conducted.

Quality assessment

High confidence can be had in the findings of this review, however, it is important to note the limitations on the generalizability and reliability of the results (as described by the authors). They use appropriate methods to identify, critically appraise and synthesise studies. They clearly report the characteristics and results of the included studies, and the methods used for the synthesis of outcomes. The authors moreover describe the extent of heterogeneity and justify the synthesis of different outcome indicators and intervention types.

Source link


Additional publications

Photo © World Bank

Scroll to Top