Conflict and Entrepreneurship in Afghanistan
Tommaso Ciarli, Saeed Parto and Maria Savona
Afghanistan remains one of the poorest countries in the world with an estimated per capita income of 300 US dollars and average mortality age of 47 years. By all accounts, progress to date for moving Afghanistan out of poverty and on to a path of economic and socio-political recovery has been arduous and slow. In recent years government and donor supported programmes and policies have increasingly been concerned about the development of the private sector and tapping into Afghanistan's pool of entrepreneurial capacity. It is not clear, however, how best to proceed since there continues to be a lack of reliable empirical evidence (and analysis) on the current state of entrepreneurship in Afghanistan.
In this article we summarize the results of a recent study that aimed to assess the impact of conflict on entrepreneurial activity in Afghanistan, and explore the household and environmental factors that are related to entrepreneurial choice. The study used the National Risk and Vulnerability Assessment (NRVA) of Afghanistan households carried out in 2005. The results were presented at the UNU-WIDER workshop on Conflict and Entrepreneurship held in Londonderry, Northern Ireland in March 2009.
According to the literature, entrepreneurship plays an important role in driving countries out of poverty traps, and even out of conflict situations. Part of the literature also stresses that, rather than entrepreneurship per se, it is productive entrepreneurship that nurtures development. On the other hand, conflict is likely to have an impact, directly or indirectly, both on the possibility of holding an entrepreneurial activity, and on the motivations to do so (whether for survival or because of the exploitation of new business opportunities).
In the mentioned study we investigated whether very different levels of conflict intensity across Afghan districts have a direct impact on the diffusion of entrepreneurial activity, and the conditions under which entrepreneurial activity is more likely to occur, despite the intensity of the conflict. Our analysis controlled for five different dimensions: (a) household features, (b) access to resources, (c) governance institutions, (d) infrastructure and (e) social capital. Based on this rich information, we were also able to provide some first evidence on the type of entrepreneurship (and underlying motivations), and on the distinct roles of direct and indirect effects of the conflict.
Empirical Evidence on Entrepreneurial Activity in Afghanistan
We define an entrepreneurial household, which we henceforth use interchangeably with an entrepreneur, as one that earns parts or all of its income from a 'small business'. To reduce the noise of definition fuzziness we exclude income from wage labour, growing crops, raising livestock or food production and sale, opium production and sale, small services such as milling, petty trade, sale of wood or transportation, and handicrafts. Those services are in fact closer to a definition of self-employment than entrepreneurship, and are thus more likely to be temporary activities.
Based on the survey we estimate that 9% of the households in Afghanistan rely on a small business as a source of income. One of the striking findings is that entrepreneurs in Afghanistan appear to have a strong tendency to adapt to ongoing conflict and continue to operate across districts with very different levels of security and conflict activity. Even when controlling for different sets of variables the results indicate a negative, though very small, impact of the intensity of conflict on entrepreneurship. The second very relevant result is that such effect is reduced by only one fourth when we further control for access to resources, social capital, formal institutions, and infrastructure. Some dimensions contribute more than others to this reduction (namely institutions, infrastructure and social capital) but none of them can firmly be considered to be the transmission chain of an indirect effect of conflict. In other words, social capital, institutions and infrastructure turn out to be more complementary than substitute explanations of the impact of conflict on entrepreneurship. Third, our analysis quite clearly suggests that entrepreneurial activity is mainly a means to survival, rather than of an entrepreneurial 'spirit' and is therefore rather risk averse. In the following paragraphs we sum up the main results that stand behind these three findings. We then turn to discuss potential indirect effects not made explicit in the analysis, and reflect on policy implications for a sustainable reconstruction of Afghanistan.
First, many of the household features replicate the evidence found in the standard literature on entrepreneurship, but there are also quite a few Afghanistan-specific effects. By far the strongest positive effect on the probability of being an entrepreneur is exerted by the number of activities in which the household is involved, followed by the household size. This suggests that it is much easier to hold a business when the risk of incurring income losses is covered by the involvement in a large number of other activities. Risk-averse entrepreneurship is confirmed by the effect of household assets: while assets do increase the likelihood of holding a business, the effect is non linear, and a too large number of assets prevents entrepreneurial activity.
Second, entrepreneurship is mainly a coping strategy in Afghanistan. This is reflected in the finding that access to resources is not related to entrepreneurial activity. Even if entrepreneurs use loans more for business investment, they access loans less than non-entrepreneurs, and not through formal credit institutions. This is consistent with risk-averse entrepreneurship, which is not aimed primarily at pursuing opportunities but at survival.
Third, a similar result applies to social capital, which does not play a substantial role in favouring entrepreneurship in a country that has seen conflict for an entire generation. In fact, when considering financial and social capital, neither seems to be an indirect channel through which the intensity of conflict across districts plays a role. But social capital is built in the very long run, and this result is likely to be because of the static nature of our analysis.
Fourth, entrepreneurial households appear to have a weak preference for communities with lower security in property rights, larger possibility of regulatory capture and rent seeking, and a smaller participation by the rest of the community in the policy making process. We also observe a small but insignificant positive relation between being part of a governmental body and having a small business. These two findings would support a hypothesis on predatory entrepreneurship, which may, or may not, be productive. Certainly, continued conflict has generated strong incentive dynamics for non-productive and destructive entrepreneurship in Afghanistan. Numerous warlords and people of influence have benefited handsomely from the conflict by getting involved in the many physical reconstruction projects that require local counterparts and contractors to be implemented. Corruption and nepotism remain rife in all manner of aid contracting in Afghanistan.
Finally, significant effects of the ongoing conflict appear to operate through the inadequate access to markets and lack of adequate infrastructure to support business activity. Arguably, these two factors are products of the ongoing conflict and hence conflict has a significant but indirect effect on entrepreneurship. At a minimum, a continued inadequacy of access to markets and infrastructure is likely to prevent the process of expansion and moving up on value chains by the productive entrepreneurs. The nonlinear relation between trade infrastructure and likelihood of entrepreneurship reinforces the hypothesis of the survival 'non–productive' entrepreneurialism: without any access to markets, communities rely on autarchic production.
Our analysis shows that entrepreneurial activity has persisted in Afghanistan despite the ongoing conflict for the last 3 decades, and is not affected by the intensity of conflict across locations. Tapping into this survivalist entrepreneurialism as a key component of the internationally supported reconstruction efforts requires a careful and clear assessment of the entrepreneurs' needs, trajectories, and ambitions, as well as a more clear assessment of the indirect channels through which conflict affects entrepreneurial activity.
While the information collected through the NRVA surveys is useful in providing a picture of how things unfold in the socio-economy, they are insufficient as the basis on which to develop intervention strategies to introduce changes aimed at supporting productive entrepreneurship. Afghanistan is a difficult context for conducting complementary population surveys on specific aspects of households' behavior and making survey data available in a timely manner so as to avoid rapid data obsolescence due to the chronic conflict.
Because of these constraints, a further conclusion we make is that research needs to focus as well on narrative-based case studies of entrepreneurial activity to contextualize and accompany formal analyses based on survey data such as we have attempted to do in our study. If we want to know how entrepreneurs cope and whether they are productive, non-productive, or destructive in conflict situations, we need also to ask entrepreneurs involved in economic activity.
To conclude, it is at best difficult to generalize on cases like Afghanistan, which are diverse, fragmented, and fluid. Future research will need to draw on the available quantitative databases – as we have attempted to do – and on locally specific case studies of selected segments of the economy for more depth to inform intervention through policy by the Government of Afghanistan and international donors.
About the authors
Tommaso Ciarli is a researcher at the Max Planck Institute of Economics, Germany. He currently works on economic development, technological and institutional change. His other main areas of research are micro evolutionary models of structural change, economic growth and distribution; organisation of industries and technological change; and diffusion of environmentally friendly technologies. He previously worked for ECLAC and UNIDO.
Saeed Parto is director of research at Afghanistan Public Policy Research Organization (APPRO) based in Kabul, Afghanistan and responsible for research programmes on aid effectiveness; socio-economic, civic, and political rights; natural resource and environmental management; governance; and political economy. He is also a visiting lecturer at the Faculty of Arts and Social Sciences, Maastricht University (the Netherlands).
Maria Savona is a lecturer at SPRU, Science and Technology Policy Research at the University of Sussex, UK; visiting research fellow at the Cambridge-MIT Institute, University of Cambridge. She has been associate professor in Economics at the University of Lille 1, France. She works on structural change and growth, the economic and employment impact of innovation in firms and sectors; the effects of internationalisation of production on innovation and economic performance.
See the UNU-WIDER project on 'Promoting Entrepreneurial Capacity'. The following WIDER research papers on entrepreneurship and economic development also discusses the relationship between conflict and entrepreneurship.
Wim Naudé, 'Entrepreneurship in Economic Development'
WIDER Research Paper 2008-20.
Wim Naudé, 'Peace, Prosperity and Pro-Growth Entrepreneurship''
WIDER Discussion Paper 2007-02.
WIDER Angle newsletter, November 2009