VoxEU Column Frontiers of economic research

Why a positive relation between entrepreneurship and conflict may not be good news

Though some studies propose that entrepreneurial activity increases during conflicts, macro evidence shows that a conflict is damaging to growth. This column argues that the conflict in Afghanistan did not contribute to economic development because it caused regressive structural changes at the micro level. It reduced employment opportunities and increased self-employment in activities that have low returns. To improve the economic resilience, self-employment in activities that are less affected by conflict should be stimulated.

Does entrepreneurship under fire contribute to economic development and growth? A significant number of studies show that some entrepreneurial activities increase during violent conflicts.1 Entrepreneurship is a key driver of economic development (Bandiera et al. 2013). However, macro evidence consistently shows that a conflict is a critical obstacle to growth (Collier 1999). Based on rich micro data that allows us to disentangle heterogeneous effects of conflict on different types of private entrepreneurial activities (PEA), we suggest that entrepreneurship under fire does not contribute to economic development.

The complex interactions between conflict and entrepreneurial activity

To date, the microeconomic literature has reached the following, apparently contrasting, conclusions on the micro relations between conflict and private entrepreneurial activities: violent conflict reduces the incentives to invest (Hoeffler and Reynal-Querol 2003); reduced investments and entrepreneurship, in turn, lower output and employment; fragile and conflict affected states should enter a vicious cycle of decreasing private entrepreneurial activities, decreasing labour demand, increasing enrolment into conflict (Blattman and Ralston 2015), decreasing labour supply, and further contraction of entrepreneurial activities, which further fuels conflict (Fearon 2004). However, substantial evidence shows that in conflict affected countries there are more people employed in conflict areas than in non-conflict areas (Iyer and Santos 2012), and that in a number of these countries private entrepreneurial activities are resilient, and even increase with conflict (Bozzoli et al. 2013, Guidolin and La Ferrara 2007).2

How does one reconcile the virtuous effects of private sector development (PSD), with the negative relation between growth and conflict at both macro and micro levels, along with the evidence of resilience and increased entrepreneurial activities under conflict? To answer this question we identify the effect of distinct conflict measures on distinct types of entrepreneurial activities (Ciarli, Kofol, and Menon 2015).

·         We find that in the case of Afghanistan, increased conflict induces an increase in low-value added entrepreneurial activities and a decrease in higher value added activities.

That is, given a U-shaped relationship between the share of entrepreneurial activities and economic development (Figure 1), the conflict rewinds the process of structural change.

Figure 1. Nascent entrepreneurship versus per capital income (2002)

Note: Nascent entrepreneurship is defined as “as the number of people actively involved in attempting to start a new business, expressed as a percentage of the adult population” (Wennekers et al. 2005, p 294)
Source: Wennekers et al. (2005), p. 302

Identifying entrepreneurship and conflict in Afghanistan

We identify the effects of conflict on private economic activity in four steps, combining spatial and instrumental variable strategies. First, we collected, cleaned, and harmonised several household and conflict databases, providing precise information on detailed types of economic activities, household background, characteristics and actors of all conflict events engaging US soldiers, and their locations.3

Second, we constructed a spatial grid in Afghanistan to facilitate analysis at a finer level of geographical detail than administrative districts (see Figure 2 for an example. 

Figure 2. Location of households with high income self-employment activity on a gridded Afghan map

Note: The households on the map are those that declare that in 2005 their main source of income is self-employment in an activity with a relatively large capital investment (see Ciarli et al. 2015) and the data website for a classification of household activities). The grid is an approximation of the square cells of ~33km length that define our areas of observation, and which replace the administraitve districts (also visible in the figure for easy comparison).
Source: Ciarli, Kofol, and Menon (2015).

Third, exploiting spatial and time variation of conflict and private entrepreneurial activity (Figure 3) we tested the probability that an Afghan household in a given area is engaged in one type of entrepreneurial activity (among small business, low and high capital non-agriculture self-employment, sale of agricultural products, and subsistence agriculture) for different indicators of conflict in the same area.

Figure 3. Location of relevant conflict events and location of households with a small business in Afghanistan (top panels: 2005; bottom panels: 2008)

                            conflict events                                                                  households with a small business

Source: Ciarli, Kofol, and Menon (2015)Note: Relevant conflicts are defined in Ciarli et al. (2015) and in the online support material. The households on the map are those that declare that in 2005/2008 their main source of income is a small business.

Fourth, we controlled for reverse causality by employing a shift share approach similar to what is commonly employed in the literature on the determinants of migration, using data on the 1979-1989 Soviet conflict.4

Higher entrepreneurial activities with lower entrepreneurial opportunities

In line with previous literature that has found resilience of entrepreneurial activity under conflict, we find that private entrepreneurial activity increases in areas of higher conflict. However, this is a sign of reduced income opportunities, rather than the other way round. Higher conflict intensity shifts households away from employment to low capital self-employment activities and agricultural sales. Self-employment with higher capital content, instead, decreases in high conflict areas (Figure 4).

Results are confirmed by the counter-intuitive increase of investment in non-capital-intensive agricultural inputs in areas under more intense conflict. We also find that the increase in low capital self-employment is stronger in high conflict areas that are also far from trade infrastructures (markets and roads). This last result further suggests that the observed self-employment is mainly of the subsistence type.

Figure 4. Relation between the number of conflict events and several entrepreneurial activity indicators

Note: Locally weighted scatterplot smoothing (LOWESS) estimating the relation between the number of relevant conflict events per district (normalised by population size) on the y-axis, and the percentage of households in the same district whose income source is mainly self-employment in the non-agricultural sector (top-left), sales of agricultural products (top-right), a small business (bottom-left), and payed employment (bottom-right), on the x-axis.
Source: Ciarli, Kofol, and Menon (2015).

Although the general pattern is consistent for a number of conflict indicators (intensity, impact, and frequency), we find interesting differences when we distinguish between parties based on number of casualties. In particular, it is the number of casualties among the civil population that drives the results of the impact measure of conflict. We find less precise estimates when we use self-reported violence, which is more markedly correlated to an increase in self-employment in agriculture-related activities.

Using various measures of annual peace frequency we study the effect of the expectation of violence (based on the pattern of past violence). We find that households’ activities are affected only by the total number of days of peace, not by the temporal pattern of conflict.


Our evidence suggests that conflict pushes households towards marginal self-employment activities and towards agriculture, although a few households do gain from doing business with the foreign army.

Going back to the question that we raised at the beginning of this article, we can provide an empirical explanation for why more entrepreneurship may be associated with stronger conflict and lower economic growth.

  • The negative effect of conflict on economic growth observed at the macro level operates through a regressive structural change at the micro level. Conflict reduces employment opportunities and increases self-employment in activities that have low returns.

Violent conflict, including when driven by a foreign coalition, moves a country to increased autarchy, characterised by a higher number of people casually self-employed with lower income (if we exclude the income generated by aid and military operations). Such regression may require a long time before the country can change direction again. Therefore, the long-term effect of conflict on the economy is likely to be negative, even though people tend to resume their lives quickly, and hold on to their entrepreneurial – some would say survival – capabilities.

‘Collateral’ damages represent a huge toll for countries that are under fire, not only on human lives, but also in terms of future economic development. In Afghanistan, the number of civilian casualties has the strongest effect on a household’s decision to invest in a private economic activity of a given type.

  • However, despite the evident negative effects of conflict, international organisations and aid agencies can improve the resilience of economic activities by stimulating self-employment in activities that are less affected by conflict, especially in areas were civilians are less targeted.

Some of these entrepreneurial activities may become leverages for economic development after the conflict.

Disclaimer: The views expressed in the column are exclusively those of the authors and do not represent the views of the instituitons or organizations they are affiliated with.

This column draws from research carried out as part of PEDL project 450. For more information see http://pedl.cepr.org/node/1220


Bandiera O, R Burgess, N Das, S Gulesci, I Rasul and M Sulaiman (2013), "Can Basic Entrepreneurship Transform the Economic Lives of the Poor?" IZA Discussion Papers IZA DP 7386, Bonn.

Blattman, C and L Ralston (2015), "Generating Employment in Poor and Fragile States: Evidence from Labor Market and Entrepreneurship Programs", SSRN Electronic Journal (June 25)

Bozzoli C, T Brück and N Wald (2013), “Self-Employment and Conflict in Colombia.” Journal of Conflict Resolution 57(1): 117–42.

Brück T, W Naudé and P Verwimp (2011), “Small Business, Entrepreneurship and Violent Conflict in Developing Countries,” Journal of Small Business and Entrepreneurship 24(2): 161–78.

Ciarli T, C Kofol and C Menon (2015), "Business as Unusual. An Explanation of the Increase of Private Economic Activity in High-Conflict Areas in Afghanistan", London.

Ciarli T, M Di Maio and S Langlotz (2015), "Difficult Relationships: Conflict, Entrepreneurship and Economic Development", SERC Discussion Paper 182, London

Collier, P (1999), “On the Economic Consequences of Civil War,” Oxford Economic Papers 51: 168–83.

Fearon, J D (2004), “Why Do Some Civil Wars Last So Much Longer than Others?” Journal of Peace Research 41(3): 275–301.

Guidolin, M, and E La Ferrara (2007), “Diamonds Are Forever, Wars Are Not: Is Conflict Bad for Private Firms?” American Economic Review 97(5): 1978–93.

Hoeffler, A, and M Reynal-Querol (2003), Measuring the Cost of Conflict, Oxford: Oxford University Press.

Iyer, L and I Santos (2012), "Creating Jobs in South Asia’s Conflict Zones", Policy Research Working Paper 6104, Washington DC.

Wennekers, S, A van Wennekers, R Thurik, and P D Reynolds (2005), “Nascent Entrepreneurship and the Level of Economic Development,” Small Business Economics 24(3): 293–309.


1 For recent reviews see Brück et al. (2011) and Ciarli et al. (2015)

2 For more examples see Ciarli, Di Maio, and Langlotz (2015).

3 Details on the harmonization of Afghan data can be found here.

4 This method is similar to accounting for the level of conflict which depends only on external physical and geographical factors (such as ruggedness, accessibility, position, etc), and not on factors which may be related to the current level of entrepreneurship.

2,100 Reads