This post by Lori Beaman, Jeremy Magruder and Jonathan Robinson.
Hundreds of millions of people in the developing world work in microenterprises. These businesses tend to be very small, often employing only a single operator, and they tend to have difficulty growing. Yet growing evidence suggests that such businesses could increase profits by increasing investment – a number of recent studies find that the marginal return to capital among small firms in developing countries tends to be very high (i.e. de Mel et al. 2008). If returns to capital are high, why don’t microenterprises borrow, invest and grow rapidly?
The obvious answer is that these firms don’t have access to credit. But while credit constraints are likely part of the explanation for the puzzle, accumulating evidence suggests that it’s not just credit that limits investment: a number of recent studies suggest that relieving credit constraints alone are not sufficient to allow firms to increase investment (i.e. Banerjee et al. 2010; Crépon et al. 2011; Karlan and Zinman 2010; Attanasio et al. 2012). Another reason to question whether credit is the sole constraint facing firms is that the return to seemingly very simple business decisions – such as marginally increasing inventory (i.e. Kremer et al. 2013) – seems to have sizeable returns (even though the amount involved in such investment decisions is fairly small).
These studies suggest that firms may not be perfectly optimizing along all dimensions. And in fact, several recent studies suggest that fairly basic training on simple management practices can have positive effects (see, for example, the management consulting provided in Bloom et al. 2012 or the “rule of thumb” training evaluated in Drexler et al. 2012).
So what’s going on? Though there are several explanations for these results, part of the story may be that firm owners have a finite amount of attention to spend and that some aspects of the business get overlooked. For example, a business owner may have to divide attention between home and work and may therefore spend less time on a particular business decision because she’s worried about what’s going on at home (Banerjee and Mullainathan 2008). If so, helping the firm owner think through that business decision may improve productivity and profits.
In a recent study, we tested this idea, focusing on one business decision which does not seem to depend on credit constraints: deciding how much change to keep on hand to break larger bills among small business owners in Kenya. Not having enough change is a surprisingly large problem for these firms: in surveys, firms lose an estimated 5-8% of profits from not having enough change on hand.
To test whether limited attention might be a reason for this phenomenon, we conducted two interventions. First, the survey instrument itself, which asked about lost sales due to insufficient change, served as a treatment. If the survey served as a type of “reminder” that made lost sales more salient, it might induce behavior change. To test this, firms were randomly enrolled into the program at different points in time. Thus, the effect of the survey could be tested by measuring lost sales between firms that had been surveyed for longer and shorter times (and that therefore received more or less “reminders”). Second, after following firms for several weeks, field officers calculated lost profits and then shared this information with firms. This “information intervention” included results for both the firm and the market average.
Interestingly, both interventions changed behavior. Firms lost fewer sales and profits appeared to increase. It also appeared as if firms started bringing in more cash in the morning. Since these interventions provided little more than subtle reminders of the change problem, it seems likely that limited attention is the main explanation for these results.
While change management is certainly not the most important decision a business makes, the results suggest that constraints such as limited attention may be relevant to understanding how microenterprises operate and how to improve their performance. This is just one study in one country, but there are a number of ongoing projects on these and similar topics out there – it’ll be fascinating to see whether and how these results generalize to different settings.