VIII. Bringing together the evidence, and reflecting on the theory

As described in this brief, studies that investigate the provision of information and a lower cost of formalisation have found at best modest effects on formalisation rates. This finding provides strong evidence against the “handcuffs” view of de Soto (1989), suggesting instead that firms rationally decide that the benefits of formalisation do not exceed the costs. De Mel, McKenzie, and Woodruff (2013) provide further convincing evidence that firms will only become formal when the benefits to doing so increase. Their experiment offered informal firms different monetary incentives to become formal, ranging from just enough to cover the cost of tax registration to an amount (LKR 40,000, or about USD 350) equivalent to approximately two months’ profit for the median firm.  They find that although no additional firms registered when only the cost of registering was covered, 20 percent registered when offered LKR 10,000 and 47 percent registered when offered LKR 40,000. By combining these incentives to register with the initial registration costs and the discounted value of future tax payments facing these firms, the authors show that the proportion of firms that were willing to become formal dropped quickly as the net cost of formality increased, as illustrated in the steep downward demand curve in Figure 1.

 Figure 1: The Demand for Formality in Sri Lankan Firms as a Function of the Net Cost of Formalising

Source: Based on De Mel et al. (2013), Figure 3.

The authors calculate that fewer than 15 percent of informal firms seem to view the net benefits of being formal as greater than zero. Interestingly, many of the firms that did not formalise when offered the maximum amount cited other barriers, such as land titling issues. Studies in other developing country contexts have found that the price of formal land titles is often prohibitive, and that even if land titling costs are subsidised there are important challenges in both the delivery and enforcement of these rights (Ali, Collin, Deininger, Dercon, Sandefur, and Zeitlin, 2016).

Reflecting on the different theories of informality, Ulyssea (2018) develops a unified theoretical model that encompasses the three main views of informality, by explicitly modelling the extensive and intensive margin of informality. In contrast to the majority of the existing literature, which only focuses on the extensive margin, considering the intensive margin allows one to break the direct association between firm and worker informality. Empirically, Ulyssea shows that the intensive margin accounts for a large share of total informal employment. His unified theoretical framework also reveals that the competing views of informality are not actually competing, but rather complementary to one another. For any given institutional framework, firms will choose whether and how much to comply given their own circumstances. Given heterogeneity of firms, all three views can be relevant at the same time. Estimating his model with data from Brazil allows him to infer the proportion of firms that corresponds to each theoretical view. The results show that the “handcuff view” corresponds to only about 9 percent of the informal firms in the data, while the “parasite view” corresponds to approximately 42 percent and the “survival view” the remaining 49 percent. The model predicts that only a very small proportion of informal firms are constrained by entry costs, and that reducing these formalisation costs would have limited effects on informality, which is consistent with the evidence from the studies described in this report. The modelling approach offers the potential for building upon the existing research that analyses private costs of formalisation to individual firms, by investigating the economy-wide effects and questions such as the often-made claim that informality leads to inefficient resource allocation in the wider economy, since informal firms who don’t pay taxes can compete away customers from more efficient tax-paying formal firms.

Previous Chapter VII. Does formalisation cause firms to grow?
Back to top
Next Chapter References