Overall, we find that the single numerical estimate of legally required time for firms to complete certain legal and regulatory processes provided by the Doing Business survey does not summarize even modestly well the experience of firms as reported by the Enterprise Surveys... The most interesting difference between these two approaches to assessing the business or investment climate is that Doing Business focuses on de jure processes and Enterprise Surveys on de facto practice.
The Doing Business estimates are problematic in four ways:
First, there is huge variance reported by firms within the same country.
Second, the average times reported de facto in the Enterprise Surveys are much, much less than de jure times reported by Doing Business... at low values of the Doing Business time (estimated assuming full regulatory compliance) there are some countries in which the median of Enterprise Surveys responses (reflecting actual experiences) is above the Doing Business value. But at higher levels (above 30 days), the reported medians from the Enterprise Surveys are nearly uniformly below the de jure Doing Business values...
Third, there is almost zero correlation across countries between the single Doing Business survey number and the Enterprise Survey responses of firms.
Fourth, for those countries with repeated Enterprise Surveys data, changes in the reported Doing Business times are not strongly associated with changes in actual times as reported in the Enterprise Surveys, and if anything, reductions in Doing Business times are associated with higher reported actual times in the Enterprise Surveys... When firms are asked about time to obtain an operating license or a construction permit, or to clear goods through customs, the responses not only do not cluster tightly around the Doing Business value, they also do not cluster around the Enterprise Surveys central tendency. A significant fraction of firms report that regulatory compliance (or their version of compliance) takes essentially no time at all, while others report significant delays... The Enterprise Surveys data show that there is typically far more variation across firms in the same country than in the typical firm across countries.
When strict de jure regulation and high rates of taxation meet weak governmental capabilities for implementation and enforcement, we argue that researchers and policymakers should stop thinking about regulations as creating “rules” to be followed, but rather as creating a space in which “deals” of various kinds are possible... It is commonly observed that policy implementation often deviates from the stated policy in firm-specific ways, but this hypothesis has not been easy to document. It appears that when strict rules meet weak state capability—or, more broadly, “institutions”—the rules bend and become more like individuated “deals” where outcomes are not the result of a neutral application of policy to the facts but rather have to be negotiated case by case.
Within this space there will be winners and losers—among firms and among officials. The structure of these deals can in some cases reduce times to get permits, but the very uncertainty itself can favor officials... one hypothesis for explaining the differential responses to “policy reform” may be that when de jure and de facto policy diverge, the impact of de jure reform might have wildly different effects
Mary Hallward-Driemeier/Lant Pritchett How Business is Done in the Developing World: Deals versus Rules, Journal of Economic Perspectives—Volume 29, Number 3—Summer 2015—Pages 121–140