Doing business project: Choice of the indicators, antinomies and possible rank biases

The Doing Business (DB) Project of the World Bank, started in 2002, has the purpose of measuring the legal and economic environment in 189 economies. Doing Business  focuses on domestic small and medium-size companies and on assessing the regulations applying to them through their life cycle. The DB report has the scope to gather  and evaluate quantitative data in order to compare economic and legal regulations across economies and over time.

The Doing Business Project of the World Bank acquired more and more reputation over time and a small variation on some sub-indicators may effect important policy measures with great implications on the global society (e.g. impact on the labour market, on political choices and on professional activities).

Therefore, it would be important to understand the general principles of the percentile rank and DB hypotheses for assigning the score to each sub-indicator. Furthermore, it would be essential to properly evaluate the rank criteria.

The assessment of some hypotheses and observed cases - carried out with a formal scheme of reasoning (for example, inequations, relational operators or model checking) - stresses the need of an empirical counterproof in order to avoid significant biases of the nature and the order of the considered phenomenon. 

For example, some hypotheses related to the calculation of the time spent on filing the procedures can lead to results that do not correspond to the empirical cases. Moreover, the choice of some synthesis indicators can change the scheme of reasoning and produce different rank results.

Actually, formal models are only an attempt to describe the reality because the world is too complex in order to be exactly described by mathematical methods. In fact we use simplifications and specific hypothesis in every model and, we do not claim that something is true or false but we state that something is more or less true with a certain degree of probability.

In some cases, if the reality is already giving us clear and simple elements, it would be reasonable to maintain these simple criteria and avoid mathematical transformations (e.g. Doing Business hypothesis for the "time sub-indicator"). In fact, our assessment shows some potential paradoxes emerging on the application of DB criteria and some biases that could occur with the DB synthesis method. In other cases, integrating the adopted DB criteria with additional indicators giving more detailed information on the analyzed data, would make the rank more significant and transparent.