"Serious flaws" in IMF labour market deregulation research

Explore similar themes
Politics & economy

The ILO recently published an analysis written by staff economist Mariya Aleksynskaof a series of four recent IMF research papers on the impact of labour market reforms.

The IMF papers concluded that deregulatory reforms could help reduce unemployment. The Fund has justified its support for substantial weakening of labour regulations, in Europe and elsewhere, on the basis of the results of these and other papers it has prepared on the same topic.

Aleksynska’s paper shows that the labour market data on which the papers were based and the way they were used by IMF researchers had serious flaws.

She states that most of the “reforms” identified by the IMF authors were actually explained by breaks in data series of which they were apparently unaware.

The ILO paper’s findings “call into question most of the empirical results of these [IMF] papers and policy advice based on them…. [T]hey risk encouraging policymakers to make hasty and ill-informed reforms on sensitive political issues with far-reaching economic and social consequences.”

The paper notes that in several cases deregulatory labour market reforms were imposed though IMF loan conditions.

Here are some of the findings of the ILO economist’s examination of the IMF research papers on labour market reforms:

• The IMF papers are based on an “Economic Freedom of the World Database” published by the conservative Fraser Institute of Vancouver, whose labour regulations component consists of a mix of hard data, composite indices and opinion survey questions, with the mix changing over time. For example, since 2002 the minimum wages sub-component used the World Bank’s “Doing Business” indicator for minimum wages, which is based on a quantitative measure of the legal minimum wage as a proportion of value added per worker. However prior to 2002, the minimum wages sub-component is based on the results of an opinion survey of business executives carried out by the World Economic Forum on their appraisal of the enforcement of the legal minimum wage in their country. These are two completely different notions, but the IMF papers treated them as one and the same.

• Several other breaks in the Fraser data series took place, many of them also in 2002 when the first edition of “Doing Business” was published. The IMF authors found that well over half of 53 “large-scale labour market reforms” said to have taken place between 1980 and 2008 happened in 2002 and appear to be explained by breaks in data series that were concentrated in that year. The IMF authors made no attempt to identify and analyse actual labour market reforms processes; they only looked for substantial changes in the composite indices.

• The indices used by the IMF authors were further flawed by the fact that when data were unavailable for specific sub-components in some years, which they frequently were, they calculated simple averages of those that were available. The ILO economist was able to replicate only 17 out of the 53 “reforms” identified by the IMF authors when excluding missing data for any sub-components.

• Three of the six sub-components of Fraser index are based on the “Doing Business” labour market flexibility indicator (“Employing Workers Index”), which the World Bank suspended in 2009 because of major conceptual flaws and ordered its staff to no longer use in analyses or for formulating policy recommendations. A footnote reveals that, astoundingly, the IMF authors “were not aware … of the sensitivity issues related to the World Bank Employing Workers Index”. The same footnote states that they were also unaware of the methodological breaks in the Fraser data series, even though these are explicitly signalled by the Fraser Institute on its web site.

• Similar to “Doing Business”, especially in early editions that heralded countries having no or minimal labour regulations as “best performers”, the IMF papers show clear bias in favour of labour market deregulation by equating more flexible labour regulations with “better quality”. Aleksynska notes that the IMF papers’ view “assumes linearity of regulations’ impact on labour market outcomes”. She contrasts this with ILO research and the World Bank’s “World Development Report 2013: Jobs”, which found no empirical evidence of a linear relationship between labour regulations and economic outcomes. In the view of these latter works and in contrast to the IMF view, “a ‘high quality’ regulation would not be the one with the lowest level, but the one which balances the need to provide fair treatment and income security to workers with the employment adjustment possibilities of firms”.