Why is it so difficult to measure the effect of the labor reform, in the eyes of candidates?

Labor reform is on the agenda for the 2022 elections. At least two presidential candidates —Lula and Ciro Gomes— talk about the possibility of revoking or revising the change to the CLT (Consolidation of Labor Laws), sanctioned in 2017, during the government of Michel Temer.

With 11.9 million unemployed at the end of the first quarter, 40% of the population employed in the informal sector and a drop of almost 9% in income in the year, Brazilians have a question: what was the effect of the labor reform on employment?

Before implementation, the promises of Temer’s economic team were that the reform — which, among other changes, instituted new hiring modalities — would create between 2 and 6 million jobs in the years following its approval.

Even considering the pandemic and institutional crises since then, there is frustration among the population with an unemployment rate that has remained in double digits for six years – even though the reform has been in place for more than four of them.

However, assessing the impacts of the reform is not a simple task. This is because it is not enough to look at data such as the unemployment rate and income, before and after the reform, to reach any conclusion, as several factors influence these variables and it is not possible to know how the economy would have behaved if the reform had not been implemented. in force.

Two recent studies attempt to circumvent these difficulties.

But there are still many more doubts than certainties about how the reform changed the Brazilian economy and how it could be changed to benefit workers and companies.

New USP study

Gustavo Pereira Serra, a researcher at Made-USP (Center for Research in Macroeconomics of Inequalities at the University of São Paulo), highlights the difficulty of evaluating the effects of labor reform on the Brazilian labor market.

“Since the reform was approved, at the end of 2017, a lot has happened in the Brazilian economy. We had political issues, a change of government, also the economic crisis caused by the Covid-19 pandemic. the impacts specifically of this reform”, says Serra.

As it is difficult, looking only at Brazil, to separate what is a direct effect of the reform, the group of researchers formed by Serra, Ana Bottega and Marina da Silva Sanches selected a sample of 12 countries in Latin America and the Caribbean that did not undergo changes in its labor laws in the analyzed period — which goes from 2003 to 2020.

By this criterion, Paraguay, Uruguay, Argentina and Costa Rica were excluded from the sample, as they underwent some type of change in labor relations in the period. Bahamas, Bolivia, Chile, Colombia, Dominican Republic, Guyana, Mexico, Nicaragua, Saint Lucia, Saint Vincent and the Grenadines and Trinidad and Tobago remained.

Based on the behavior of the unemployment rate in these countries, economists build a “synthetic Brazil”. That is, a projection of how the Brazilian unemployment rate would have behaved if the labor reform had not been approved here.

This model is controlled for a number of other variables, such as GDP growth, inflation, exchange rates and interest rates, as all these indicators influence the unemployment rate.

Using this tactic, economists find that labor reform would have reduced the country’s unemployment rate between 2018 and 2020 by 1 percentage point, on the three-year average.

Then, economists perform a test to see if this variation found is statistically significant.

And here comes the cold water bath.

Applying the same methodology to countries where there was no reform, such as Chile, Guyana and Trinidad and Tobago, economists find even greater differences between the synthetic model and the effective unemployment rate of these countries, which indicates that it cannot be concluded that the difference observed in Brazilian data is a result of the labor reform.

“We cannot say that job creation was greater with the reform than it would have been without the reform”, says Serra.

He points out that a difficulty for the analysis of the Brazilian case is the short time of the reform in force.

This is because studies using a similar methodology, analyzing labor reforms carried out in Argentina, Australia and Germany, found reductions in the unemployment rate of these countries between 1.19 and 3.44 percentage points in 12 years.

“On the one hand, the great expectations that the Temer government had for the reform were definitely not met”, observe Serra, Bottega and Sanches in the study.

“On the other hand, our results also suggest caution to conclude anything about the reform so soon, as we cannot say that it has been the cause of the increases in the unemployment rate in recent years, nor can we predict the effects it will have in the long term. “

Study on a specific point of the reform

Another analysis of the labor reform has had a lot of repercussion in recent weeks, although the study is not yet published and it is not possible, for now, to evaluate the methodology used by the researchers.

The study analyzes only one point of the reform: the rule that transfers the costs of the company’s lawyer to the worker, in case he loses a labor lawsuit in court.

“We created a mathematical model to replicate what Brazil was like exactly before the reform, in 2017”, says Danilo Paula de Souza, a postdoctoral researcher at Insper and one of the authors of the study alongside professors Raphael Corbi, Rafael Xavier Ferreira and Renata Narita, from FEA-USP (Faculty of Economics, Administration, Accounting and Actuarial Science at the University of São Paulo).

“So we do a counterfactual exercise, which is to analyze what this post-reform Brazil would be like, without looking at the data [da economia real]why employment may have gone up or down [na economia real] for various reasons, such as foreign policy issues, etc.”, adds Souza.

The researchers then alter this mathematical model to include the possibility that the worker will have to bear part of the cost of the lawsuit if he loses it.

“Then we redo all the calculations and see how firms and workers would behave in this ‘counterfactual’ world, where we simulate the implementation of this point of reform within the model,” says Corbi, also an author of the study.

“Within the model, this change results in a drop in the number of labor lawsuits of 30%, or 800 thousand lawsuits – very similar to what was actually observed in the Labor Court. And a 1.7 percentage point reduction in unemployment , which represents 1.7 million more jobs”, adds the researcher.

According to him, this means that companies, before the reform, anticipated the potential high cost of dismissal and hired fewer people. With the reform, companies would have greater security in hiring and in terms of dismissal costs and, therefore, would hire more, in the researchers’ view.

Adriana Marcolino, sociologist and technician at Dieese (Inter-Union Department of Statistics and Socioeconomic Studies), views the results of this study with caution.

“I don’t know the methodology of the study, since it is not published, but I believe it is ineffective to take a measure that reduced the number of labor lawsuits and assess that there were more hirings in the job market because of that”, says Marcolino.

The sociologist assesses that it is negative that there has been a reduction in the number of labor lawsuits, not because labor relations have become more efficient, but because, in her view, an obstacle has been created for workers to claim their rights in court.

“We know that in Brazil there is a great lack of compliance with basic issues related to work — payment of wages, working hours, health conditions. Instead of creating a regulation that would solve these problems more effectively, we created mechanisms to inhibit the worker to claim these rights”, he says.

“With that, the employer may have had a cost reduction with lawsuits, but I don’t see how this directly generated jobs, other than a mathematical exercise.”

In October of last year, the Federal Supreme Court (STF) ruled that it was unconstitutional for workers entitled to free justice to pay the fees of loss of suit, as the amounts paid to the lawyers of the winning party in a lawsuit are called.

winners and losers

For economist Claudio Ferraz, professor at the University of British Columbia, Canada, and scientific director of the JPAL (Poverty Action Lab) for Latin America, there are several factors that explain why there are still so few consistent studies on the effects of the reform. labor on the Brazilian economy.

The first problem, in Ferraz’s assessment, is the lack of data. This is because the main source of studies on the formal labor market in Brazil is Rais (Annual Social Information Report), whose microdata were released with a delay of about two years compared to the reference year by the Ministry of Economy, a period that recently became even longer.

In addition, as the reform came into effect at once for all Brazilian workers, it is not possible to compare groups affected and not affected by the change in the law, which is a widely used way to assess the impact of public policies, observes the researcher.

The fact that the reform is quite complex, having affected several aspects of employment relations at the same time, also makes it difficult to analyze the impacts, assesses Ferraz.

In addition to changing the rule in relation to attorney fees, the reform brought several other changes to the law, such as the introduction of intermittent work, the prevalence of the agreement between worker and employee over the legislation and the non-mandatory of union dues.

“It is possible to look at pieces of the reform, as the recent study by Raphael Corbi, Renata Narita and co-authors does, which looks at the reduction in the number of lawsuits in labor courts and how this may have affected companies and the legal cost of firms with lawsuits “, says Ferraz.

“But this is a very small part of the reform, whose big change I consider to be the flexibility of the market in terms of hiring”, he says.

For the economist, a long-time advocate of evidence-based public policies and the use of data for policy evaluation, any changes in labor reform must take into account that different groups may have been affected in different ways by the change in rules.

“Usually, when we evaluate public policies, we look at the average effect, but the labor reform is an example of how we can have ‘winners’ and ‘losers,'” says Ferraz.

He cites the example of women with small children, who may have benefited from the greater flexibility of contracts made possible by the reform. On the other hand, the economist believes that there may be many “losers”, with the advance of subcontracting, outsourcing of jobs and the loss of quality of jobs.

“To say that, by definition, this is a reform that makes everyone’s life worse, I don’t think it’s true. That’s why I get a little scared when people say that everything needs to be reversed, because in fact, we didn’t even evaluate the what happened and the effects are heterogeneous.”

For the researcher, the PT’s initiative to look at what has been done in Spain, with the revision of the labor reform there, is a first step in a process that demands a precise diagnosis before talking about changes.

“The starting point is to look at the similarities between the Brazilian reform and others. Secondly, it is necessary to understand that this is a microeconomic reform, so we need to look at the different sides of the reform, what happened with companies and workers. when we know what happened, then we can think about what policies we can implement.”