Itaú BBA - Unemployment approaching a peak in Brazil

Macro Vision

< Back

Unemployment approaching a peak in Brazil

July 27, 2017

We have revised the trajectory of the unemployment rate in our scenario

Please open the attached pdf to read the full report 

• The stability of the unemployment rate since January was partly a function of an increase in the employment in the informal sector, which is expected to lose strength ahead.

• Still, recent developments reinforce the assessment that some types of occupation are less strongly correlated with economic activity.

• Using models that take into account the sensitivities of different types of occupation to economic activity, we have revised the trajectory of the unemployment rate in our scenario. The new trajectory anticipates peak unemployment (from 3Q18 to 1Q18) and lowers the upper limit of the unemployment rate (from 14.3% to 13.4%, in seasonally adjusted data).

The decline in the formal employment offset by increases in other occupations

The unemployment rate was virtually flat in 2017, hovering between 12.9% and 13.2% according to our seasonal adjustments. This stability contrasts with the 3.1-pp increase observed in 2016 (see left chart below). The April and May figures[1] differed from our forecasts of a continuous upward trend.

The recent stability in the unemployment rate is largely a consequence of a rise in the employment in the informal sector, which has offset an increase in the working age population and a drop in the formal employment (see chart on the right). See the appendix to this report for a breakdown of the employment by type of occupation.

Excluding formal jobs in the private sector, employment is only loosely correlated with economic activity

The continuous PNAD (available since March 2012) shows a steep increase in formal employment in the private sector until mid-2014, and a decline since then (in line with the recession). On the other hand, the remainder of employment showed an upward trend, hovering around trend in the same period.

In order to assess performance before March 2012, we extended the series retroactively based on the former monthly employment survey, the CAGED and the annual PNAD. Formal employment in the private sector rose steadily throughout most of thepre-2012period, with a brief stagnation due to the economic downturn at the end of 2008.  

Quarterly models show that GDP has strong explanatory power for formal employment in the private sector (table below). The estimate including only the official sample of the continuous PNAD (since 2012) suggests a looser relationship in the short term, but a strong co-integration structure nevertheless. Recent performance is consistent with the models.

An equivalent specification is unable to forecast the remainder of employment, both in the official sample and in the expanded sample (table below).

An alternative specification for the employment excluding formal jobs in the private sector provides better results. We used three factors in the model (see table below): i) the growth of the economically active population (EAP) itself, to take into account the effect of organic growth on the labor force through types of occupation that are less correlated with the economic cycle; ii) the increase of the employment in the formal sector as a proportion of the EAP, to take into account oscillations in labor market formalization; and finally, iii)GDP growth.

The econometric results help us map the factors behind the recent increase of employment in the informal sector.

The first factor is the transition to informal employment of workers who lost their jobs in the formal sector. The negative coefficient of the increase in the formal sector employment is consistent with the advance of labor market formalization between 2003 and 2011, and with the increase in the employment components associated with informal work in the current scenario of formal job destruction.

Secondly, the coefficient associated with EAP growth is more statistically significant than GDP growth. Therefore, organic EAP growth could have an influence regardless of economic activity. In addition, the recession leads to a drop in real household income, which leads to the entry of additional family members into the labor market.[2] It is reasonable to assume that such additional participants are more predisposed toward self-employment or paid work in the informal sector.

We forecast that the unemployment rate will hit 13.4% in 1Q18 and recede slowly thereafter

Using models that show the sensitivity of different types of occupation to activity, we revised the trajectory of the unemployment rate in our scenario (GDP growth rates of 0.3% in 2017 and 2.7% in 2018).

The new trajectory anticipates that unemployment will peak between around 1Q18 (previously: 3Q18). It also lowers the upper limit of the unemployment rate from 14.3% to 13.4% (in seasonally adjusted data), a much less pronounced gain compared with the level posted in May (13.1%).

Finally, the newly approved labor reform will have a positive impact on labor market efficiency, and may affect the equilibrium unemployment rate in the long run. However, cyclical factors related to the pace of resumption of economic activity will be the main determinants of the evolution of the unemployment rate in the coming quarters.


Artur Manoel Passos


[1]  All continuous PNAD data are disclosed in three-month moving averages.

[2]  This is one of the effects mentioned in “Oferta de Trabalho e Ciclo Econômico: os Efeitos Trabalhador Adicional e Desalento no Brasil”, by Gustavo Gonzaga and Mauricio Cortez Reis(RBE v. 55,n.2, 2011). The increase in the participation rate since 2014 suggests that the additional worker/necessity effect was stronger than the discouragement effect in the current recession.

Please open the attached pdf to read the full report



< Back