What Explains the Behavior of Consumption in May?

BRAZIL ECONOMICS - Report 13 Jul 2020 by Affonso Pastore, Cristina Pinotti, Marcelo Gazzano and Paula Magalhães

In relation to April, real retail sales in the augmented sense (including automobiles and construction materials) in May recovered more than expected, generating a wave of optimism and revision of the projections about the severity of the recession in 2020. However, a detailed analysis suggests caution. The improvement in May was very heterogeneous, both among segments and states. There are two possible explanations for this improvement: a) the income transfers to the most vulnerable people changed the pattern of consumption and increased spending (because lower income means the marginal propensity to consume is near one); and b) the predominant cause came from changes in the intensity of social distancing. For example, although the largest income transfers occurred in states of the North and Northeast regions and the lowest in states of the South and Southeast, sales expanded most in the latter two regions. Through out-of-sample projections based on our model to explain real retail sales, we show that based on: a) real labor income; b) the real interest rate; and c) new credit to households; and omitting income transfers from the model, the performance of real retail sales that effectively occurred would have been much larger (than the case when including the income transfers). Indeed, in April and May there would have only been a small decline in relation to January and February. In other words, the deviation between the projection and what happened can only be due to an omitted variable, which is social distancing. We also show that in the states that achieved the greatest success in social distancing (leading to a lower death rate), growth in sales was stronger from April to May. Everyone is learning how to analyze the economy in a crisis like the present one, so we see no reasons to alter scenarios and projections with each “surprising” number. For this reason, we have not changed our projection for 7.5% contraction of GDP in 2020.

Now read on...

Register to sample a report

Register