Research

This paper explores the optimal design of monetary policy in a multisector model where agents' preferences are non-homothetic. Non-homotheticity derives from the existence of a minimum consumption requirement for food, which households need to satisfy for subsistence. We find that the introduction of a minimum consumption requirement reduces the weight on food inflation in the optimal index that the monetary authority should target. We identify three motives for such  prescription. First, non-homothetic preferences turn the stabilization of food inflation more costly, as it requires larger deviations of output from the efficient level. Second, proximity to the subsistence level turns the demand for food insensitive to monetary policy. Inflation in this sector thus becomes difficult to control. Third, non-homothetic preferences imply that households spend only a small share of any additional income on food. This means that prices in this sector have a reduced impact on aggregate consumption demand. Hence, responding to inflation in this sector becomes less relevant. Importantly, our results provide a rationale for targeting an index that excludes (or attaches a limited weight to) food inflation, a usual practice amongst central bankers.

How does household heterogeneity affect the transmission of an energy price shock? What are the implications for monetary policy? We develop a small open economy TANK model that features labor and an energy import good as production inputs (Gas-TANK). Given complementarities in production inputs, higher energy prices reduce the labor share of total income. Due to borrowing constraints, this translates into a drop in aggregate demand. Higher price flexibility insures firm profits from adverse energy price shocks, further depressing labor income and demand. We illustrate how the transmission of shocks in a RANK versus a TANK depends on the degree of complementarity between energy and labor in production and the degree of price rigidities. Optimal monetary policy is less contractionary in a TANK and can even be expansionary when credit constraints are severe. Finally, the contractionary effect of an energy price shock on demand cannot be generalized to alternate supply shocks, as the specific nature of the supply shock affects how resources are redistributed in the economy.

This paper studies the gains from wage flexibility in a New Keynesian model with price and wage rigidities and incomplete asset markets. When a fraction of households consume solely out of their labor income and have no access to financial markets, the real wage, and therefore, the relative nominal rigidities between wages and prices, directly determine the economy’s aggregate demand. We show that when wages are flexible relative to prices, economic downturns are accompanied by a pronounced decline in real wages, which depresses aggregate demand, and exacerbates the economy’s volatility. In this context, we conclude that enhancing wage flexibility when prices are highly rigid is an undesirable policy prescription.

We examine the measure of inflation that central banks should target given the sectoral composition of the economy and the characteristics of its sectors. We build a New Keynesian model including two sectors—agriculture and non-agriculture—with sticky wages and idiosyncratic productivity shocks. We show that targeting the sector unaffected by the shock curbs wage inflation and minimizes welfare losses arising from wage and price volatility. Consequently, the optimal target assigns reduced weight to sectors with high shock volatility and persistence. This holds regardless of sector sizes and price rigidities. To reduce wage inflation volatility and its transmission to prices, our empirical exercise suggests excluding agricultural inflation from the target in an estimated version of the model using data for Brazil.

We study a New Keynesian model where production inputs and pricing decisions are made under information frictions. Firm production is constrained by inputs that are chosen before shocks are realized, based on firms’ expectations of future demand. We show that the assumption of real rigidities versus nominal rigidities is not innocuous, as assuming the presence of either or both affects the passthrough of demand shocks to aggregate output and inflation. When the choice of production inputs is made under imperfect information about demand shocks, the impact on inflation is amplified while the impact on output is dampened. When both production inputs and pricing decisions are made under imperfect information about demand shocks, the pass-through to output is amplified while the impact on inflation is dampened. Additionally, we show that expectations about demand can behave similarly to a supply shock, as these expectations influence the natural level of output and enter the New Keynesian Phillips curve in a manner analogous to a cost-push shock.