Job Market Paper
Networks, Phillips Curves and Monetary Policy
Online appendix and Supplemental material
Abstract:
I develop an analytical framework for monetary policy in a multi-sector economy with a general input-output network. I derive the Phillips curve and welfare as a function of the underlying production primitives. Building on these results, I characterize (i) the correct definition of aggregate inflation and (ii) how the optimal policy trades off inflation in different sectors, based on the production structure. Correspondingly, I construct two novel inflation indicators. The first yields a well-specified Phillips curve. Consistent with the theory, this index provides a better fit for Phillips curve regressions than conventional consumer price specifications. The second is an optimal policy target, which captures the tradeoff between stabilizing aggregate output and relative output across sectors. Calibrating the model to the U.S. economy, I find that targeting consumer inflation generates a welfare loss of 0.8% of per-period GDP relative to the optimal policy, while targeting the output gap is close to optimal.
Research Papers in Progress
Aggregate vs Cross-section: a Network Approach
Abstract: I model a currency union, with segmented labor markets and a fully general production network. I study local versus aggregate Phillips curves and transfer/spending multipliers. I derive the full matrix of elasticities of regional prices to regional output gaps and transfers as a function of primitives. I argue that cross-sectional regressions which estimate “local” consumer-price Phillips curves are mis-specified, because they ignore heterogeneity in the own effect and cross-regional spillovers. In addi-tion, the relative response of aggregate inflation and output gap depends on the underlying shock, and even conditional on a shock (for example a money supply shock), it cannot be inferred from cross-sectional regressions. I propose an alternative specification for the Phillips curve (based on a price index that weights sectors according to their value added in local consumption) which depends on two parameters only: one governs the response to own output gaps, and the other governs the spillover from foreign output gaps, weighted by net trade flows between the corresponding regions. With this inflation measure the coefficient on the own output gap is the same as the slope of the aggregate Phillips curve. I derive moment conditions that generalize the standard assumptions in cross-sectional studies, whereby aggregate monetary policy is orthogonal to local business cycles.
“Monetary Policy in the Global Economy”
Abstract: I consider a global production network, where every country has multiple sectors that trade in inter-mediate inputs both domestically and internationally. Producers in each country-sector pair can price different shares of their output in different currencies, also conditioning on the destination. I derive the response of local inflation and output to local and foreign shocks to sector-level productivity, monetary policy and exchange rates. I plan on using this framework to revisit traditional questions, such as the effect of competitive devaluations and monetary policy spillovers, and to evaluate the extent to which an increase in import shares (both in production and final consumption) can reduce the sensitivity of inflation to local employment.
A General Equilibrium model of Outsourcing and Wage Inequality
Abstract: A growing literature documents that outsourcing puts downwards pressure on the wage of low-skill workers in instances where it entails a change in the employment relationship (it excludes workers from earning firm premia) without affecting the organization of production. I consider a complemen-tary question: I define outsourcing as a shift from producing “in house” to buying components “on the market”, with the resulting increase in intermediate input flows across industries. I model the in-centives behind it as coming from a tradeoff between demand volatility and specialization on one side, and coordination frictions on the other. With respect to standard models based on attention vs coordination costs or property rights theory my setup allows to rationalize the formation of a full network with cross-industry linkages (not just a vertical chain) and the fact that producers serve mul-tiple customers, without the counterfactual implication of decreasing firm size. The model is tractable enough to derive general equilibrium implications, and I focus on the wage distribution. A fall in out-sourcing costs for a set of tasks is equivalent to a factor-biased increase in the productivity of the worker type performing those tasks. The effect on relative wages across types then depends on whether they are complements or substitutes. I assume that workers cannot learn the tasks that are not specific to their type, while in principle they can be employed on any of the tasks that their group performs. This introduces an inherent substitutability between workers belonging to the same group. If skilled workers have a comparative advantage on products with volatile demand, which are more likely to be outsourced, a fall in outsourcing costs reduces inequality at the bottom end of the distrib-ution and increases inequality at the top end.