July 20, 2016
By Paul Beaudry, Dana Galizia and Franck Portier
There is a long tradition in macroeconomics suggesting that market imperfections may explain why economies repeatedly go through periods of booms and busts. This idea can be captured mathematically as a limit cycle. In this paper we present both a general structure and a particular model with the aim of giving new life to this mostly dismissed view of fluctuations. We begin by showing why and when models with strategic complementarities can give rise to unique-equilibrium dynamics characterized by a limit cycle. We then develop a fully-specified dynamic general equilibrium model that embeds a demand complementarity that allows for a limit cycle. Booms and busts arise endogenously in our setting because agents want to concentrate their purchases of goods at times when purchases by others are high, since in such situations unemployment is low and therefore taking on debt is perceived as being less risky. A key feature of our approach is that we allow limit-cycle forces to compete with exogenous disturbances in explaining the data. Our estimation results indicate that US business cycle fluctuations in employment and output can be well explained by endogenous demand-driven cycles buffeted by technological disturbances that render those fluctuations irregular.
This is going to be a controversial paper because it revisits theories that have been discredited, sometimes with choice words. Beaudry and Portier have successful in revisiting old theories or bringing distinct strands of literature together. We’ll whether this on does as well.
July 12, 2016
The last NEP-DGE report has two very interesting papers on frictions in the business cycle. I could not bring myself to feature only, so here are both. The first is interesting in that it can account for the movement of both the size and quantity of asset liquidity in the market through a cycle, the second in that it shows that the costless vacancy creation hypothesis in a typical labor search model has important implications, especially if you want to account for long recoveries.
Search-based endogenous asset liquidity and the macroeconomy
By Wei Cui and Sören Radde
We endogenize asset liquidity in a dynamic general equilibrium model with search frictions on asset markets. In the model, asset liquidity is tantamount to the ease of issuance and resaleability of private financial claims, which is driven by investors’ participation on the search market. Limited market liquidity of private claims creates a role for liquid assets, such as government bonds or at money, to ease financing constraints. We show that endogenising liquidity is essential to generate positive comovement between asset (re)saleability and asset prices. When the capacity of the asset market to channel funds to entrepreneurs deteriorates, investment falls while the hedging value of liquid assets increases, driving up liquidity premia. Our model, thus, demonstrates that shocks to the cost of financial intermediation can be an important source of flight-to-liquidity dynamics and macroeconomic fluctuations, matching key business cycle characteristics of the U.S. economy.
The slow job recovery in a macro model of search and recruiting intensity
By Sylvain Leduc and Zheng Liu
Despite steady declines in the unemployment rate and increases in the job openings rate after the Great Recession, the hiring rate in the United States has lagged behind. Significant gaps remain between the actual job filling and finding rates and those predicted from the standard labor search model. To examine the forces behind the slow job recovery, we generalize the standard model to incorporate endogenous variations in search intensity and recruiting intensity. Our model features a vacancy creation cost, which implies that firms rely on variations in both the number of vacancies and recruiting intensity to respond to aggregate shocks, in contrast to the textbook model with costless vacancy creation and thus constant recruiting intensity. Cyclical variations in search and recruiting intensity drive a wedge into the matching function even absent exogenous changes in match efficiency. Our estimated model suggests that fluctuations in search and recruiting intensity help substantially bridge the gap between the actual and model-predicted job filling and finding rates in the aftermath of the Great Recession.
July 8, 2016
By Daniel Borowczyk-Martins and Étienne Lalé
Employed individuals in the U.S. are increasingly more likely to work part-time involuntarily than to be unemployed. Spells of involuntary part-time work are different from unemployment spells: a full-time worker who takes on a part-time job suffers an earnings loss while remaining employed, and is unlikely to receive income compensation from publicly-provided insurance programs.We analyze these differences through the lens of an incomplete-market, job-search model featuring unemployment risk alongside an additional risk of involuntary part-time employment. A calibration of the model consistent with U.S. institutions and labor-market dynamics shows that involuntary part-time work generates lower welfare losses relative to unemployment. This finding relies critically on the much higher probability to return to full-time employment from part-time work. We interpret it as a premium in access to full-time work faced by involuntary part-time workers, and use our model to tabulate its value in consumption-equivalent units.
This paper is a bigger deal than it seems, and they is actual debate about this topic. It is matters quite a bit. People may not like it ti be working part-time, but the continuation value is so much higher than under unemployment. Of course, this result may not translate to other labor markets with different employment probabilities and different income support schemes.
July 6, 2016
By Michael Keane and Nada Wasi
In this paper we specify and estimate a life-cycle labour supply model that expands on earlier work by simultaneously including human capital accumulation, saving and bequests, an active extensive margin, a realistic specification of the Social Security system, an accounting for private pensions and health expenditures, and a realistic specification of the progressive tax structure. By accounting for all these features, we develop new insights into how taxes affect life-cycle labour supply. For instance, we find that labour supply elasticities vary in important ways with age, education and the tax structure itself. We also show how human capital affects elasticities on the intensive vs. extensive margins.
This is a nice paper with a very rich model. I hope it is used to answer some policy questions, which is where I see its real potential.
July 1, 2016
By Roozbeh Hosseini and Ali Shourideh
We study policy reforms aimed at overhauling retirement financing. We develop a novel approach by considering optimal reforms: policy reforms that minimize the cost for the government while respecting the distribution of welfare in the economy. Our model is an OLG model with life-cycle features and bequest motives where individuals are heterogeneous in their earning ability and mortality. Theoretically, we show that due to the negative correlation between earnings ability and mortality, postretirement distortions to saving decisions are a robust feature of any optimal policy. We, then, use this framework to quantitatively analyze optimal reforms. Our quantitative exercise shows that an optimal reform relative to the status-quo must have three key features: First, post-retirement assets must be subsidized while bequests must be taxed. On average, optimal marginal subsidies on assets for individuals above age 65 is 3.2 percent, while optimal marginal tax on their bequest is 60 percent. Second, pre-retirement transfers must increase while social security benefits must become less generous in the aggregate and more progressive towards low income groups. Finally, earnings tax reform does not contribute to optimal reforms, i.e., optimal marginal taxes on earnings remain very close to the status-quo. The optimal policies reduce the present discounted value of net tax and transfers to each generation by 15 percent.
I find it frustrating that despite ample research that shows that pension reforms are possible, not much is happening. This is yet another paper that shows that these reforms are not just a pie in the sky from economists, they are politically feasible: They are a Pareto improvement. Let us see this whether this one sticks somewhere.
June 24, 2016
By Koichi Miyazaki
The cost of attaining higher education is growing in some developed countries. More young people borrow larger amounts than before to finance their higher education. Several media reports indicate that student loans might affect young people’s decision making regarding important life events such as marriage, childbirth, purchasing a house, and so on. Specifically, this paper focuses on how the burden of student loans affects young people’s decision making with regard to the number of children to have, and studies the fertility rate, gross domestic product (GDP) growth rate, and growth rate of GDP per capita using a three-period overlapping generations model. A young agent needs to borrow to accumulate his/her human capital, although for some reason, s/he faces the borrowing constraint. In the next period, the agent repays his/her debt as well as determines the number of children to have. Under this setting, this paper analyzes how the tightness of the borrowing constraints affects the growth rates of the population, GDP, and GDP per capita. The paper finds that when rearing children is time-consuming, the population growth rate decreases as the borrowing constraints are relaxed. Moreover, the paper shows a case in which the GDP growth rate decreases as the borrowing constraints are relaxed, whereas the growth rate of GDP per capita still increases. In addition, I show that if the cost of rearing children is mainly monetary, then the population growth rate is not necessarily decreasing as the borrowing constraints are relaxed. The paper also calibrates the model using U.S. data.
Very interesting question. The paper is a nice exploratory work that shows what could matter. It has a lot of potential for further work, in particular extending the model to more than three periods, so that a richer life-cycle profile can be taken into account, including the other big choices mentioned in the abstract.
June 16, 2016
By Victor Aguirregabiria and Arvind Magesan
This paper extends the Euler Equation (EE) representation of dynamic decision problems to a general class of discrete choice models and shows that the advantages of this approach apply not only to the estimation of structural parameters but also to the computation of a solution and to the evaluation of counterfactual experiments. We use a choice probabilities representation of the discrete decision problem to derive marginal conditions of optimality with the same features as the standard EEs in continuous decision problems. These EEs imply a fixed point mapping in the space of conditional choice values, that we denote the Euler equation-value (EE-value) operator. We show that, in contrast to Euler equation operators in continuous decision models, this operator is a contraction. We present numerical examples that illustrate how solving the model by iterating in the EE-value mapping implies substantial computational savings relative to iterating in the Bellman equation (that requires a much larger number of iterations) or in the policy function (that involves a costly valuation step). We define a sample version of the EE-value operator and use it to construct a sequence of consistent estimators of the structural parameters, and to evaluate counterfactual experiments. The computational cost of evaluating this sample-based EE-value operator increases linearly with sample size, and provides an unbiased (in finite samples) and consistent estimator the counterfactual. As such there is no curse of dimensionality in the consistent estimation of the model and in the evaluation of counterfactual experiments. We illustrate the computational gains of our methods using several Monte Carlo experiments.
Interesting approach that is not obvious at first sight. Indeed, why iterate on first order conditions in a discrete choice setup? But it works well.