This paper examines how naïve present-biased preferences affect the transmission of anticipated neutral total factor productivity (TFP) shocks in a one-sector real business cycle model. Relative to time-consistent preferences, present-bias weakens the immediate response to favorable productivity news and shifts a larger share of adjustment toward the date on which the productivity change is realized. For realized news, present-bias attenuates the pre-realization declines in output and investment, although it does not eliminate the one-sector co-movement problem. When the announced productivity improvement is not realized, consumption reverses and the anticipated expansion in output and investment disappears. These qualitative results remain robust across the reported local parameter range. As supplementary empirical discipline, an IRF-based simulated method of moments exercise compares model and empirical responses to productivity news. The well-conditioned weighting criteria select a present-bias parameter value of 0.97, which is at the lower boundary of the available local solution range. This boundary result is conditional on the selected moments and news-identification strategy and is not interpreted as a precisely identified structural estimate of present-bias.
Will smart (programmed) machines do to humans what the internal combustion engine did to horses—render them obsolete? If so, can putting people out of work or, at least, good work leave them unable to buy what smart machines produce? The answer, as we illustrate in a simple, two-period, two-good life-cycle model, is yes. Our goods are output and prayers. Output is produced with code, new and legacy, and capital. This function is effectively a robot with code for brains and capital for brawn. ”Robots are Us” captures our key assumption that code is the offspring of human brains, specifically that of high-skilled agents. Such agents are free to produce new code or pray (for others). Low-skilled agents only pray. An improvement in code tech, modeled as greater durability, increases current demand for (and supply of) new code. But it decreases future demand. What explains this dynamic reversal of Say’s Law? When code lasts longer, today’s fresh code means even more legacy code tomorrow. But tomorrow’s greater legacy code reduces tomorrow’s need for/demand for new code and, thus, coders. Hence, a code-tech shock generates a boom-bust cycle in the coders’ wage, which, thanks to skilled-worker mobility equals their prayer wage. Output, saving, and investment boom and bust in unison with wages. The investment bust can leave the economy’s capital stock and, potentially, output and lifetime utility below their starting values. Such immiserating growth is more likely when code can, as AI now does, self improve. Whether or not future generations are worse off, the big winners are early generations who ride the boom and avoid the bust. As we discuss, cross-cohort redistribution can share the gains from tech shocks and preclude immiseration. What about limiting IP rights? Surprisingly, doing so makes immiseration more likely. Better tech worsening the economy? This is as counterintuitive as Leamer (1999)’s finding that a minimum wage can raise GDP and lower inequality. Both studies pull rabbits out of hats, which was Edward Leamer’s lifetime practice and delight.
This paper aims to contribute to the literature on early warning systems by assessing the role of bank-specific variables in explaining and predicting systemic banking crises. We employ a multinomial logit regression to address bias in crisis identification and apply the Bayesian Model Averaging (BMA) technique to select potential crisis indicators, thereby reducing uncertainty in developing a credible and robust model. The study uses a sample of 48 developed and developing countries from 1980 to 2015. Our findings highlight the significance of microeconomic variables in explaining financial fragility and crisis duration. These variables are also relevant for predicting the onset of crises, both in-sample and out-of-sample, although they are less important for correctly identifying crises when using BMA to select the optimal combination of early warning indicators. Additionally, the results demonstrate that aggregated bank-specific characteristics are crucial for explaining and predicting banking defaults.
This paper uses a two-stage specification search to study growth miracles in China, Singapore, and South Korea. The first stage estimates a scalar additive-functional model that decomposes per capita GDP into deterministic trend, martingale, and stationary components. Bayesian posterior estimates indicate highly persistent transient growth gaps and substantial volatility. The second stage uses the method of simulated moments to calibrate a stochastic Ramsey growth model that replicates key moments of the estimated statistical model. Specification searches over the long-run growth parameter ν reveal that fixing ν = 0.02 to match US growth leaves the Ramsey model’s distribution too far to the left for China, while estimating ν separately changes the moment targets and improves the in-sample Ramsey fit. Our analysis illustrates how priors and likelihoods jointly determine posterior distributions and inferences about parameters.
This paper addresses two fundamental macroeconomics questions. First, are macro shocks large enough to alter the course of the economy? Second, are they large enough to materially impact economic welfare? Lucas and many others have addressed these issues, but do so primarily in the context of representative agent models. We study these questions using a large-scale, general equilibrium, stochastic, overlapping generations model. We consider 80 generations overlapping in an economy buffeted by realistically calibrated total factor productivity and capital depreciation shocks. The model is solved using Marcet’s projection method taking explicit account of the full state space, which comprises 81 variables. Our findings, some recapitulated from prior studies by Hasanhodzic and Kotlikoff, suggest macro shocks are second order both with respect to their impact on aggregate variables and individual welfare. Specifically, the probability that the stochastic economy’s long-run aggregates materially deviate from their deterministic counterparts is less than one percent. Furthermore, the realized (simulated) lifetime utility of generations born in the long run rarely differs from deterministic long-run utility levels by more than 1 percent, measured as consumption-compensating differentials. These findings are supported by the model’s small equity premium. Moreover, adding a bond market leaves our findings essentially unchanged. Our RBC-in-OLG findings suggest that what really impacts economic welfare is not macro shocks, but government policy, particularly intergenerational redistribution policy. Our paper comprises, in the tradition of Edward Leamer, a specification search, but one involving alternative theory and computation, not alternative estimators. And, like so much of Leamer’s work, it’s iconoclastic. It shows that large-scale, life-cycle macro models can be computed as readily as intergenerationally altruistic representative agent models, which have, regrettably, come to dominate all aspects of macroeconomic research.
