Research Database

This paper estimates the expectations of older male workers in the form of a 130 million element Markov transition probability matrix specifying the joint stochastic process for workers’ income, health, martial and employment status, conditioned on workers’ decisions about labor force participation and collection of Social Security benefits. The estimated transition matrix will be used in subsequent work to estimate the unknown parameters of workers’ utility functions under the assumption that their behavior is governed by the solution to a dynamic programming model. The paper also discusses some of the problems involved in constructing good measures of workers’ states and decisions.

During 1870–1913, Canada had a well-diversified branch banking system while banks in the U.S. unit-banking system were less diversified. Canadian banks could issue large-denomination notes with no restrictions on their backing, while all U.S. currency was essentially an obligation of the U.S. government. Also, experience in the two countries with regard to bank failures and panics was quite different. A general equilibrium business cycle model with endogenous financial intermediation is constructed that captures these historical Canadian and American monetary and banking arrangements as special cases. The model's predictions contradict conventional wisdom about the cyclical effects of banking panics. Support for these predictions is found in aggregate annual time series data for Canada and the United States.

This paper is a study of bank panics under the U.S. National Banking System in 1864–1913. During this period, bank deposits in the United States, like those in Great Britain and Canada, were not insured by the government. Unlike the United States, however, neither of those countries had any bank panics. The U.S. panics were caused essentially by two unique features of the U.S. banking system: prohibitions on bank branching and pyramiding of bank reserves. In the paper, a model which includes these features is constructed, and it is shown that bank panics can occur even though all agents are rational. In this model, bank panics can be eliminated by a combination of reserve requirements, central bank loans, and occasional restrictions on cash payments by banks. The conclusion is that to eliminate bank panics, deposit insurance is not necessary.

The effects of stochastic inflation on equity prices and the equity premium are studied in a pure-endowment asset-pricing model with a cash-in-advance constraint. Stochastic inflation affects the equity premium through two channels: the assessment of an inflation tax and the presence of an inflation premium. Real and monetary versions of the model are simulated and the comparative dynamic results corroborate the conclusion that inflation has quantitatively important effects.

The other important result is that the equity premium in the real version of a model—a continuous state-space generalization of Mehra and Prescott (1985)—and the monetary model is very sensitive to the conditional variance of endowment growth. When the standard deviation of endowment growth is increased from 3.49 percent (the estimated value) to 5.59 percent, the real model can generate an equity premium of 2.8 percent in the range of the risk aversion parameters considered by Mehra and Prescott. The monetary model displays similar sensitivity and can generate an equity premium of 5.81 percent.

For the first-order univariate autoregression without constant term, the joint p.d.f (corresponding to a flat prior) for the true coeffecient p and the least squares estimate p-hat is estimated by Monte Carlo and graphically displayed. The graphs show how the symmetric distribution of p|p-hat coexists with the assymetric distribution of p-hat|p. Treating tail areas of the p-hat|p distribution as if they summarized evidence in the data about the location of p amounts to ignoring the shrinkage in the variance of p-hat|p as p approaches one. Prior p.d.f.'s implicit in treating classical significance levels as if they were Bayesian conditional probabilities are calculated. They are shown to depend sensitively on p-hat and to put substantial probability on p's above one.

This paper reports an optimistic forecast of U.S. output and inflation trends in 1990–91. Generated by a Bayesian vector autoregression (BVAR) model of the U.S. economy using data available on November 30, 1989, the forecast is more optimistic than a consensus forecast. The key to the model's greater optimism for real growth is its outlook for strong consumer spending. The model's optimism is defended by examining historical precedents as well as comparing the track records of the model and consensus forecasts. The model's measures of forecast uncertainty, however, suggest that its predictions should be taken cautiously. An appendix explains how the model can be used to generate conditional forecasts.

This paper describes a method developed to predict the advance (first) estimate of inflation-adjusted gross national product (real GNP) using hours-worked data. Besides generating fairly accurate forecasts of advance GNP, the method has two implications. First, the Commerce Department seems to weigh the hours-worked data most heavily in its early estimates of real GNP but less and less so in its revised estimates. Second, analysts attempting to predict current-quarter outcomes in real time need to consider the availability and reliability of data at the time the forecasts are made.

In this paper I develop a practical extension of the Method of Simulated Moments (MSM) estimator for limited dependent variable models to the panel data case. The method is based on a factorization of the MSM first order condition into transition probabilities, along with the development of a new highly accurate method for simulating these transition probabilities. A series of Monte-Carlo tests show that this MSM estimator performs quite well relative to quadrature-based ML estimators, even when large numbers of quadrature points are employed. The estimator also performs well relative to simulated ML, even when a highly accurate method is used to simulate the choice probabilities. In terms of computational speed, complex panel data models involving random effects and ARMA errors may be estimated via MSM in times similar to those necessary for estimation of simple random effects models via ML-quadrature.

This paper studies the accuracy of two versions of the procedure proposed by Kydland and Prescott (1980, 1982) for approximating the optional decision rules in problems in which the objective fails to be quadratic and the constraints linear. The analysis is carried out in the context of a particular example: a version of the Brock-Mirman (1972) model of optimal economic growth. Although the model is not linear quadratic, its solution can nevertheless be computed with arbitrary accuracy using a variant of the value function iteration procedures described in Bertsekas (1976). I find that the Kydland-Prescott approximate decision rules are very similar to those implied by value function iteration.

The paper discusses a new, fully recursive approach to the adaptive modeling, forecasting and seasonal adjustment of nonstationary economic time-series. The procedure is based around a time variable parameter (TVP) version of the well known “component” or “structural” model. It employs a novel method of sequential spectral decomposition (SSD), based on recursive state-space smoothing, to decompose the series into a number of quasi-orthogonal components. This SSD procedure can be considered as a complete approach to the problem of model identification and estimation, or it can be used as a first step in maximum likelihood estimation. Finally, the paper illustrates the overall adaptive approach by considering a practical example of a UK unemployment series which exhibits marked nonstationarity caused by various economic factors.

Leamer (1983) suggested to study the range of estimators β˅0 in the model y=Xβ + δ when imposing linear constraints of the form M (Cβ - c) = 0 where only C and c are fixed. However the extremes may come from models with a bad R^2, say. In this paper we give the exact bounds when only considering models with R^2 ≥ (1 - δ) R^2 max + δR^2 min. These exact bounds can be found from calculating only two regressions. We apply our techniques to study the velocity of money.