Knowledge that Transforms

This paper addresses the issue of credibility from an econometric perspective. It develops a rational expectations model of inflation in which the dynamics are driven by the level of government spending and by the effect of past inflation rates on the value of real taxes. Government expenditure is modeled as an exogenous autoregressive process subject to discrete changes in regime. The regimes are defined by whether the level of spending is or is not consistent with the rate of inflation targeted by the government as part of a stabilization program. In making their money demand decision, the agents need to construct probability inferences regarding the state of the expenditure process. Credibility is quantified by the agents' inferred probability that the joint observation of inflation, the nominal interest rate, and government spending is generated by the reformed expenditure regime. In an application to Israel, results indicate that the failed stabilization program of November 1984 was less than fully credible to the agents. The uncertainty about the true nature of the expenditure process partially explains the volatility of the rate of inflation in this period. In contrast, for the July 1985 program the agents correctly inferred a change in the regime driving the government spending process.

We present a theory of rational behavior in which individuals maximize a set of stable preferences over goods with unknown addictive power. The theory is based on three fundamental postulates: that consumption of the addictive good is not equally harmful to all, that individuals possess subjective beliefs concerning this harm, and that beliefs are optimally undated with information gained through consumption. Although individual actions are optimal and dynamically consistent, addicts regret their past consumption decisions and regret their initial assessment of the potential harm of the good. Addict-prone individuals who believe "it could not happen to them" are most likely to be drawn into a harmful addiction.

This paper examines the argument presented in The Bell Curve. A central argument is that one factor--g--accounts for correlation across test scores and performance in society. Another central argument is that g cannot be manipulated. These arguments are combined to claim that social policies designed to improve social performance cannot be effective. A reanalysis of the evidence contradicts this story. The factors that explain wages receive different weights than the factors that explain test scores. More than g is required to explain either. Other factors besides g contribute to social performance, and they can be manipulated.

It is well known that under the assumptions of constant returns to scale, perfect competition, and the absence of factor hoarding, primal and dual productivity measures should be highly correlated. The apparent lack of correlation is usually attributed to fixed factors of production. In this paper I propose an alternative explanation by relaxing the assumption of perfect competition. By controlling for the presence of a markup component, I demonstrate that both productivity measures are in fact highly correlated for U.S. manufacturing. The analysis also provides an alternative method of estimating a markup of prices over marginal cost that avoids certain difficulties inherent in some existing methods of estimation.

When banks act as delegated monitors of borrowing firms in a finite economy, two factors help banks dominate direct lending: portfolio diversification, which increases with bank size, and bank capitalization, which diminishes with size. With free entry into banking, intermediated equilibria are possible even when direct lending cannot overcome autarky. There are usually multiple intermediated equilibria; these may not be Pareto-ranked by bank size, since smaller banks are better captialized and may Pareto-dominate larger banks. Even when one large bank would be most efficient, assigning a monopoly bank charter to coordinate beliefs on the single bank equilibrium may be unattractive: in some cases, a monopoly bank cannot overcome autarky even though free-entry banking can, and in other cases, the monopoly bank reduces production from the direct lending level. Journal of Economic Literature Classification Numbers: G21, L13, O16.

This paper examines cross-subsidy, moral hazard, and bank liability issues related to the provision of federal deposit insurance by "rerunning" its implementation, i.e., determining fair premium values, over the period 1927-1932. The pre-1933 period was characterized by historically high asset-price volatility, a large number of bank failures, and a weak federal safety net. In this economic context, we find a high degree of self-insurance on the part of the banks in our sample, both in terms of higher overall capital levels and a strong correlation between capital levels and asset volatility. Potentially large, regional cross-subsidies among banks were also found. Journal of Economic Literature Classification Number: G21.

We show that the pattern of short-term negative serial covariances for stock returns over different return measurement intervals is consistent with the implications of inventory-based microstructure models. We develop different testable implications of these models and document supporting evidence. Our findings indicate that to a large extent the short-horizon return revearsals can be explained by dealer-inventory-related market microstructure effects. Journal of Economic Literature Classification Numbers: G14, G20.

Controversial econometric studies of mortgage data show that mortgage loan applications by some minorities are denied more frequently than are applications by whites with similar observable default risk factors. But recent evidence indicates that minority borrowers also default more frequently than whites with similar observable risk. This paper presents a simple equilibrium model of discriminatory credit rationing and finds parametric restrictions consistent with both these empirical findings. However, in this model, proposed antidiscrimination policies have surprising side effects. Thus, policy analysts accepting this empirical evidence should not expect to derive model-free conclusions about the effects of proposed policies. Journal of Economic Literature Classification Numbers: G21, G28, D63.

I show that dual trading reduces the net order flow and market depth. Trading volume and gross (of commission fees) profits of informed traders are lower with dual trading, while trading volume and gross losses of uninformed traders are unaffected. When the broker′s commission income is independent of the customer′s trading volume, the competitive commission fee is lower with dual trading. The utility of uninformed traders (net of commission fees) increases with dual trading, while the net profits of informed traders decrease. Journal of Economic Literature Classification Numbers: G12, G13, D82.

We develop an integrated model in which a risk-neutral informed trader optimally chooses any combination of a market buy, a market sell, a limit buy including the limit buy price, and a limit sell including the limit sell price. Limit orders undercut the market maker and generate transactions inside the bid-ask spread. The informed trader exploits limit orders by submitting market orders even when the terminal value is inside the spread. When the terminal value is above the bid, a combined market buy-limit sell is more profitable than a market buy only. We obtain an analytic solution. Journal of Economic Literature Classification Numbers: D40, D82, G12, G14.