EconSouth (Second Quarter 1999)
Research Notes highlights some of the research recently published by the Federal Reserve Bank of Atlanta. For complete text of these articles or papers on this Web site, see the links below.
Economic models can shed light on crime
Polls identify crime as the number one public worry. Crime also exacts tremendous costs not factored into official measures of well-being, and it is a favorite subject of politicians' campaign promises. But the public seems largely unaware that crime responds to economic conditions and incentives and that a number of studies by economists have important implications for public policies related to crime.
In a recent article, Zsolt Becsi, an Atlanta Fed economist, introduces the economics and crime literature by describing a simple supply-and-demand crime model in which criminals supply crime, the public demands protection from crime and the government provides public protection. He uses the model to show how crime responds to a variety of demographic and economic factors and what results to expect from public policy proposals.
Using state data from 1971 to 1994, the author outlines broad regional differences and trends in the patterns of crime in the United States. While in the 1990s crime has fallen dramatically nationwide in almost all categories, not all regions have benefited equally. In particular, southeastern states have seen distinct worsening in crime rates relative to other regions.
The author points out some measurement problems with crime data, such as underreporting and undercounting of crimes, that must be considered when interpreting these data. To help control for such problems, he focuses part of his analysis on data categories that are less subject to these measurement problems.
Becsi also cautions that correlations between crime and possible explanatory factors can be distorted by the aggregation of data over time. To deal with this problem, he subjects the data he analyzes to a panel regression approach that estimates the effects of demographic and economic variables on crime over time.
Becsi's analysis suggests several policies that could help alleviate crime, including providing unemployment insurance, increasing welfare expenditures and increasing imprisonment rates.
The results of his analysis mirror those of some other researchers. Generally, the demographic and economic variables explain crime rather well, and estimates for the most part conform to the economic model of crime.
Former Fed president leaves a legacy for monetary policy debate
Monetary policy was freed from the straightjacket of pegging U.S. Treasury interest rates following the Treasury-Federal Reserve Accord in 1951. This newfound freedom led to a growing debate inside and outside the Federal Reserve System about the appropriate measures to use as operating guides. A recent article by Atlanta Fed visiting scholar R.W. Hafer examines the contributions of Malcolm Bryan, president of the Atlanta Fed from 1951 through 1965, to this debate and to the evolution of monetary policy since the 1951 accord.
Bryan parted company with most of his colleagues on the Federal Open Market Committee (FOMC) by trying to steer policy away from a focus on interest rates and money market conditions to placing more weight on the monetary aggregates. Hafer reviews the transcripts of the committee meetings during Bryan's tenure, which reveal Bryan's concerns about preventing the disruptive effects of short-run fluctuations in money growth and the longer-term effects of expansive Fed actions — namely, inflation.
Bryan's approach to monetary policy was a dramatic departure from the prevailing views of the FOMC at the time. He took the new and controversial research results coming out of monetary economics and tried to implement them. His contributions went beyond merely implementing others' ideas, however. His introduction of short-run aggregate growth targets — growth cones — stands out as a significant and innovative development in U.S. monetary policy. Even though Bryan's targets and procedures were not adopted by the FOMC at the time, his strategy for monetary policy would resurface as the inflation produced by the policies he fought against became unacceptable.
Hafer also notes that Bryan's contribution is important from a larger perspective. Bryan's willingness to advocate a controversial view within the FOMC promoted an airing of diverse views and concerns that helped foster an environment in which alternative theories and approaches to economic analysis could be used for improving monetary policy.
A simple model provides lessons on practical forecasting issues
Constructing forecasts of the future path for economic series such as real gross domestic product growth, inflation and unemployment is vital to applied economic analysis for business and government. Model-based forecasts are easier for independent researchers to replicate and validate than forecasts in which expert opinion is used to adjust a model's results. In addition, the forecaster can formally investigate the source of systematic errors in model forecasts, and a forecast model's performance can be established before it is used by a decision maker.
A recent article by John C. Robertson and Ellis W. Tallman, senior economists at the Atlanta Fed, describes a particular model-based forecasting approach. The model studied is a vector autoregression (VAR) comprising six U.S. macroeconomic variables. The idea underlying forecasting with a VAR model is first to summarize the dynamic correlation patterns among observed data series and then use this summary to predict likely future values for each series.
Robertson and Tallman discuss various methods that attempt to improve VAR forecast accuracy and then provide some empirical evidence that compares the accuracy of forecasts created using these various techniques.
The authors also focus particular attention on the technical hurdles that must be addressed in a real-time application of their model and methods for overcoming those hurdles. The solutions to these technical problems include conditional forecasting to handle the staggered release of data and matching quarterly with monthly data.
In emphasizing the practical problems of forecasting economic data using a statistical model, Robertson and Tallman draw on experience in using such a model at the Federal Reserve Bank of Atlanta. Although the model studied is small and highly aggregated, it provides a convenient framework for illustrating several practical forecasting issues. The focus on a simple model provides potential users with a road map of how one might implement such a forecasting model in specific applications.