Citation

Partyka SC. Accid. Anal. Prev. 1984; 16(3): 211-222.

Copyright

(Copyright © 1984, Elsevier Publishing)

DOI

unavailable

PMID

unavailable

Abstract

Simple models of traffic fatalities were developed using only the readily available factors of population size (from the Bureau of the Census estimates and projections) and the size of the potential labor force and the number of unemployed workers (from the Bureau of Labor Statistics estimates based upon household surveys). The fatality trends from 1960 through 1982 appear consistent with changes in the numbers of unemployed workers, employed workers, and people not available for the labor force. The long-term (23-yr) model using employment factors produces an estimate of 45,158 fatalities in 1982 versus the current estimate of 44,000 fatalities which was derived from preliminary State reports. Short-term (1975-1982) annual fatality trends reflect recent sociological changes. A model based upon the number of unemployed workers and the number of employed workers produces an estimate of 44,709 fatalities in 1982, as compared to the State-reported estimate of 44,000 fatalities. A similar model of teenage (15-19 yr old) traffic fatalities produces an estimate of 6679 fatalities in 1982, as compared to the current estimate of 6700. The quantification of the relationship between economic factors and the number of traffic fatalities may be useful in understanding the effects of safety programs and other changes to the driving environment. However, the reader is cautioned on three important points about any modeling analysis. First, a model produces estimates of the effects of variables included in the analysis, but does not imply estimates of the effects of variables left out of the analysis. In particular, these models do not imply that other factors (such as safety programs and improvements) are unrelated to fatality decreases. Second, a model describes the observed coinciding of changes among variables, but does not of itself imply cause and effect. Inferences of causal relationships (between, for example, the economy and fatalities via changes in travel patterns) are made by people--usually subject matter specialists--to explain the model. Third, predictions from the model are the responsibility of the analyst making them. Predictions are especially tenuous when based upon data beyond the range of the historical experience used to create the model.