ASSESSMENT OF LAND-USE AND SOCIOECONOMIC FORECASTS IN THE BALTIMORE REGION

Accuracy of forecasts for population, labor force, employment, and car ownership from 1962 to 1975 in the Baltimore area are examined. Comparisons are made at three levels of zonal aggregation-city and suburbs, traffic districts, and traffic zones. The lack of information about household size and household income made inferences from the results incomplete. The results show that regionwide forecasts were accurate for all the variables except population. However, allocation of these forecasts between city and suburbs, to traffic districts, and to traffic zones was quite inaccurate. The correlation coefficient between predicted and actual changes varied from 0.93 to 0.17 for the city zones and from 0.28 to 0.02 for the suburban zones. The corresponding ranges at the traffic-district level were from 0.86 to 0.61 and from 0.36 to 0.30, respectively. The results in the paper point toward large errors and uncertainties in the independent variables of traditional travel-demand models.