Gerald A. Carlino

Are U.S. regional incomes converging?: A time series analysis

Abstract The time series properties of per-capita income in U.S. regions are tested for consistency with the neoclassical growth models prediction of per-capita income convergence. Two conditions are required for convergence. Shocks to relative regional per-capita incomes should be temporary (stochastic convergence), and initially poor regions should catch up to rich regions (β-convergence). We find evidence for stochastic convergence across U.S. regions during the 1929–1990 period after allowing for a trend break in 1946. We also find that U.S. regions have achieved β-convergence. These findings support the neoclassical models prediction of conditional convergence.

THE DIFFERENTIAL REGIONAL EFFECTS OF MONETARY POLICY

This paper examines whether monetary policy has similar effects across regions in the United States. Impulse response functions from an estimated structural vector autoregression reveal a core of regionsNew England, Mideast, Plains, Southeast, and the Far West that respond to monetary policy changes in ways that closely approximate the U.S. average response. Of the three noncore regions, one (Great Lakes) is noticeably more sensitive to monetary policy changes, and two (Southwest and Rocky Mountains) are found to be much less sensitive. A state-level version of the model is estimated and used to provide evidence on the channels for monetary policy.

The Concentration/Conduct Relationship in Bank Deposit Markets

This study investigates the structure/conduct/performance relationship in retail deposit markets. The study explicitly incorporates conduct as the link between structure and performance in local deposit markets. It attempts to determine whether banks typically behave competitively or strategically, and whether their conduct is influenced by market concentration. The empirical investigation is guided by an equilibrium model of a retail deposit market. The model is applied to regression equations for local (MSA) MMDA and three- and six-month CD rates. The empirical results indicate that strategic conduct is the norm in MMDA and in three- and six-month CD markets. Copyright 1991 by MIT Press.

The differential regional effects of monetary policy: evidence from the U.S. States

This paper uses time-series techniques to examine whether monetary policy has similar effects across U.S. states during the 1958-92 period. Impulse response functions from estimated structural vector autoregression models reveal differences in state policy responses, which in some cases are substantial. The paper also provides evidence on the reasons for the measured cross-state differential policy responses. The size of a states response is significantly related to its industry mix, evidence of an interest rate channel for monetary policy. The state-level data offer no support for recently advanced credit-channel theories of the monetary policy transmission mechanism.

Testing neoclassical convergence in regional incomes and earnings

Abstract The time-series properties of per capita income and per capita earnings in the regions of the United States are tested for consistency with the neoclassical growth models prediction of convergence. We find evidence for per capita income convergence for U.S. regions during the 1929–1990 period after allowing for a trend break in 1946. These findings support the neoclassical models prediction of convergence. The evidence for per capita earnings convergence is, however, less conclusive. Shocks to per capita earnings are found to be more persistent than shocks to per capita income. This implies that the regional distribution of transfer payments tends to smooth the effects of deviation on relative regional per capita earnings and reinforce trends in per capita income convergence.

Accounting for differences in aggregate state productivity

Abstract The purpose of this paper is to analyze the determinants of aggregate productivity at the state level by utilizing the new GSP data. State differences in aggregate productivity are modeled as a Hicks-neutral shifter term in an aggregate production function framework. The model attempts to account for variations in aggregate productivity resulting from state-to-state differences in industry mix, human capital characteristics, public investment in infrastructure, and metropolitan structure. In addition, a random-effects extension of the model is employed to capture the effects on state productivity of other important localized factors that are not directly entered into the empirical model. The findings indicates that a states industrial mix, infrastructure, education level, and metropolitan structure all effect productivity. Relative state productivity rankings that allow cross-state comparisons of overall productivity and identify the sources of productivity differences are developed. For comparison, the model is re-estimated using manufacturing data. An important finding is that the productivity rankings based on the aggregate state output differ markedly from rankings based solely on data for the manufacturing sector. This raises questions about making inferences about overall productivity based solely on manufacturing data.

Declining city productivity and the growth of rural regions: A test of alternative explanations

Prior to the 1970s the U.S. witnessed a net flow of migrants from its nonmetropolitan to its metropolitan region. Since then, however, a dramatic reversal of this long-standing tendency for people to concentrate in metropolitan places has been observed-nonmetropolitan places have gained migrants largely as the result of increased outmigration from metropolitan areas. An open question has been to what extent the employment profile between metropolitan and nonmetropolitan places parallels that of population. This paper introduces a data series compiled from County Business Patterns that is used to analyze the employment trends for the first time. The analysis reveals that not only is total employment growing fastest in nonmetropolitan counties, but it also appears that the turnaround in employment growth preceded the reversal in the rate of population increase. Manufacturing, which had faster nonmetropolitan growth rates even during the 195Os, appears to have led in the development of nonmetropolitan counties. This shift of manufacturing has attracted other sectors as well as people to nonmetropolitan places. Part of the recent growth of nonmetropolitan counties represented nothing more than a continuation of suburbanization or spillover of people and jobs which permitted one more outward bound into adjacent nonmetropolitan counties. But there is more to nonmetropolitan development than simply metropolitan spillover. Nonmetropolitan counties far removed from metropolitan ones are also growing faster than metropolitan counties. It is shown that this rural renaissance is part of a general deconcentration

Employment deconcentration: a new perspective on America's postwar urban evolution

In this study we show that during the postwar era the United States experienced a decline in the share of urban employment accounted for by the relatively dense metropolitan areas and a corresponding rise in the share of relatively less dense ones. This trend, which we call employment deconcentration, is distinct from the other well-known regional trend, namely, the postwar movement of jobs and people from the frostbelt to the sunbelt. We also show that deconcentration has been accompanied by a similar trend within metropolitan areas, wherein employment share of the more dense sections of MSAs has declined and that of the less dense sections risen. We provide a general equilibrium model with density-driven congestion costs to suggest an explanation for employment deconcentration.

Aggregate metropolitan employment growth and the deconcentration of metropolitan employment

Abstract In this paper we document that the disparity in employment densities across US metropolitan areas has lessened substantially over the postwar period. To account for this deconcentration of metropolitan employment , we develop a system-of-cities model in which an increase in aggregate metropolitan employment causes congestion costs to increase faster for the more dense metro areas. A calibrated version of the model reveals that the (roughly) two-and-a-half-fold increase in postwar aggregate metropolitan employment implies, by itself, more deconcentration than actually observed. Thus, rising aggregate metropolitan employment is a powerful force favoring deconcentration, although some benefit of greater employment density appears to have partially offset the effects of rising congestion costs for the more dense metro areas.

How strong is co-movement in employment over the business cycle? Evidence from state/sector data

This study measures the extent of co-movement in employment across states and industries at business-cycle frequencies. The strength of co-movement is quantified using the bi-variate and multi-variate measures of cohesion developed in Crous, Forni, and Reichlin (2001). The data indicate that cohesion is generally positive for the state/industry pairs, although the distribution masses around a relatively low value. The results suggest that cohesion has risen over time and that cohesion increases with spatial aggregation. Evidence is presented revealing that the measured degree of co-movement is sensitive to the chosen periodicity of the data and that there is much greater cohesion across states for a given industry than across different industries within a state. An investigation into the sources of cross-state variation in cohesion reveals that important determinants include the strength of input-output linkages within each state, the different effects of monetary policy actions on each states employment, and the degree of industrial diversity within a state. No state-level support is found for Sheas (1996) hypothesis that industries that locate together co-move to a greater extent than do those that are more spatially diffused.