Approaches to modelling overdispersion in the analysis of migration

"In this paper the modelling of overdispersion in generalised Poisson and multinomial models of migration flows and rates is considered, and its importance within the wider question of substantive model specification is shown. It is argued that substantive specification and the modelling of overdispersion are closely interrelated. Simplified ways of estimating the form of overdispersion--moments methods and pseudo-likelihood--are considered wherever possible. Overdispersion is set within the broader context of correlation effects which relate to migration--correlation across different destinations, between spatially adjacent origins and destinations, and between different time periods. The data used to illustrate the modelling development [concern England and Wales and] are drawn from the Office of Population Censuses and Surveys Longitudinal Study, the decennial Census, and the National Health Service Central Register."     

What do you want from the 2001 census? Results of an ESRC/JISC survey of user views

"The author describes the results of a survey of user views about the next Census of Population in the United Kingdom, to be held in 2001. Some 140 respondents reported their views, which included strong support for a question on income, endorsement of the new one number census methodology and support for postcode-based outputs. The author sets these views in the context of the Census Development Programme being carried out by the U.K. Census Offices and the proposals for outputs which are being discussed with the main census user sectors of central and local government, business, the Health Service, and the academic community."     

Comparison of ward population estimates from FHSA patient registers with the 1991 Census

"Postcode addresses from National Health Service patient registers for Norfolk and Suffolk [counties in England] current on census day 1991 were assigned to census wards, and estimates of populations in wards were produced for the total population and for twelve age-sex groups. These were compared with adjusted counts of usual residents from the 1991 Census." The results indicate that family health service authority registers "are an acceptable alternative to the census for population estimation purposes. This supports recent arguments for wider use of population registers and suggests that they may be particularly valuable as a source of intercensal information."     

Migration models incorporating interdependence of movers

"In this paper we describe a development of the Poisson model in which households are assumed to migrate independently, but the expected number of migrants is given by a Poisson distribution generalized by an empirically given household-size distribution.... The model is fitted to data on short-distance migration within the English county of Hereford and Worcester.... The data set used comes from the Special Migration Statistics (series II) produced as part of the 1981 British Census." The sparse nature of the data raises problems in assessing goodness of fit, because the deviance value is unusually low. This is tackled using simulation methodology.     

Research policy and review 30. How to add value to migration data from the 1991 census

"The form in which migration information is likely to be provided from the 1991 Census [of the United Kingdom] is reviewed, and suggestions are made about how this provision can be improved for the academic research community, building on the experience of the 1981 Census. The recommendations add considerable value to the information at relatively low cost."     

The migration response to the economic turnaround in New England

"This paper is an investigation of the dynamics of interstate migration flows to and from New England since 1975. The main goal of the research is to study the timing and volume of the lagged migration adjustment to the regional economic turnaround and the temporal stability of patterns of regional inflows and outflows. The paper is an examination of the responsiveness of interstate migration to changing regional economic conditions based on annual interstate migration data and a set of cross-sectional destination-specific Poisson regression models." The data are from a series developed by the Bureau of the Census based on Federal income tax returns.     

PRIMARY, RETURN AND ONWARD MIGRATION IN THE U.S. AND CANADA: IS THERE A DIFFERENCE?

"Using microdata sources available from the U.S. Census Bureau and Statistics Canada, this paper contrasts primary, return and onward migration in the two countries. These classifications are based on information on the region (state or province) of birth and region of residence at the start and end of the census period. To study the propensities to make a primary, return or onward migration, rates are calculated...and compared with previous census periods. Analysis reveals that both countries have experienced similar temporal patterns, and that primary, return and onward migration patterns tend to be similar. Analysis using a nested logit model further reveals that return and onward migration in both countries can be similarly explained." Data are from the Public Use Micro Sample (PUMS) for the United States and the Public Use Micro File (PUMF) for Canada.     

Functional forms and the relevance of contiguous migration in the study of migration and employment growth

Previous studies on interstate migration have dealt with total migration without decomposing it into contiguous and noncontiguous migration flows. Since migration over borders of contiguous states may merely reflect changes in residence without changes in economic activities, the use of total migration data might reduce the reliability of empirical results. In this study, a simultaneous-equation model containing two equations, one for migration and the other for employment growth, has been specified in a general form and estimated by the 2SLS method for total, contiguous and noncontiguous migration flows. Results obtained from the use of 1970 Census data show that noncontiguous migration behaves significantly different from contiguous migration, and that noncontiguous migration, rather than total migration, should be used for the study of factors affecting interstate migration. Results also indicate that the log-linear functional form commonly used in empirical studies cannot be accepted statistically.     

Atmospheric nitrogen in the Mississippi River Basin--emissions, deposition and transport

Atmospheric deposition of nitrogen has been cited as a major factor in the nitrogen saturation of forests in the north-eastern United States and as a contributor to the eutrophication of coastal waters, including the Gulf of Mexico near the mouth of the Mississippi River. Sources of nitrogen emissions and the resulting spatial patterns of nitrogen deposition within the Mississippi River Basin, however, have not been fully documented. An assessment of atmospheric nitrogen in the Mississippi River Basin was therefore conducted in 1998-1999 to: (1) evaluate the forms in which nitrogen is deposited from the atmosphere; (2) quantify the spatial distribution of atmospheric nitrogen deposition throughout the basin; and (3) relate locations of emission sources to spatial deposition patterns to evaluate atmospheric transport. Deposition data collected through the NADP/NTN (National Atmospheric Deposition Program/National Trends Network) and CASTNet (Clean Air Status and Trends Network) were used for this analysis. NOx Tier 1 emission data by county was obtained for 1992 from the US Environmental Protection Agency (Emissions Trends Viewer CD, 1985-1995, version 1.0, September 1996) and NH3 emissions data was derived from the 1992 Census of Agriculture (US Department of Commerce. Census of Agriculture, US Summary and County Level Data, US Department of Commerce, Bureau of the Census. Geographic Area series, 1995:1b) or the National Agricultural Statistics Service (US Department of Agriculture. National Agricultural Statistics Service Historical Data. Accessed 7/98 at URL, 1998. http://www.usda.gov/nass/pubs/hisdata++ +.htm). The highest rates of wet deposition of NO3- were in the north-eastern part of the basin, downwind of electric utility plants and urban areas, whereas the highest rates of wet deposition of NH4+ were in Iowa, near the center of intensive agricultural activities in the Midwest. The lowest rates of atmospheric nitrogen deposition were on the western (windward) side of the basin, which suggests that most of the nitrogen deposited within the basin is derived from internal sources. Atmospheric transport eastward across the basin boundary is greater for NO3- than NH4+, but a significant amount of NH4+ is likely to be transported out of the basin through the formation of (NH4)2SO4 and NH4NO3 particles--a process that greatly increases the atmospheric residence time of NH4+. This process is also a likely factor in the atmospheric transport of nitrogen from the Midwest to upland forest regions in the North-East, such as the western Adirondack region of New York, where NH4+ constitutes 38% of the total wet deposition of N.     

Monitoring intercensal migration in the United Kingdom

"This author demonstrates the utility of the National Health Service Central Register of patient reregistrations for providing continuous information for monitoring changes over time in migration behaviour within the United Kingdom. An information system has been constructed that contains annual time-series data with which to illustrate trends in the volume, composition, and geographical distribution of migration during the 1980s. Issues of data reliability and system expansion are amongst those discussed."     

Analysis of alternative labor-force population migration forecasting models

This study evaluates the ability of a model using economic and demographic variables to predict migration patterns. The regression model tested is based on an analysis of migration hypotheses, operationalized on a past time period, and evaluated with the method used by the Bureau of the Census. Dollar value estimates using recent Federal revenue sharing allocations are made with the alternative forecasting methods for one State in each of the Bureau of the Census regions. The results show that the model that is developed provides a firmer basis for projection confidence than does one which relies solely on trend extrapolation.The authors would like to thank Hugh Folk and Ken Cogger for several helpful comments. This paper was supported in part by a grant from the Employment and Training Administration, U. S. Department of Labor.     

The importance of place of birth in migration and population redistribution analysis

"In this paper, U.S. Census data is used to examine further the importance of place of birth on migration streams and to trace the impacts of such 'native dependence' on age patterns of migration, multiregional life expectancies, and spatial population projections."     

Short-term forecasting of internal migration

A new methodological approach to the forecasting of short-term trends in internal migration in the United States is introduced. "Panel-data (or longitudinal-data) models are used to represent the relationship between destination-specific out-migration and several explanatory variables. The introduction of this methodology into the migration literature is possible because of some new and improved databases developed by the U.S. Bureau of the Census.... Data from the Bureau of Economic Analysis are used to investigate the incorporation of exogenous factors as variables in the model." The exogenous factors considered include employment and unemployment, income, population size of state, and distance between states. The author concludes that "when one...includes additional parameters that are estimable in longitudinal-data models, it turns out that there is little additional information in the exogenous factors that is useful for forecasting."     

Does migration function to reduce earnings differentials by race and gender?

Traditionally, studies of migration have focused on two functions of migration: first, as a means to redistribute society's labor force to its most efficient use and, second, at the level of the individual, as a human capital investment. This study investigates the role of migration in the determination of earnings differentials by race and gender. Microdata from the United Status Census are used to estimate earnings functions by race, gender, and migration status. Strong evidence is provided suggesting that unexplained portions of earnings differentials, reflecting wage discrimination and other unobserved heterogeneity, are substantially reduced for both race and gender through the process of interstate migration.I am grateful to Michael J. Greenwood, Janet M. Rives, and an anonymous referee for insightful comments.     

A partial hierarchical regionalization of 3,140 US counties on the basis of 1965-1970 intercounty migration

"Because of a special tabulation recently completed by the US Bureau of the Census for the 1965-1970 period, interareal migration flows--for the first time--can be examined on the level of the 3,140 counties.... Preliminary results, restricted to six (of a theoretically possible 3,139) levels, are discussed here. They provide some indications of the full results, as well as an upper bound (40,000) on the number of largest doubly standardized values it is necessary to consider in the full clustering."     

Migration age profiles for local authority areas in England, 1971-1981

"This paper is a report on continuing research on the application of age profiles in migration analysis. In particular, it is an update of previous work published in this journal with an analysis of 1981 Census data for local authority areas in England. The changes in migration patterns over the intercensal period from 1971 to 1981 are described." The focus is on internal migration     

基于PSI的水体提取方法在地理国情普查中的应用

引入像元形状指数(PSI),利用像元形状指数能够反映地物匀质性的特点,结合水体的光谱特征,开展高分辨率遥感影像的水体信息提取,实验表明基于像元形状指数和光谱特征的方法较好解决了建筑物阴影等地物的干扰,有效提高了高分辨率影像水体提取精度,该方法在地理国情普查水体覆盖数据生产中取得了较好的效果,为利用高分辨率遥感影像开展地理国情普查地表覆盖数据生产提供了一种新的方法。     

Leaving home,returning home: Migration as a labor market choice for Alaska Natives

We investigate determinants of the pattern of recent migration of Alaska Natives between rural homelands and urban areas. A review of the literature on moving in the North American North suggests that economic opportunities draw migrants, although such opportunities must be viewed in the context of the mixed subsistence-cash economy prevailing in rural areas of the region. Consequently, we model Native migration as a simultaneous decision with labor market participation in a mixed economy. Estimated equations explaining individual Alaska Native migration choices using the U.S. Census Public Use Microdata Sample (PUMS) are consistent with the model, and also suggest that perceived opportunities differ between women and men.The authors acknowledge financial support from the National Science Foundation, grant OPP-9521459. An earlier version of this paper was presented to the 1998 annual meeting of the Western Regional Science Association     

A supply-side interregional model of the U.S. manufacturing industry: 1960–78

This paper presents a multiregional simultaneous model of the U.S. manufacturing industry based on the supply-side paradigm, focusing on the behavior of the primary inputs, capital stock and labor force, across the nine Census Regions. The estimated structure shows that the regional investment sensitively responds to the marginal productivity of capital in the respective region relative to the national average, while the net interregional migration basically follows the relative regional wages. As an application, the model is used for evaluating the economic value of migration and analyzing the relationship between the overall national growth and interregional equalization of income.Presented at Ninth Pacific Conference, Molokai, 1985.