c^.^^'/ rCDS'SD-z-. 
 
 SPECIAL 
 
 DEMOGRAPHIC 
 
 ANALYSES 
 
 
 Appraising the 
 
 Trend and 
 Reasons 
 for Moving 
 
 ^ 
 
 U. S. Department of Commerce 
 BUREAU OF THE CENSUS 
 
SPECIAL 
 
 DEMOGRAPHIC 
 
 ANALYSES 
 
 CDS 80-2 
 
 Appraising the 
 
 Trend and 
 Reasons 
 for Moving 
 
 Larry H. Long 
 
 and 
 
 Diana DeAre 
 
 Issued November 1980 
 
 j-KZ-^K 
 
 f ¥ 
 
 ^rAT^so^""* 
 
 U.S. Department of Commerce 
 
 Philip M. Klutznick, Secretary 
 
 Luther H. Hodges, Jr., Under Secretary 
 
 Courtenay M. Slater, Chief Economist 
 
 BUREAU OF THE CENSUS 
 Vincent P. Barabba, Director 
 
BUREAU OF THE CENSUS 
 
 Vincent P. Barabba, Director 
 Daniel B. Levine, Deputy Director 
 George E. Hall, Associate Director 
 for Demographic Fields 
 
 CENTER FOR DEMOGRAPHIC STUDIES 
 James R. Wetzel, Chief 
 
 Acknowledgments— Marie Pees of Population Division prepared or supervised 
 the computer tabulation of all data appearing in the study except for table 7, 
 which was programmed by Mike Fortier of the Center for Demographic 
 Studies. Clerical and statistical assistance were provided by Joyce E.Williams 
 of Population Division. Ruth Breads typed successive versions of the text 
 and tables. 
 
 Parts of this work were originally presented at the annual meeting of the 
 Population Association of America, Philadelphia, April 26-28, 1979. 
 
 Library of Congress Cataloging in Publication Data 
 
 Long, Larry H 
 
 Migration to nonmetropolitan areas. 
 
 (Special demographic analyses; CDS-80-2) 
 
 Bibliography: p. 
 
 Supt. of Docs, no.: C 3261 : 
 
 1. Migration, Internal— United States. I. DeAre, 
 Diana R., joint author. II. United States. Bureau of 
 the Census. III. Title. IV. Series. 
 HB1965.L58 304.8'2'0973 80-607185 
 
 For sale by the Superintendent of Documents, U.S. Government 
 Printing Office, Washington, D.C. 20402 
 
Preface 
 
 This study is another in a series of publications from the Census Bureau's 
 Center for Demographic Studies. The purpose of these publications is to pro- 
 vide insight and perspective on important demographic trends and patterns. 
 Most bring together data from several sources and attempt to enhance the use 
 of Census Bureau data by pointing out the relevance of the statistics and popu- 
 lation developments for policy analysis and policy planning. A distinguishing 
 feature of the studies is the inclusion of broad speculative analyses and illustra- 
 tive hypotheses offered by the authors as an aid in identifying the reasons 
 underlying population trends. 
 
 Larry H. Long is a senior research associate in the Census Bureau's Center for 
 Demographic Studies. His research has focused on regional population patterns 
 and the changing relationships of central cities, suburbs, and nonmetropolitan 
 areas. Recent publications on these topics include The City-Suburb Income 
 Gap: Is It Being Narrowed by a Back-to-the-City Movement? (coauthored 
 with Donald Dahmann; U.S. Bureau of the Census, 1980) and "Back to the 
 Countryside and Back to the City in the Same Decade," a chapter in Back to 
 the City: Issues in Neighborhood Renovation, edited by Shirley Laska and 
 Daphne Spain (Pergamon Press, 1980). Dr, Long received a Ph.D. degree in 
 sociology from the University of Texas at Austin. 
 
 Diana DeAre received a Ph.D. degree in geography from the University of 
 Texas at Austin, v^here she was an affiliate of the Population Research Center. 
 She has been with the Census Bureau's Population Division since September 
 1975. Her research interests include population distribution, small-town 
 growth and decline, rural development strategies, and the farm and nonmetro- 
 politan populations. 
 
 Ill 
 
Digitized by the Internet Arcliive 
 
 in 2012 with funding from 
 
 LYRASIS JVIembers and Sloan Foundation 
 
 http://archive.org/details/nnigrationtononnneOOIong 
 
One of the most unexpected demographic developments in the United States in 
 the 1970's was the shift of nonmetropolitan areas to net inmigration. Partly 
 because of this change, the nonmetropolitan sector, which includes many small 
 towns and rural areas, experienced noticeably faster rates of total population 
 growth in the aggregate than did metropolitan areas. To many people, these 
 changes seemed to imply a reversal of the longstanding association of rural-to- 
 urban migration with rapid growth of large urban areas, and there was even a 
 suggestion in the data and discussions that new forces might be governing 
 population redistribution in the United States and new motives might be 
 shaping the residential location decisions of individuals. 
 
 Understanding these changes requires an analysis of reasons for moving and the 
 motivations that underlie decisions to live in one place rather than another. 
 Some insights into reasons for the new patterns of the 1970's have been . 
 gleaned from examination of the types of nonmetropolitan counties that had 
 net inmigration. For example, many counties shown by the 1970 census to 
 have concentrations of retirees have grown in population in the 1970's, suggest- 
 ing retirement as a reason for moving that has helped to shift the nonmetro- 
 politan sector to net inmigration. Also, net inmigration to many nonmetro- 
 politan counties with recreational facilities implies that the growth of leisure 
 time and the tendency to spend at least some of it in a rural setting have been 
 factors in the nonmetropolitan migration turnaround. A number of other 
 features of nonmetropolitan counties have been associated with a renewal of 
 population growth in the 1970's (Beale, 1977; Morrill, 1978). 
 
 Economic changes have clearly played a role. A renewed search for energy has ^ 
 produced net inmigration in coal-mining areas of rural West Virginia and some 
 of the Rocky Mountain States, and other economic changes— like the shift of 
 many light-manufacturing jobs to nonmetropolitan locations to take advantage 
 
 1 
 
of lower taxes, less expensive land, and cheaper labor costs— have decentralized' 
 jobs and enhanced employment opportunities in nonmetropolitan areas (see, 
 for example, U.S. Department of Housing and Urban Development, 1978)j 
 Expansion of many jobs in services, like local government, has also facilitated 
 the change to net inmigration in the nonmetropolitan sector in the 1970*s 
 (Carpenter, 1977). Employment as well as population grew more rapidly in 
 nonmetropolitan areas than in metropolitan areas in the 1970's, suggesting an 'I 
 economic basis for the population shift (Regional Economic Analysis Division, \jt 
 1978; Renshawetal., 1978; Wardwell and Gilchrist, 1980). _J 
 
 Among researchers, however, there is near-universal agreement that economic 
 factors alone cannot fully account for or explain the change in the 1970's in 
 migration between the metropolitan and nonmetropolitan sectors. A turn- 
 around in migration patterns has occurred even in relatively isolated counties 
 of modest income levels (Beale, 1977), and^thjs and other evidence have sug- 
 gested to many persons that noneconomic considerations involving quality^of^—r 
 life or environmental amenities have become more important in individual 1^ 
 "^decisions to move or not and that this development has favored a tilt of -^ 
 population toward less urban, more residentially attractive settings in small 
 towns or rural locations (Beale, 1975; Goldstein, 1976; Morrison, 1976; 
 McCarthy and Morrison, 1978). The theory is that either residential prefer- 
 ences have changed to favor nonmetropolitanai:easLj>.nndjyjduals_have become 
 more willing or better able to act on the basis of longstanding preferencesToT- 
 living in a nomrietro poll tan setting, even at the sacrifice of income maximiza- 
 tion (Carpenter, 1977; Roseman; 1977; Wardwell, 1977; Ploch, 1978; Blundell, 
 1980). 
 
 This monograph presents a perspective on the duration and motivational basis 
 of this new population pattern, which emerged in the early 1970's. We first 
 want to ask the obvious question: 
 
 • Is there any evidence, according to the most recently available data, of a 
 slowdown or abatement in the shift of population growth and net inmigra- 
 tion in favor of nonmetropolitan areas? 
 
 After addressing this question, we want to go beyond the basic population 
 figures and inquire, to the extent feasible, into the behavioral bases underlying 
 the development of net inmigration to nonmetropolitan areas. Data on reasons 
 for moving provide some insights along these lines, so drawing upon the Annual 
 Housing Survey, the investigation that follows is the first on a nationwide basis 
 of reasons for moving to nonmetropolitan areas of the United States in the 
 1970's. The questions we address include: 
 
 • To what extent do nonemployment motives underlie the net migration to 
 nonmetropolitan areas in the 1970's? Would nonmetropolitan areas still have 
 net inmigration if only persons moving for employment-related reasons were 
 considered? 
 
u« 
 
 How many of the migrants to nonmetropolitan areas stay close enough to 
 allow them to commute into a metropolitan area for employment? Do these 
 exurban movers differ from other metropolitan-to-nonmetropolitan movers 
 in terms of reasons for moving? That is, are they like city-to-suburb movers 
 but simply moving farther out, or do their reasons for moving imply differ- 
 ent motives that carry them slightly farther from an urban core than city-to- 
 suburb movers? 
 f In what types of migration streams— between metropolitan areas, between 
 nonmetropolitan counties, from metropolitan to nonmetropolitan or from 
 nonmetropolitan to metropolitan areas— are economic motives most preva- 
 lent? Conversely, in which of these various types of migration do retirees 
 and persons moving for ostensibly noneconomic reasons play a significant 
 role? 
 
 Because effective policy planning requires an assessment of how long current 
 patterns will continue, do available data provide any basis for speculating 
 about whether present trends toward population deconcentration can con- 
 tinue in the face of energy shortages and rising prices of energy? 
 
 THE NEW PATTERN: IS IT REAL? HAS IT ENDURED? 
 
 An apparent new trend in population and migration patterns was announced 
 on November 23, 1973, when the Census Bureau issued a press release with 
 data from the March 1973 Current Population Survey showing more persons to 
 have moved from than to metropolitan areas in the three years since the 1970 
 census (U.S. Bureau of the Census, 1973). An independent data source, the 
 Census Bureau's annual estimates of population by county, yielded two widely 
 accepted "proofs" that the Current Population Survey data did indeed repre- 
 sent a new pattern of population redistribution. .First, the net inmigration to 
 nonmetropolitan areas could notbejltdbMeid, sijnplyJolh 
 lation beyond the statistical boundaries of metropolitan areas, for even non- 
 Inetropolitan counties not adjacent to metropolitan areas shifted to net in- 
 migration and grew more j;a£idj;^_th aji met^ in the 1970*s (Beale, 
 1975). Secpnd, theJtrj^tiojiaUHDositiw^associatioa betw^e 
 county's incom£J£w4--af^d™^ts^tkel^hood (and rale ).„„Q^^ net inmigration dis- 
 appeared in th.£JL97n!s,._and_Uierejwas^ that the rate of net 
 in migration JoMionrneJj^o^^ was highest for those with the lowest 
 in^come levels (Beale^1^^977). Clearly, more net movement to relatively isolated 
 rural counties and those with modest income levels occurred, raising the possi- 
 bility that monetary incentives had declined in significance as a reason for 
 moving. 
 
 Concomitant with these empirical discoveries arose the question of whether the 
 new pattern might be only a temporary manifestation that would soon revert 
 to the traditional trend. One economist suggested that the pattern observed in 
 1970-73 reflected "temporary cyclical adjustments resulting from the serious 
 increase in unemployment rates in a number of metropolitan areas" (Kain, 
 
1975, p. 224). But the overall pattern was not a mere aberration, for both data 
 sets— the annual Current Population Survey and the annual county estimates of 
 population and net migration— have confirmed the existence of the pattern for 
 each year through 1978 or 1979. Thus, the new pattern has endured about 
 8 or 9 years. 
 
 Our basic concern here is whether there is any observable change in the pattern 
 itself, for example, whether there is any evidence of a slowdown in the shift of 
 population and net inmigration to the nonmetropolitan sector. The net in- 
 migration to nonmetropolitan areas in the early 1970's was a pre-energy-crisis 
 migration, and the oil embargo of the winter of 1973-74 and subsequent spot 
 shortages and rising prices of gasoline could have caused individuals and firms 
 to reassess their decisions regarding relocation. Other researchers have sug- 
 gested (Beale, 1976; Phillips and Brunn, 1978) that at some point rising prices 
 and declining availability of gasoline and diesel fuel put constraints on the 
 extent to which population can deconcentrate, because nonmetropolitan 
 manufacturing plants highly depend on truck transportation and nonmetro- 
 politan residents depend more on private cars for transportation than city or 
 suburban residents. Moreover, because per capita and per family incomes 
 remain lower in nonmetropolitan areas (Long and Dahmann, 1980), the ability 
 to absorb gasoline price increases in family budgets may be lower among non- 
 metropolitan than metropolitan residents. 
 
 Data recently created provide an opportunity for an initial test to examine 
 whether events in the 4 years after the 1973-74 oil embargo measurably slowed 
 the shift of population toward nonmetropolitan areas. The Census Bureau's 
 annual estimates of total population and components of change (net migration 
 and natural increase) for counties are now available from 1970 through 1978, 
 and with such data we can ask whether the net shift of population toward 
 nonmetropolitan areas was as great in the post-energy -crisis years of 1974-78 as 
 in the pre-energy-crisis years 1970-74. 
 
 A difficulty in accurately making such a test is that the official definitions of 
 metropolitan and nonmetropolitan changed between the beginning and end of 
 the 8-year study period. For example, after the commuting data from the 1970 
 census became available in 1973, more than 100 counties were added to stand- 
 ard metropolitan statistical areas (SMSA's), reflecting the extension of subur- 
 banization and the fact that as a result a number of nonmetropolitan counties 
 had become functionally parts of metropolitan areas. In addition, new SMSA's 
 are created as nonmetropolitan cities grow into metropolitan status, and based 
 on the Census Bureau's intercensal population estimates and special censuses 
 since 1970, a number of counties that were nonmetropolitan in 1970 have 
 been reclassified as new metropolitan areas. Thus, in order to compare 1970-74 
 and 1974-78 population change in metropolitan and nonmetropolitan areas, 
 one must deal with the fact that official statistical definitions of metropolitan 
 and nonmetropolitan differed in 1970, 1974, and 1978. 
 
 Accordingly, we assembled table 1 to show 1970-74 and 1974-78 average 
 
00 
 
 T3 
 
 o 
 
 n3 
 
 Z 
 
 ■a 
 
 c 
 
 c 
 *■*-> 
 
 3 
 Q. 
 O 
 
 a. 
 c 
 IS < 
 
 si 
 
 
 
 n 
 
 <D 
 
 
 
 
 3 
 
 <4- 
 
 «J 
 
 C 
 
 (U 
 
 V 
 
 i- 
 
 M 
 
 ( > 
 
 
 c 
 
 r 
 
 
 f« 
 
 
 bfi 
 
 sz 
 
 
 S? 
 
 o 
 
 
 >- 
 
 (4- 
 
 o 
 
 3 
 
 < 
 
 s 
 
 c 
 
 CO ON 
 
 
 E 
 
 
 r^ 
 
 3 
 
 c 
 
 o 
 
 
 Tt 
 
 <+- 
 
 c 
 o 
 
 en 
 
 c 
 
 W) 
 
 2 
 
 r— 
 
 ffl 
 
 c 
 
 
 
 ro 
 
 bfl 
 
 
 
 F 
 
 
 
 ^- 
 
 
 ^^ 
 
 o 
 
 <a 
 
 r^ 
 
 <r 
 
 a» 
 
 
 o 
 
 
 
 
 r^ 
 
 
 
 
 <J\ 
 
 
 -a £^ 
 
 r3 Q 
 
 15 
 
 3 
 
 r 
 
 
 c 
 o 
 
 
 c 
 
 c 
 
 
 r— 
 
 f« 
 
 ra 
 
 3 
 
 
 ft}. 
 
 sz 
 
 n 
 
 
 o 
 
 o 
 
 
 n 
 
 M- 
 
 n 
 
 
 % 
 
 o 
 
 
 ^ 
 
 (U 
 
 rt 
 
 r^ 
 
 < 
 
 2 
 
 o 
 
 o 
 
 CM 00 
 
 o r^ 
 
 O <N 
 
 r- VO 
 
 U-l CN VO r- 
 
 iS 
 
 c 
 
 ! 
 
 00 
 
 o 
 
 c 
 
 D. 
 
 (Ti 
 
 O 
 
 -C 
 
 4-> 
 
 U 
 
 <U 
 
 o 
 
 ^ 
 
 4i 
 
 _c 
 
 00 
 
 c 
 
 0) 
 
 T5 
 
 W1 
 
 
 c 
 
 O 
 
 (Tl 
 
 O 
 
 J^ 
 
 o 
 
 6< 
 
 r— 
 
 
 a> 
 
 
 
 
 ^ 
 
 
 «j 
 
 
 H 
 
 
 ^ o 
 < 
 
 o m 
 
 00 «N 
 
 00 «N 
 
 CN r— 
 
 
 CN T- 
 
 O -^ CN CN CN CN 
 
 ■«t o a\ lo 00 vo 
 
 lo 00 o\n^ I— r- 
 
 "^o CO r^ vo <o 
 
 ■* 00 
 
 CN r>. 
 
 rn ox 
 
 O CN 
 
 -^ VO 
 
 O CN 
 00 CN 
 
 <»« t. 
 00 -t 
 
 ■«t in 
 
 00 r- 
 
 00^ Tt 
 
 O CN 
 
 m in 
 
 c 
 o 
 
 I I 
 
 - o. 
 
 (/) 
 
 1/1 
 
 
 
 y < 
 
 - - r_' c "^ ^' c < ^ 
 oto^ oto^ 0(yiS 
 ;^.2 LO ;^ ;^ cyi p ^ to 
 
 a> -a to a> -a wi " 
 9 — O ^ " 
 
 00 "" 
 
 C U- 
 
 LU (U 
 
 a> 1- 
 
 *-> c 1- 
 
 f^-S ^ 
 <u oc ^ 
 
 ^ C V, 
 ^ UJ 3 
 
 |l^ 
 
 ^ .E ^ 
 ^ 1/1 *^ 
 to r- > 
 — < ^ 
 
 5 U « 
 "» Z ci 
 
 n] L. 
 
 .i£ J^ D. 
 
 •^^ s 
 
 t/1 to c 
 
 ~ ^ c 
 
 O 3 O 
 
 o. o -^ 
 2 o J2 
 
 +J ■<-> 3 
 
 E fe o 
 
 ^ ■+- °- 
 
 "• c c 
 o 
 
 ■^ '7, ^-o -m 
 0\ o^ 
 
 = .ti ^ 
 *-> -x o 
 
 " a-- 
 2 o o 
 
 if 5 § 
 
 a> C rj 
 ■■ ° 3 
 
 S • f^ 
 
 ^ TD -^ 
 O (/) "rt 
 
 P (U « 
 <U "5 to 
 
 z 2 w^ ~ 
 2 </i 
 
 ^ LU 
 
annual rates of change in total population, net migration, and natural increase 
 for metropolitan and nonmetropolitan areas as defined in 1970, 1974, and as 
 of January 1980. We took 1980 as the last date rather than 1978 in order to 
 obtain the most extensive definition of what is metropolitan. The official 
 determination of what is to be considered metropolitan rests with the Office of 
 Federal Statistical Policy and Standards (OFSPS), formerly a part of the Office 
 of Management and Budget but now a part of the Commerce Department (still 
 independent of the Census Bureau, however). A committee sponsored by 
 OFSPS meets periodically to decide whether additional counties meet the 
 established criteria for metropolitan (SMSA) status. In somewhat oversimpli- 
 fied terms, an SMSA consists of a county with a city (or twin cities) of at least 
 50,000 population; an adjacent county may be included if there is significant 
 commuting Into the central county. What is not part of an SMSA is considered 
 nonmetropolitan. 
 
 TRegardiess of whether one uses the 1970, 1974, or 1980 SMSA definition, the 
 average annual rate of population growth declined in both metropolitan and 
 nonmetropolitan areas between the 1970-74 and 1974-78 intervals. This de- 
 cline reflects a decrease in the national rate of population growth, from 0.91 
 percent per year in 1970-74 to 0.78 percent per year in 1974-78. The total 
 population increased in both the metropolitan and nonmetropolitan sectors. 
 This point is important because although a number of individual metropolitan 
 areas declined in total population in the 1970's (U.S. Bureau of the Census, 
 1980), the total metropolitan population of the nation increased. The non- 
 metropolitan sector, however, grew more rapidly during both intervals, irre- 
 spective of which definition of metropolitan is used. 
 
 ''Note that except for the most restrictive (1970) metropolitan definition, both 
 the metropolitan and nonmetropolitan sectors are shown to have net inmigra- 
 tion (i.e., positive rates of population change due to migration). Both sectors 
 can have net inmigration because the data in table 1 include migration from 
 abroad. It is not possible to exclude such migration from the table, but survey 
 data indicate that when migration from abroad is excluded, the metropolitan 
 sector shows substantial net outmigration to the nonmetropolitan sector (U.S. 
 Bureau of the Census, 1975; Bowles, 1978). 
 
 Clearly, the nonmetropolitan population has continued to grow more rapidly 
 than the metropolitan population, even in the post-energy-crisis era and regard- 
 less of the metropolitan definition. But has the shift of population to the 
 nonmetropolitan sector slowed down? Such a slowdown could be indicated by 
 a decline in the degree to which nonmetropolitan population growth exceeded 
 that of metropolitan areas; tables 2, 3, and 4 provide this and other measures 
 extracted from the basic data in table 1 . 
 
 Table 2 shows the ratio of nonmetropolitan rates of population change to 
 metropolitan rates in order to facilitate comparison of the degree to which the 
 nonmetropolitan sector grew more rapidly than the metropolitan in the two 
 time periods. The table reveals that the ratio rose between the 1970-74 and 
 
Table 2. Degree of Difference Between Metropolitan and Nonmetropolitan 
 Average Annual Rates of Total Population Growth and Natural 
 Increase, 1970-74 and 1974-78, According to Changing 
 Definitions of SMSA 
 
 SMSA definition 
 
 Ratio of nonmetropolitan 
 
 to metropolitan population 
 
 growth 
 
 1970-74 
 
 1974-78 
 
 Ratio of nonmetropolitan 
 
 to metropolitan natural 
 
 increase 
 
 1970-74 
 
 1974-78 
 
 1970 definition 
 1974 definition 
 1980 definition 
 
 1.99 
 1.54 
 1.48 
 
 2.25 
 1.87 
 1.80 
 
 0.89 
 .85 
 .83 
 
 0.95 
 .95 
 .94 
 
 Table 3. U.S. Population in Metropolitan Areas in 1970, 
 1974, and 1978, According ro Changing 
 Definitions of SMSA 
 
 (Percent) 
 
 
 
 
 SMSA definition 
 
 1970 
 
 1974 
 
 1978 
 
 1970 definition 
 1974 definition 
 1980 definition 
 
 69.0 
 73.2 
 74.2 
 
 68.4 
 72.9 
 73.9 
 
 67.8 
 72.4 
 73.5 
 
 Table 4. National Population Growth Occurring 
 In Metropolitan Areas, 1970-74 and 
 1974-78, According to Changing 
 Definitions of SMSA 
 
 (Percent) 
 
 
 
 SMSA definition 
 
 1970-74 
 
 1974-78 
 
 1970 definition 
 1974 definition 
 1980 definition 
 
 52.4 
 63.6 
 65.9 
 
 48.8 
 58.8 
 60.8 
 
 1974-78 intervals, indicating a more rapid population shift to the nonmetro- 
 politan sector in the more recent interval. Observe, for example, that for the 
 1970 definitions the nonmetropolitan sector's average annual rate of popula- 
 tion growth was 1.99 times as great as that of the metropolitan sector in the 
 1970-74 period; by 1974-78 the nonmetropolitan growth rate was 2.25 times 
 as great. Even with the updated, more expansive definitions of metropolitan in 
 use in 1974 and 1980, the same conclusion applies: the growth-rate differential 
 widened in favor of nonmetropolitan areas in the post-energy -crisis period. 
 
8 
 
 Hence, we tentatively conclude that energy developments since the 1973-74 oil 
 embargo did not slow down, at least by 1978, the shift in population growth 
 toward the nonmetropolitan sector. Although rates of natural increase and net 
 migration declined between the two time periods in both sectors, the shift 
 accelerated in 1974-78 relative to 1970-74 because the declines were greater in 
 metropolitan areas. The data suggest that persons living in nonmetropolitan 
 areas or wishing to live there have been able to make adjustments and accom- 
 modations to rising energy prices without changing their residential location 
 decisions enough to measurably slow the shift of population growth away from 
 metropolitan areas. 
 
 Both natural increase and net migration contributed to widening the differen- 
 tial in rates of population growth between metropolitan and nonmetropolitan 
 areas. Sustained net inmigration to an area often involves young persons in the 
 reproductive years and thereby tends to have a positive effect on natural in- 
 crease through this alteration in age composition. This type of effect may help 
 to explain why, at a time when birth rates for the Nation as a whole are falling, 
 rates of natural increase fell less in nonmetropolitan than in metropolitan areas. 
 The point is that net inmigration during one time period tends to instill a 
 momentum to population growth during later time periods, and the momen- 
 tum of population growth that began with the net inmigration to nonmetro- 
 politan areas in 1970-74 continued with even stronger force in 1974-78. 
 
 IS THE NATION UNDERGOING 
 DEMETROPOLITANIZATION? 
 
 From at least 1900 to 1970 the United States underwent metropolitanization, 
 whereby a growing proportion of the population at each census was found to 
 be living in metropolitan areas (Taeuber, 1972). Because metropolitan areas 
 had slower rates of population growth than the nonmetropolitan sector in the 
 1970's, can we say that the opposite process-demetropolitanization-was at 
 work? One of the ironies of nonmetropolitan population growth is that if it is 
 rapid enough and continues long enough, it transforms the character of non- 
 metropolitan localities to metropolitan, either as a result of the growth of cities 
 or towns to metropolitan status in their own right or as a result of fusing 
 nonmetropolitan counties with existing metropolitan areas. In the 1970's the 
 proportion of the population living in metropolitan areas increased 6^r only as 
 a result of the reclassification of nonmetropolitan counties to metropolitan 
 status. Because of such reclassifications, the percent of U.S. population living 
 in metropolitan areas rose from 69.0 percent in 1970 to 73.5 percent in 1978 
 (table 3). 
 
 The percent of the U.S. population living in territory classified as metropolitan 
 in 1970 fell from 69.0 percent to 67.8 percent in 1978. Even if we take the 
 more expansive definition of metropolitan in use in 1980 and extend it back to 
 1970, the percent of population living in metropolitan territory fell from 74.2 
 percent in 1970 to 73.5 percent in 1978. These may seem like very small 
 
changes, but they deviate from historical patterns. In the past the percent of 
 population living in metropolitan areas increased from three sources: (1 ) popu- 
 lation growth in existing metropolitan territory at a rate above the national 
 average, (2) population growth on the fringes of metropolitan areas and the 
 subsequent inclusion of nonmetropolitan counties in existing metropolitan 
 areas, and (3) the birth of new metropolitan areas as nonmetropolitan cities 
 and towns grow large enough to be redefined as metropolitan. At the present 
 time the Nation's metropolitan population increases entirely from the second 
 and third sources. 
 
 Because a majority of Americans live in metropolitan areas (using the 1970 or 
 1980 definitions of metropolitan), it should come as no surprise to find that 
 more than one-half of the Nation's population growth between 1970 and 1978 
 occurred within the borders of metropolitan areas (table 4). Between 1970 and 
 1974 the Nation's population grew from 203.3 million to 211.3 million— 8 
 million persons. Under any of the three definitions of metropolitan, more than 
 one-half of this increase went to SMSA's. In 1974-78 total population growth 
 for the country as a whole was down to 6.6 million, and except for the most 
 restrictive (1970) definition of SMSA's, more than one-half of this growth 
 occurred in SMSA's. Importantly, however, metropolitan areas have been 
 absorbing a declining proportion of the Nation's population growth— by all 
 three definitions of metropolitan areas. 
 
 Of course, analyzing population distribution on a metropolitan- 
 nonmetropolitan basis is only one perspective on population distribution, and 
 there are many other ways of measuring population concentration or decon- 
 centration (e.g., see Long, forthcoming). If available on an intercensal basis, 
 more complex measures of spatial patterns might provide fuller insights into 
 the effects (if any) of energy developments on population deconcentration 
 trends of the 1970's. Moreover, there are likely to be considerable lags between 
 systemic shocks (like rapidly rising prices of gasoline) and changes in popula- 
 tion distribution, so that the effects of energy developments in 1974-78 might 
 not be evident in the data but might manifest themselves over a much longer 
 period of time. There has not been a great deal of research to draw upon in 
 trying to assess how quickly households change their residence or in other ways 
 adjust their behavior in response to rapidly changing energy circumstances. 
 
 We want to emphasize that the data presented up to now are from the Census 
 Bureau's population estimates for individual counties summed to metropolitan 
 and nonmetropolitan aggregates. Because the data are estimates, they are sub- 
 ject to error. The Census Bureau initially began to extend its intercensal popu- 
 lation estimates down to the county level in the late 1960's, but not until the 
 1970's were data available annually. The 1980 census will allow assessment of 
 the accuracy of techniques used to prepare population estimates for individual 
 counties in the 1970's, and although discrepancies between the estimates and 
 the census population counts for counties are sure to occur, some will offset 
 each other when counties are grouped into metropolitan and nonmetropolitan 
 categories. 
 
^ 
 
 10 
 
 Our basic conclusion is simply that available data indicate that the 
 metropolitan-to-nonmetropolitan shift of population first observed in the early 
 1970*s accelerated in 1974-78. The present is a good time to ask this question 
 because although the 1980 census will provide a more detailed snapshot view in 
 comparison with 1970, it can offer relatively little in assessing whether the 
 trend observed in the early 1970*s had changed by 1980. 
 
 GROSS FLOWS AND REASONS FOR MOVING 
 
 Strong forces appear to have continued in 1974-78 to push people away from 
 metropolitan areas generally and to pull them toward the less densely settled 
 nonmetropolitan counties. One way of assessing the reasons behind these push 
 and pull forces examines the types of counties experiencing changes in migra- 
 tion patterns. In the 1970's this approach tended to concentrate on pull factors 
 because of the focus on types of nonmetropolitan counties that changed from 
 net outmigration in the 1960's to net inmigration in the 1970's. Econo mic 
 CP reasons fprnioving are suggested by nonmetropolitan turnaround counti£SJ;hat 
 attract manufacturing or experience renewed oil and gas exploration or coal 
 mining. A noneconomic basis for the nonmetropolitan turnaround is suggested 
 by counties with growing concentrations of retirees or without an obvious 
 economic explanation for growth. 
 
 Perhaps the most direct method of assessing why people move is simply to ask 
 them. A strong representation of noneconomic reasons for moving to non- 
 metropolitan areas is suggested by a number of special surveys taken in recent 
 years in selected groups of counties in the Midwest (Williams and Sofranko, 
 1979), the Upper Great Lakes (Voss and Fuguitt, 1979), and Pennsylvania 
 (Dejong and Keppel, 1979). Although these regional surveys throw consider- 
 able light on the subject, the broader question remains: For the United States 
 as a whole, would nonmetropolitan areas still have net inmigration if only 
 persons moving for economic reasons were considered? A negative answer to 
 this question would imply that essentially noneconomic considerations have 
 played a large, possibly dominant role in the nonmetropolitan turnaround (at 
 least at the national level) and whatever trends might put more people in a 
 position to move (or not to move) out of other than strictly job-related con- 
 siderations might also tend to continue the shift of the U.S. population toward 
 nonmetropolitan counties. 
 
 In the 1970's the largest nationwide surveys to ask persons moving to non- 
 metropolitan (and other) localities to give their reasons for moving were the 
 Annual Housing Surveys conducted by the Census Bureau for the Department 
 of Housing and Urban Development. We used the 1975 survey because it 
 provided a mid-decade perspective and included a one-time supplement on 
 place of work. The same questions on mobility status and main reason for 
 moving were included in the surveys of 1975 through 1978. This part of the 
 1975 questionnaire is reprinted as figure 1. 
 
11 
 
 Figure 1 . Questions on mobility status and reason for moving in tlie 1 975 
 annual housing survey 
 
 Section IMC - OCCUPIED UNITS (Include URE) Continued 
 
 CHECK 
 ITEMQ 
 
 □ URE household (See 
 "(See Check Item A(3) 
 
 □ Head moved here dur 
 
 □ Head has lived here 
 
 Item 7. poge /) - Skip to 105. poge 31 
 page 14) 
 ng the last 12 months - Ask 83 
 2 months or longer — Skip to 102a, page 30 
 
 83. What wo* th* address of . . .'s (hood) 
 previous residence? 
 
 ^, 
 
 
 Address (Number and street) 
 
 City or town 
 
 County State ZIP code 
 
 @ 
 
 OR 
 
 1 □ Outside the United States - Skip to 102a, 
 
 1 1 1 1 1 "*''° 
 
 84. Whot is the main reoson . . . (head) moved 
 from his previous residence? 
 
 (Write all reasons mentioned below, and then 
 mark the main reason.) 
 
 
 ' EMPLOYMENT 
 
 \(\7i) 1 □ Job transfer 
 
 1 2 □ Entered or left U.S. Armed Forces 
 
 1 3 □ Retirement 
 
 1 4 □ New job or looking for work 
 
 1 5 □ Commuting reasons 
 
 1 6 □ To attend school 
 
 1 7 □ Other 
 
 1 FAMILY 
 
 \(m) 8 □ Needed larger house or apartment 
 
 1 9 □ Widowed 
 
 j 10 □ Separated 
 
 1 " □ Divorced 
 
 1 1 2 □ Moved to be closer to relatives 
 
 ] 13 □ Newly married 
 
 I 14 1 [Family increased 
 
 1 15 □ Family decreased 
 
 1 16 □ Wanted to establish own household 
 
 1 17 □ Other 
 
 
 
 
 
 
 
 
 
 
 
 
 
 '. ■' •' ''';(;■';". 
 
 
 1® 
 
 OTHER 
 
 18 □ Neighborhood overcrowded 
 
 19 □ Change in racial or ethnic composition 
 
 of neighborhood 
 
 20 □ Wanted better neighborhood 
 
 21 □ Wanted to own residence 
 
 22 □ Lower rent or less expensive house 
 
 23 □ Wanted better house 
 
 24 □ Displaced by urban renewal, highway 
 
 construction, or other public activity 
 
 25 □ Displaced by private action 
 
 26 □ Schools 
 
 27 □ Wanted to rent residence 
 
 28 □ Wanted residence with more conveniences 
 
 29 □ Natural disaster 
 
 30 □ Wanted change of climate 
 
 31 □ Other 
 
 HeacJs of households who moved in the 12 months preceding the survey, which 
 was conducted in October, November, and December 1975, gave the locations 
 of their previous residences and then were asked, "What is the main reason. . . 
 (head) moved from his previous residence?" The interviewer was instructed to 
 write on the lines provided on the questionnaire the reason or reasons for 
 moving. For respondents giving more than one reason, the interviewer asked 
 
12 
 
 which was the "main" reason and then marked one of the 30 reasons listed on 
 the questionnaire or else marked the "other" category (reason 31). In process- 
 ing the data, multiple responses were not retained; so we have no information 
 on the frequency with which each of the individual reasons was mentioned or 
 the way in which respondents chose the main reason among several offered. 
 
 In the survey, household heads were identified in the manner traditionally 
 employed by the Census Bureau (see, e.g., U.S. Bureau of the Census, 
 1977a). In husband-wife couples, the husband was considered the head for 
 purposes of data collection and tabulation, but automatic designation of the 
 husband as the household head is being discontinued. 
 
 About 62,000 households were interviewed in the 1975 national sample. Migra- 
 tion responses were coded in terms of the boundaries of standard metropolitan 
 statistical areas defined at the time of the 1970 census. (Additional informa- 
 tion about the sample, including estimates of sampling errors, may be found in 
 U.S. Bureau of the Census, 1977a.) 
 
 The basic data on migration flows (in terms of households) between the metro- 
 politan and nonmetropolitan sectors and the main reasons for moving in the 12 
 months preceding the 1975 Annual Housing Survey appear in table 5. We have 
 rearranged the order of the categories on the questionnaire (figure 1), and 
 because of small cell frequencies we have collapsed some of the categories, 
 usually in obvious ways although neighborhood dissatisfaction includes reasons 
 18, 19, 20, and 26 and the miscel laneous category under housing and neighbor- 
 hood encompasses reasons 24, 25, 27, and 29. 
 
 The number of households (770,000) moving to nonmetropolitan areas, the 
 number (651,000) moving from nonmetropolitan areas, and the net inmigra- 
 tion of 119,000 households to nonmetropolitan areas during the year preced- 
 ing the 1975 Annual Housing Survey agree fairly closely with independent data 
 from the Current Population Survey (U.S. Bureau of the Census, 1977b). The 
 most Important conclusion from table 5 is the demonstration that the net 
 inmigration to nonmetropolitan areas results from households moving for 
 reasons not directly related to employment. Note that the number of house- 
 holds moving to metropolitan areas for employment reasons (316,000) approx- 
 imately equals the number of households moving from metropolitan areas for 
 employment reasons (313,000). These figures mean that metropolitan areas 
 about broke even among households whose main reason for moving was 
 employment related. 
 
 But metropolitan areas lose migrants whose main reason for moving is not 
 directly employment related. These nonemployment reasons include a variety 
 of family considerations and housing and neighborhood aspirations, but being 
 more specific is difficult because many individual reasons have cell frequencies 
 far too small to draw firm conclusions about net movements between sectors. 
 
 Metropolitan areas appear to attract persons experiencing marital breakups 
 (widowhood, divorce, or separation), persons entering or leaving the Armed 
 
13 
 
 c 
 
 I 
 
 "o 
 
 Q. 
 
 O 
 
 u. 
 
 0) 
 
 E 
 
 c 
 o 
 z 
 
 -o 
 
 c 
 
 c 
 
 O 
 Q. 
 O 
 
 k- 
 
 •M 
 
 c 
 
 CD > 
 00 3 
 
 .E (^ 
 
 § ^ 
 
 ^ X 
 
 xi 
 
 :a < 
 
 o m 
 
 ^2 
 
 O <U 
 
 X5 
 
 >- 00 
 
 X) c 
 
 <u a> 
 tj o 
 
 O ^ 
 QlQ- 
 
 
 o ^ <l> 
 
 Il- 
 ls 
 
 .ti Q. c 
 O 2 o 
 
 r: o 
 
 o ° 3 
 
 i: c o 
 
 <o '(5 ■£ 
 
 O o = 
 
 i: 2 o 
 12 
 
 c — 
 
 c« o ^r 
 
 tf Q. </> 
 
 ^ 2 § 
 
 2 i 
 
 P »^. "^ "*. «N CO Tf r-; CN r-; p 1— , T— , CN »-; <N OJ Vq 00 1—, 00 00 Tj; .-; 
 
 O oovovof^tNodvoiniXcNr^r- CNi— r— TtoN»— '--»nco 
 
 o r^copc>Jc^r^oqvqo^T^<^^^^or-;CNf<^^-;l/^o^vqT-;ln«/^ 
 O o in CO T-^ T- r>^ vo iX »— * ro u^* r^i r^ c>I 1-^ Tf <N <N in «n fN r->* cr> 
 
 momCNvovocNONoofOCNTf 
 
 t— vooocnt— »noO'^cocovor^'*r*oovo'^t— ONCNCNino 
 lo T-or^CNr- CNTfcncnT— -^ r~ rom»— »— cocn 
 
 VO m T— T- r- 
 
 o rooor^<T>ONvofNco«nvooor-oO'— OimvoCTNCNcnvooor-. 
 
 r- en t— T— r— r- 
 
 V ♦J "*i o .E 
 
 i| 
 It 
 
 E 
 
 .r. -a -a 
 
 >1 § 
 
 r ^ o 5 
 
 l_ <J <« «» J_ «« 
 
 o 1 ^ s s o S 
 
 J2 C 
 
 1- 
 
 o 
 
 c 
 
 0) 
 
 
 
 I/I 
 
 « 
 
 
 ^ 
 
 88 
 
 3 
 
 3 
 
 u 
 
 c 
 
 3 J= 
 
 
 O 
 
 >- 
 
 0) 
 
 o i: 
 
 c 
 
 ^ 
 
 c 
 
 u. 
 
 •c 2 
 
 m 
 
 0) 
 
 ^ 
 
 (U 
 
 Sjf 
 
 « 
 
 go 
 
 o 
 
 ^ 
 
 S.5P 
 
 « 
 
 (D 
 
 o 
 
 o 
 
 u a> 
 
 •— 
 
 -J 
 
 1- _J CQ Z S 
 
 3 < 
 
 O 
 is 
 
 Ui < Qi 
 
 ll-s 
 
 «> o £ 
 
 5 O O 
 U Z Z 
 
14 
 
 Forces, and students. The number of sample cases in these categories is too 
 small to allow the conclusion that metropolitan areas actually have net inmigra- 
 tion of household heads with these characteristics, but other studies have 
 suggested reasons why metropolitan areas (especially central cities) appeal to 
 persons in at least some of these categories (Long and Click, 1976; Munick and 
 Sullivan, 1977). 
 
 Note that the employment-related reasons appear to be more often reported by 
 migrants to than from metropolitan areas. About 48.5 percent of households 
 moving to metropolitan areas gave one of the employment reasons, compared 
 with 40.7 percent of household heads leaving metropolitan areas. Unless one 
 adopts a more expansive concept of employment-related reasons than shown in 
 table 5, the conclusion is that strictly employment-related reasons account for 
 less than a majority of households moving between the metropolitan and non- 
 metropolitan sectors. 
 
 In general, the data tend to support the thesis that the shift of migration in the 
 1970's to favor nonmetropolitan areas is substantially the product of persons 
 moving for reasons not directly related to jobs. The net inmigration to non- 
 metropolitan areas seems to result from persons moving for such nonemploy- 
 ^ ment reasons as a desire to be closer to relatives (conceivably some of these 
 migrants are earlier rural-to-urban migrants "going home") and newly estab- 
 lished households and others with a desire to own their own homes or to 
 obtain larger houses. Lower housing prices in nonmetropolitan locations also 
 seem to have drawn migrants from metropolitan areas, and dissatisfaction with 
 neighborhood conditions in metropolitan areas appears to propel r^^^^^ 
 nonmetropolitan areas. 
 
 The nonmetropolitan sector probably has net inmigration of retirees, but the 
 number of sample cases in the 1975 Annual Housing Survey is too small to 
 draw firm conclusions. According to table 5, an estimated 43,000 household 
 heads moved to the nonmetropolitan sector in the 12 months preceding the 
 survey, and 12,000 moved in the opposite direction. The apparent difference 
 between the two numbers is not large enough to draw statistically reliable 
 conclusions about the net exchange of retirees. These figures are perhaps sur- 
 prising in that the representation of retirees in the metropolitan-to- 
 nonmetropolitan stream is not higher, for the effect of retirees has played a 
 prominent role in analyses of the nonmetropolitan migration turnaround (e.g., 
 Beale, 1975; Morrison, 1976). An estimated 5.6 percent of household heads 
 migrating to nonmetropolitan areas in the 12 months preceding the 1975 
 Annual Housing Survey named retirement as the main reason for moving. Even 
 if these retirees were excluded from the stream of outmigrants, metropolitan 
 areas still would not have net inmigration. Hence, these data on self-reported 
 reasons for moving suggest that retirees account for a relatively modest propor- 
 tion of the metropolitan-to-nonmetropolitan migration stream and do not, by 
 themselves, account for the turnaround in nonmetropolitan migration. 
 
 Two qualifications need to be made to this conclusion. First, the effect of 
 
IS 
 
 retirement on nonmetropolitan population change may be far greater than the 
 number of relocating retirees would imply, for clearly there can be multiplier 
 effects whereby retirement migration generates employment for persons still in 
 the labor force. Second, the number of household heads who in the survey 
 reported retirement as the main reason for moving may greatly understate the 
 true number of retirees among migrants. An earlier study (Long and Hansen, 
 1979) of Annual Housing Survey data on interstate migrants in the mid-1 970's 
 indicated that the number reporting retirement as the main reason for moving 
 was only about one-sixth as large as the number receiving pension income. 
 Many of the migrants with what appeared to be retirement income gave an 
 employment-related reason for moving or said the move occurred to be closer 
 to relatives, to achieve a change in climate, or for some other consideration 
 rather than retirement as the main reason. The broad conclusion is that many 
 retirees reenter the labor force or for some other reason are difficult to identify 
 separately in surveys as retirees. For these reasons, the effects of retirement on 
 population distribution are difficult to assess accurately. 
 
 A final observation about table 5 is that a fairly substantial number of spatially 
 mobile households did not report a reason for moving or gave a reason that 
 could not be assigned to one of the 30 prelisted categories. Together, these two 
 groups constituted 85,000 household heads moving to nonmetropolitan areas 
 (about 11 percent of the total) and 55,000 household heads moving in the 
 opposite direction (8.5 percent of total households moving to metropolitan 
 areas). These figures seem to imply a net inmigration to nonmetropolitan areas 
 of persons in these residual categories. To investigate this possibility, we 
 examined the "not classified" responses written on the questionnaires of the 
 1979 Annual Housing Survey. We found that a sizeable number expressed what 
 might be considered prorural attitudes: "wanted out of a big city," "wanted a 
 farm," and "wanted to live in the country" were some of the handwritten 
 entries. Among the not classified group we did not find a single case of a 
 person expressing the opposite sentiments that would indicate a preference for 
 living in a big city or a metropolitan environment. This exercise suggested to us 
 the real possibility that nonmetropolitan areas have a small net gain of migrants 
 expressing motivations not represented in the 30 coding categories; such a 
 conclusion serves to underscore the heterogeneity of reasons underlying migra- 
 tion to nonmetropolitan areas. 
 
 MIGRATION OF EMPLOYED HOUSEHOLD HEADS 
 
 In assessing the role of employment considerations in migration between 
 metropolitan and nonmetropolitan areas, it is necessary to control for 
 migrants' employment status. Perhaps the net outmigration from metropolitan 
 areas, as shown in table 5, can be attributed entirely to persons not in the labor 
 force. If so, then one would understandably expect to find a higher representa- 
 tion of noneconomic reasons among households moving from than to metro- 
 politan areas. In order to take these considerations into account, table 6 shows 
 
16 
 
 c 
 
 CD 
 
 W) 
 
 c 
 
 > >> 
 
 o <u 
 
 ^ t 
 
 o ^ 
 
 en c 
 
 «•§ 
 
 C 
 C 
 < 
 
 m 
 
 C ,— 
 
 c c 
 
 Z (u 
 
 __ o 
 
 ^ fl) 
 c ^^ 
 n ^ 
 60 <^ 
 
 ^ c 
 
 -^ CD 
 "O ^ 
 
 t^ c 
 
 O — 
 
 Q^ Q 
 
 2 
 o 
 
 in 
 
 2 -a 
 
 , c 
 
 o o ^ 
 
 oj c "o 
 
 C a) D. 
 
 c~ 2 
 
 O c 2 
 
 2 o o 
 
 ^ 2 
 
 ■O 4) 
 
 rS ^ 
 
 -£= tS 
 
 2 - 
 
 o ^ 
 
 o 
 
 X 
 
 2 2.*^ 
 
 4J c o 
 
 C ra c 
 
 c-^ 2 
 
 o ° ^ 
 
 ^ 2 
 
 lll 
 
 goo 
 ^ 2 
 
 o o ::j 
 
 1. 4-1 ._ 
 
 « c o 
 
 F 5 Ci. 
 
 c ~ 2 
 
 o o 1^ 
 
 Z Q- I 
 
 'I i § 
 
 2 o o 
 
 
 O JD 
 
 ^ 2 
 
 c 
 
 o o 2 
 
 « c "o 
 
 C (TJ Q. 
 
 c~ 2 
 
 O O ^-; 
 
 Z CX 4J 
 
 2 ° 
 
 §11 
 
 ^ 2 
 
 ^^ <N ro VD vo 
 
 r~ r- IT) 
 
 to 00 »n I en 
 
 O CO in r- 
 
 m o ''^ CN <r> 
 
 o in 00 r-^ 
 
 '<^' O d r-^ 
 U-^ CS CO 
 
 O <N VO «— 
 
 lo a\ fn 
 
 CN T- 
 
 I ^ 
 
 r- VO VO VO O ■* 
 00 00 O Tt CN r- 
 Tf CN r- r- 
 
 fO 00 a^ -^ o^ vo 
 m 00 O 'O 
 
 in (N T— T— 
 
 o 
 
 T3 
 
 
 E 
 
 O 
 
 
 
 JZ 
 
 
 2 
 
 2^ 
 
 3 
 
 
 c 
 
 o 
 
 
 s 
 
 JC 
 
 
 nj 
 
 "(5 
 
 
 u 
 
 
 
 *- 
 
 o 
 
 c 
 
 c 
 
 h- 
 
 (U 
 
 c 
 
 !^ 
 
 o 
 _o 
 
 o .E 
 
 o-g I i§ 
 
 c^Z U O 
 
17 
 
 r-; r-^ "^ ^ "^ o^ 
 tT tt CO 1-^ -"t in 
 
 en r- T— 
 
 I— ' CN T-* * ' ""^ O VO rn VX5 CN 
 CN 
 
 CN r- T- r- r- ,— 
 
 oj ^— ^ oq o^ o in in in vo <N p vq p o CN cn p on 
 CNcn inCNr-^r-»r-cN ^— »— inm r- inro 
 
 rn m 
 
 00 in 
 m CN 
 
 ""^T— ^vq^noq•^cnr-Tt•■^mTto^oocn^t 
 v^T— CNVOcnrocN«— rocNCNcn ' inro 
 
 vDa\ooo I cnti-cnt- 
 
 p^ m r— vo 
 
 00Tfinvor^CNcn<x>infnT)-Tfr-vO'<;l-v^ 
 
 lO r- CN T- CO T— 
 
 CN 
 
 ■^vo^tr^ooocNiooococNooin-^oooo 
 
 CO r-OOCNt— >— T— r-T— r- CNt— 
 
 
 9 
 
 
 
 
 
 o 
 
 
 2 
 
 o 
 
 sz 
 
 
 ■M 
 
 3 
 
 
 
 
 o 
 
 c 
 
 c 
 .2 
 1^1 
 
 
 
 c 
 •a 
 c 
 
 
 a> 
 
 .2 
 
 E 
 
 iS 
 
 <t> 
 
 BO 
 
 >- 
 
 o 
 
 f« 
 
 o 
 
 (TJ 
 
 JZ 
 
 C 
 
 F 
 
 h-^ 
 
 
 SO 
 
 3 
 
 n 
 
 
 
 
 
 
 o 
 
 4) -a 
 o t 
 
 -= c 
 
 w) o 
 «j o 
 
 O u 
 
 2 = 
 o o 
 
 II 
 
 SZ (U 
 
 .— (/I 
 
 o e 
 
 o 
 
 4-> U 
 
 a> 
 
 
 « JZ c <+- tz ti 
 
 Z b^ ^ o -j^ o 
 
 C 4) ro Q. 
 
 ^ M- « 
 
 a> £ O O 
 
 < Ol U Z Z 
 
 O -o 
 
 o O 
 <u 00 
 jO a 
 
 E a. 
 « I 
 
 I§ 
 
 1- « 
 
 o£ 
 
 z I 
 
18 
 
 the same type of data as table 5 but controls for whether the household head 
 had a job in the week preceding the survey. 
 
 Table 6 demonstrates that even among household heads who had a job in the 
 week preceding the survey, metropolitan areas (as defined in 1970) appeared to 
 have had net outmigration. An estimated 533,000 household heads so em- 
 ployed moved out of metropolitan areas in the 12 months preceding the 1975 
 survey, compared with 481,000 who moved in. Metropolitan areas also appear 
 to have had net outmigration of household heads who did not have a job in the 
 week preceding the survey; of these nonworking household heads (some of 
 whom were unemployed and some simply not in the labor force), an estimated 
 237,000 left metropolitan areas and only 1 70,000 moved to them. 
 
 Among employed household heads moving for employment-related reasons (as 
 identified in table 6), there is an approximate balance in the flow between 
 metropolitan and nonmetropolitan areas. According to the table, among 
 employed household heads (i.e., with a job in the week preceding the 
 survey), the number moving to metropolitan areas— an estimated 286,000— 
 approximately equalled the number leaving metropolitan areas— an estimated 
 288,000— in the 12 months preceding the 1975 Annual Housing Survey. These 
 data mean that the overall net migration to nonmetropolitan territory seems to 
 be essentially the product of: (1) persons not working (columns 3 and 4 of 
 table 6) and (2) employed persons moving for reasons not directly employment 
 related. The nonemployment reasons that induce employed household heads to 
 move to nonmetropolitan areas encompass a wide variety of factors associated 
 with housing aspirations and dissatisfaction with the metropolitan neighbor- 
 hood from which the household moved. Limitations of sample size generally 
 prevent more precise characterization of these broad sets of reasons. 
 
 Among employed household heads, however, employment-related reasons 
 account for at least one-half of migration between metropolitan and nonmetro- 
 politan destinations. As shown in table 6, 54.0 percent of employed household 
 heads moving to nonmetropolitan areas cited one of the employment-related 
 reasons, as did 59.5 percent of employed household heads moving in the 
 opposite direction. Even when we limit the comparison to employed household 
 heads, the data suggest that noneconomic considerations are somewhat more 
 important in moves from than moves to metropolitan areas. 
 
 DISTANCE FROM A METROPOLITAN AREA 
 
 Some migrants leaving metropolitan areas do not go far. In fact, some stay 
 close enough to commute into metropolitan areas for employment, and they 
 and others who do not move far from metropolitan areas live in territory that 
 might be called exurban or some other term that connotes a residence slightly 
 beyond established suburban developments and not clearly metropolitan or 
 nonmetropolitan in character but becoming part of the suburban fringe of 
 expanding metropolitan areas. Most metropolitan areas are spatially expanding; 
 
19 
 
 even some metropolitan areas that decreased in population in the 1970's had 
 population growth in their outer counties and in this way expanded outward 
 into nonmetropolitan territory (Long and Dahmann, 1980). 
 
 The effect of the exurbanites on metropolitan-nonmetropolitan contrasts is 
 sometimes unclear, for surveys tend to be tabulated into categories that may 
 suggest a sharper distinction than actually exists. Persons moving to territory 
 statistically defined as nonmetropolitan but lying very close to a metropolitan 
 area may be moving for the same reasons as city-to-suburb movers and in other 
 ways might be thought of as suburbanites who are merely moving slightly 
 farther than others from an urban core. If so, then it is hardly surprising to find 
 what seems to be a large proportion of reasons that appear to be noneconomic 
 (at least not directly job related) among metropolitan-to-nonmetropolitan 
 migrants, for many past studies have sought to explain suburbanization in 
 terms of a desire for homeownership, more space, better schools for children, 
 or other essentially noneconomic motivations (Goodman, 1979; Spain, 1980). 
 
 In an attempt to test the hypothesis that migrants to nonmetropolitan territory 
 adjacent to metropolitan areas move for the same reasons as city-to-suburb 
 movers, we constructed table 7, which shows the distribution of reasons 
 for moving among city-to-suburb movers and three groups of households 
 leaving metropolitan areas as defined in 1970. For the metropolitan-to- 
 nonmetropolitan movers we show: (1 ) those who went to counties that became 
 parts of metropolitan areas between 1970 and 1975; (2) those who probably 
 moved slightly farther out, to counties not incorporated into but adjacent to 
 metropolitan areas as defined in 1975; and (3) those who moved still farther 
 out, to nonmetropolitan counties not adjacent to metropolitan areas. 
 
 Most households moving from metropolitan areas go to counties either adja- 
 cent to metropolitan areas or no more than one county away from metro- 
 politan areas. From table 7 one can conclude that among households leaving 
 SMSA's (as defined in 1970) in the 12 months preceding the 1975 survey, 
 about 20.3 percent went to counties incorporated into metropolitan areas 
 between 1970 and 1975; another 48.3 percent went to counties still nonmetro- 
 politan in 1975 but adjacent to metropolitan areas whose boundaries had been 
 updated to 1975. Altogether, 68.6 percent of the nonmetropolitan-bound 
 households went to counties very recently redefined as belonging to SMSA's or 
 else adjacent to redefined SMSA's. These data should not be interpreted, how- 
 ever, as proving that the nonmetropolitan migration turnaround is explainable 
 as metropolitan spillover, for past analyses have shown that although about 60 
 percent of the net inmigration to nonmetropolitan territory went to the adja- 
 cent counties, the nonadjacent counties experienced a turnaround from net 
 outmigration in the 1960's to net inmigration in the 1970's. This change in 
 pattern in the nonadjacent nonmetropolitan counties is generally accepted as a 
 demonstration that new forces seem to be governing population redistribution 
 to the nonmetropolitan sector in the 1970's (Beale, 1975). 
 
 The purpose of table 7 is to see if different motives govern movement from 
 
20 
 
 Table 7. Reasons Reported by Household Heads for Moving Fronri Cities to 
 
 Suburbs and for Moving Out of Metropolitan Areas in the 12 Months 
 Preceding the 1975 Annual Housing Survey, According to an Indicator 
 of Distance From Metropolitan Areas 
 
 (Percent) 
 
 
 
 Movers from 
 
 SMSA's as defined in 1970 
 
 
 Movers* from 
 central cities 
 
 
 
 
 
 To counties 
 
 To counties 
 
 To counties 
 
 
 to balance of 
 
 added to 
 
 adjacent to 
 
 not adjacent 
 
 
 SMSA as 
 
 SMSA's 
 
 SMSA's as 
 
 to SMSA's 
 
 
 defined in 
 
 between 1970 
 
 defined in 
 
 as defined 
 
 Main reason for moving 
 
 1970 
 
 and 1975 
 
 1975 
 
 in 1975 
 
 Household heads (thous.) 
 
 1,003 
 
 156 
 
 372 
 
 240 
 
 Employment 
 
 8.3 
 
 34.6 
 
 35.2 
 
 53.1 
 
 Job transfer 
 
 .6 
 
 19.2 
 
 11.0 
 
 19.1 
 
 New job or looking 
 
 
 
 
 
 for work 
 
 2.0 
 
 12.2 
 
 22.8 
 
 30.7 
 
 Commuting 
 
 4.5 
 
 3.2 
 
 1.1 
 
 - 
 
 Other 
 
 1.2 
 
 - 
 
 .3 
 
 3.3 
 
 Family 
 
 24.6 
 
 15.4 
 
 21.0 
 
 13.7 
 
 To be closer to 
 
 
 
 
 
 relatives 
 
 ■■■'"'•'* ■■'\9" ■'■■■'^ 
 
 ■■■■' ^;i'-''*-'' ■ 
 
 '^"■■'■■^T3""'''- ^ 
 
 7.1 
 
 Marriage and house- 
 
 
 
 
 
 hold formation 
 
 16.0 
 
 7.7 
 
 .7.3 
 
 1.7 
 
 Marital dissolution 
 
 4.3 
 
 1.9 
 
 1.6 
 
 2.1 
 
 Other 
 
 3.5 
 
 '' .6 
 
 #.» 
 
 2.9 
 
 Housing and neighbor- 
 
 
 
 
 
 hood 
 
 52.5 
 
 22A 
 
 17.2 
 
 7.8 
 
 Larger house or 
 
 
 
 
 
 apartment 
 
 12.1 
 
 2.6 
 
 4.0 
 
 .4 
 
 To own residence 
 
 14.0 
 
 6.4 
 
 1.9 
 
 .4 
 
 Lower rent/less 
 
 
 
 
 
 expensive house 
 
 6.1 
 
 2.6 
 
 3.0 
 
 2.5 
 
 Better house/more 
 
 
 
 
 
 conveniences 
 
 8.2 
 
 1.9 
 
 1.1 
 
 .8 
 
 Neighborhood 
 
 
 
 
 
 dissatisfaction 
 
 8.7 
 
 5.1 
 
 4.6 
 
 3.3 
 
 Miscellaneous 
 
 3.6 
 
 3.8 
 
 2.7 
 
 .4 
 
 Enter or leave Armed 
 
 
 
 
 
 Forces 
 
 ' s ■ ' 
 
 1.3 
 
 1.1 
 
 5.4 
 
 Attend school 
 
 .5 
 
 4.5 
 
 7.3 
 
 2.9 
 
 Retirement 
 
 .6 
 
 5.8 
 
 5.6 
 
 5.0 
 
 Change of climate 
 
 .8 
 
 1.3 
 
 2.2 
 
 2.5 
 
 Not classified 
 
 10.0 
 
 12.2 
 
 5.6 
 
 7.1 
 
 Not reported 
 
 2.5 
 
 3.2 
 
 4.8 
 
 2.1 
 
 —Indicates no sample cases fell in this category. 
 * Excludes intermetropolitan migrants. 
 
21 
 
 cities to suburbs and from metropolitan areas to adjacent and nonadjacent 
 nonmetropolitan counties. The table indicates that these different groups of 
 movers cite substantially different reasons for moving. For one, movement to 
 exurbia is more strongly governed by employment-related considerations than 
 movement to suburbia. Only 8.3 percent of movers from cities to suburbs cited 
 one of the employment-related reasons compared with around 35 percent of 
 the movers to the two exurban categories (columns 2 and 3 of table 7). Con- 
 versely, the housing and neighborhood reasons that account for 52.5 percent of 
 the city-to-suburb moves account for only 22.4 percent and 17.2 percent of 
 moves to the two exurban groups of counties. Just why these differences 
 should occur between the suburban and exurban movers is not clear. Perhaps 
 the exurban movers simply follow jobs that decentralize into the countryside. 
 Even the exurban movers who commute into metropolitan areas for work (data 
 not shown) appeared more likely to report one of the employment-related 
 reasons for moving than the city-to-suburb movers. In general these types of 
 data indicate that households moving to the fringes of metropolitan areas do so 
 for a somewhat greater variety of reasons than reported by households moving 
 to more traditional suburban areas. 
 
 It may come as a surprise to find that employment-related reasons account for 
 a majority of households moving to the most remote group of counties, those 
 not adjacent to SMSA's as defined in 1975. These counties have had high rates 
 of outmigration in the past (Beale, 1975), and there is a temptation to think of 
 them as offering few employment opportunities. Explanations of their growth 
 in the 1970's (e.g., Morrison, 1976) have stressed their attractions to retirees 
 and their recreational opportunities, which probably provide jobs for local 
 residents rather than pull in jobseekers from urban areas. But the remote 
 nonmetropolitan counties are a heterogeneous group that includes many areas 
 with renewed coal mining and other newly developed employment opportuni- 
 ties. At any rate, a majority (53.1 percent) of households moving from metro- 
 politan areas to the nonadjacent nonmetropolitan counties reported 
 employment-related reasons. This seems like a higher representation of 
 employment-related reasons for moving than for any of the other groups of 
 migrants examined so far, although limitations of sample size prohibit firm 
 conclusions. 
 
 Retirees might be expected to be relatively more numerous among migrants to 
 the nonadjacent than to the adjacent counties, based upon analyses that have 
 emphasized the footlooseness of the growing number of retirees and their 
 apparent preference for rural settings (Beale, 1975; Morrison, 1976). Actually, 
 the proportion of retirees among migrants to the nonadjacent nonmetropolitan 
 counties (5.0 percent) does not exceed the proportion among migrants to 
 nonmetropolitan counties adjacent to SMSA's (5.6 percent). The conclusion is 
 simply that like most other persons moving out of metropolitan areas, the 
 majority of retirees go to counties no more than one county away from metro- 
 politan areas. 
 
22 
 
 Migrants from metropolitan areas wiio report employment-related reasons for 
 moving appear to be somewhat more likely than others to go to one of the 
 relatively remote nonadjacent counties. Data compiled for table 7 indicate that 
 about 41 percent of households reporting employment-related reasons for 
 moving from SMSA's went to nonadjacent counties, compared with 24 percent 
 for those reporting family reasons and 16 percent for those reporting housing 
 and neighborhood reasons. The sample sizes are too small to permit firm con- 
 clusions, but these comparisons suggest the expected: those who leave metro- 
 politan areas for the types of reasons most commonly cited by city-to-suburb 
 movers are likely to stay closer to metropolitan areas than those who look for 
 work or take jobs in nonmetropolitan settings. 
 
 TYPES OF MIGRATION 
 
 Up to now, we have identified only one group of movers— those leaving metro- 
 politan areas for nonmetropolitan counties not adjacent to metropolitan 
 areas— for whom the employment-related reasons accounted for a majority of 
 moves. Less than 50 percent of the movers to or from metropolitan areas (as 
 defined in 1970) reported one of the employment-related reasons identified in 
 the 1975 Annual Housing Survey. In an attempt to gain a fuller perspective on 
 the role of economic reasons in accounting for different types of migration, 
 table 8 was produced. It shows the distribution of main reasons for moving for 
 each of four major types of migration: between metropolitan areas, between 
 nonmetropolitan counties, and the two already discussed (from nonmetro- 
 politan to metropolitan areas and from metropolitan areas to nonmetropolitan 
 territory). Reasons for moving as reported by several groups of intra-area 
 movers are also shown. 
 
 For none of the four types of migration did the employment-related reasons 
 account for a majority of moves. The percent of household heads citing one of 
 the employment-related reasons was 46.1 among intermetropolitan migrants, 
 41.9 among those migrating between nonmetropolitan counties, 48.5 among 
 those going from nonmetropolitan to metropolitan areas, and 40.7 among 
 those leaving metropolitan areas for nonmetropolitan locations. Differences 
 among these four types of movers in the percent citing the employment-related 
 reasons for moving are not always statistically significant, and the major con- 
 clusion is that these reasons explain less than one-half of migration within the 
 United States, according to the definitions of migration shown in the first four 
 columns of table 8. 
 
 No easily identifiable set of reasons explains a majority of these moves. The 
 several family-related reasons explain between 15.6 percent and 19.4 percent 
 of the four types of migration, and various housing and neighborhood consider- 
 ations explain another 7.1 to 15.8 percent of moves between metropolitan 
 areas, between nonmetropolitan counties, or between metropolitan and non- 
 metropolitan destinations. All things considered, the data seem to suggest a 
 
23 
 
 heterogeneous set of factors and a diversity of motivations underlying each of 
 the four types of movement. 
 
 Exact comparisons of these results with past nationwide surveys of reasons for 
 moving are impossible because of differences in the universe to whom ques- 
 tions were asked and differences in the way questions were asked and answers 
 were coded. The earlier national surveys generally concluded that employment 
 considerations constituted the major motivation underlying most moves be- 
 tween counties or economic areas (U.S. Bureau of the Census, 1947, 1966; 
 Lansing and Mueller, 1967). Such a conclusion does not seem fully consistent 
 with the present results, but because of lack of comparability among the sur- 
 veys, it is impossible to demonstrate conclusively that reasons for moving have 
 changed. They probably are changing, however, because of changes in the 
 composition of the work force (especially the increase in the number and 
 proportion of women workers), changes in household composition (e.g., fewer 
 persons per household and more households maintained by single parents), 
 occupations and age of workers, and retirement policies and government pro- 
 grams like unemployment compensation. The net effect of these changes seems 
 to imply a shift of reasons for moving away from many traditional economic 
 motivations to a more heterogeneous set of factors (Long and Hansen, 1979). 
 
 Distance probably positively affects the likelihood of citing one of the 
 employment-related reasons. The employment-related reasons in table 8 were 
 reported by 50.8 percent of households moving between States in the 
 mid-1 970's and by an even higher proportion of interregional migrants (Long 
 and Hansen, 1979). As commuting fields have expanded, more households have 
 been able to move from one county to another or to make other types of 
 interarea moves that do not necessarily entail job changes as they once did. 
 
 Most strictly local moves can be attributed to one of the family, housing, or 
 neighborhood reasons. In fact, the combination of family reasons and the 
 housing and neighborhood reasons explained 77.1 percent of cIty-to-suburb 
 moves, 65.9 percent of suburb- to-city moves, 80.8 percent of moves within 
 cities, and 78.6 percent of moves within nonmetropolitan counties. Differences 
 among these percentages are not always statistically significant, and the simple 
 conclusion is that the vast majority of each of these types of local moves can 
 be attributed to one of the family, housing, or neighborhood considerations. 
 
 CONCLUSION 
 
 Annual estimates of population changes for individual counties Indicate that 
 the shift of population growth toward nonmetropolitan areas, first observed in 
 the early 1970's, was not measurably slowed by energy developments In the 
 4 years following the 1973-74 oil embargo. On the contrary, the net in- 
 migration to nonmetropolitan counties beginning around 1970 seems to have 
 established a momentum that added population to the nonmetropolitan sector 
 at an even faster pace, relative to metropolitan areas, in 1974-78 than in 
 
24 
 
 o 
 
 c 
 
 JO 
 
 ■a 
 
 60 
 
 C 
 
 n 
 U 
 
 ■o 
 
 trj 
 <U 
 X 
 
 O 
 
 o 
 X 
 >^ 
 
 JO 
 
 -a 
 
 (U 
 
 tl 
 o 
 o. 
 <t> 
 c^ 
 
 60 
 C 
 '> 
 
 o 
 
 c ^ 
 
 - 1 
 
 C^ 60 
 
 SI 
 = 1 
 
 i-l 
 
 ^ 60 
 D. C 
 
 E -o 
 
 o ^ 
 o i 
 
 00 
 
 -Si 
 
 r9 
 
 s = >• 
 
 Ji 5 ^ 
 
 ^ '-^ = 
 
 c o o 
 
 o Q. o 
 
 .E o ui 
 
 '•Z. "^^ ^ 
 
 
 
 
 5 ^ 
 0) O o 
 
 « .ti ^ T^ 
 
 ^00°. 
 
 Q- c 
 
 Is 25 
 
 t +j o rr 
 O O ^ CL 
 Z ex 
 
 J- O _ (/I 
 
 o> flj j2 *-• 
 
 « =1 o 
 
 CO O °- o 
 
 fO 
 
 (/) 
 
 <i> ^ ^ n 
 
 ._ c 
 
 (/) — 
 
 rfl Q. 
 
 2i E 
 
 r- « 
 
 
 <N 00 U-i CM VO 
 t>-* * CM* CN r-* 
 
 
 
 '^. '^ *-i "^ '^ ^. ^. "^ ^ ^. ^. ^ ^. ^ '^. 
 r-* T-* 00 Tt rn T-^ <N O vd CN CN vD 
 
 CN r- ly^ T- r- ,- 
 
 '«T^*o^a^o^'-;a^■^CNVO^~.T-;mp-.r«.o^^— CNcncN 
 \£ * 1-^ CN * o * ON in "t 00 CN o r^' (50 Tt in 
 
 en to CN CN o vo 00 vo vo 10 (^ O 00 r-^ ■^^ cn in rn 
 r-T— r-fOr-^Tfcnfor^r«^r^oocNr~>*oo 
 
 CN T- m r- r- 
 
 >X) in CN r-; -^^ oj CN r— o r- ^^ r-^ cn r-; <n t^ cn oj cn cn on 
 rn * cn ' 
 
 cn cn vo o in CN vo ON O cn »n in «— . O 
 § 00* * CN "t' f-' Tt 
 
 CN r-; vo cn in >^, ■ 
 vOTfcncNCN'^tvooooorn 
 
 O r-; rn o CN CN r-; 00 vo ON "^ cn r-; cn r^ CN rn r- in in vo 
 l!!!^ o in cn T- r-' r^' vd in r- cn in CN CN CN T-' 'sj- CN CN m* in 
 
 oovovocncNoovoinincNf^ 
 
 •^ T— CN r- 
 
 CN T- O r- r- CN r- CN ON VO 00 r- 00 
 CN >— ' r-* * ' "^' ON r- 
 
 Tl- ONONr-^r-r^Tj^-^^^OOinOOVDOONU^ONOTtCNVD 
 
 2 »— r-cnrtcNON'^oocncnincNcn»-^cncNCN ooi— 
 
 °° "t T- CN »- r- 
 
 r- 10 o CN CN vq vo 10 CN cn Tj; Tf 00 00 o CM CN cn cn o 
 vd r-^ C3 CM CM in ■^' Tt cn cn r-* cs r- * r-^ "«t r^ -^ r-^ CM 
 
 -it CM 04 r- r- 
 
 c ^ 
 
 u 
 
 ?Ǥ 
 
 D 
 
 E - 
 E " 
 o 
 
 ^ >. 
 
 I- -a Q 
 
 I si 
 
 U Ofi _ 
 
 <u .2 iS «- 
 
 -" »- I- r- 
 
 o 
 
 
 U O E H S S O 
 
 00 O 
 
 (« o 
 
 (V Q 
 
 2 ^ 
 
 O D < 
 
 o "Sis 
 
 .2? 1^ o ^ E 
 
 <u ": «_ c J' 
 
 Z 2 « « ~ 
 
 c t5 « 
 
 UJ < 0^ 
 
25 
 
 '". i^. "*. 
 
 CN »-^ 00 
 
 * O CO 
 
 CN 00 a\ 
 
 ' ON en 
 
 "*. *'°. "^ 
 ' p-* en 
 
 00 o in 
 ' O CN 
 
 '": ^. '^. 
 CN r>.' en 
 
 00 -<*, r- 
 
 1-^ «n CO 
 
 Tt Tt q 
 r-* --t r-* 
 
 vo in q 
 fO «X -^ 
 
 Tf vo q o oq 00 oj ON r-. Tt q ON 00 r-; cn r^^ oq oq r- t— t-^ r-^ q q 
 
 ^ ^^A #-^\ ^s.1 r^ /Vk v4- /«^ /■~^ «*^v 1^^ ^_ ^.1 i^-v ^''"S /K.I 
 
 r^ 0\'^mOfnr-inr-CN«n 
 CN r- U-> 1 — 
 
 oor-cooN'— >ocNr~r-;T-;0\r-;iovqoNvo 
 
 * CN CO * O * O "«t Tf ON CN IT-' VO 00 '<t "^ 
 
 I ^. 
 
 ir)VOioocnr-;qp«. 
 
 U-i * r-^ CN r-^ Tl- t-^ in 
 
 rn CN 
 en r^ 
 
 -r-cNoqvorncNvo 
 It o r-^ T-^ r^ vo 
 
 CN oq Tf 
 * o\ cn 
 
 CN q 
 
 ON cn 
 
 ■>!tcNcncNvqr^CN«— inoNoqvoqincsvooNCSvou-) 
 ON *'^rn?-^r- * "or^cnvovooor^vorrm *»— 
 r- t— cn cs cn 
 
 yo CN 
 
 irj r-> 1— cn -^ r- CN CN CN in rn in vo q ir^ oq r-; 
 ON *CNin»-in 'r^'<l-mcNCNin«nr-*oocN 
 
 CN r- m t— r- 
 
 CN CN r-_ irj ON 
 
 cn q in oq r^ '—. i^ ^, '^. ^. ^. ^. <» ■^. ^. 
 ?2 "^' O o r-^ t--* rn rn vo' r-^ cn id cn cn cn 
 
 ''' LTV rsl (T» __ n 
 
 ^^cn^ONOOrn^ 
 cn cn CN rn ' in cn 
 
 r- in O ■* CN CJN ON 
 
 00 
 
 oqoNqinu^iniocNqvoqqcNCNqrn 
 CN «— »— in«n t— men 
 
 ■^ooor-mooinmONio 
 
 ® '^ ^ '"I '^l T "'J ^. "^ ^. f^. ^ "^ *'°. • ; ^. -% ■ . 
 
 °9 cn in o^ «n rn m* cn oo' cn i— * m' cn cm* r-' -"t cn cn cn cn vo 
 
 •^ m t- CN r- ,- 
 
 00 <^ fN oq q CN oq cn Tf CN oq vo r-; Tf vo r- r-^ r- oq r-; rf vo r- oq 
 £^ r^oocncncNCNCNincNCNOcNCN r-cN«--'^m 'cNincN 
 
 '^ in CN CN »- r- 
 
 -^1^ 
 
 *= X 
 
 ^ 3 
 
 ^o 
 
 _ •"— ' c. •- «-\ ■*-• 
 
 C 
 
 o t: 
 
 P ° 3 -5, o 
 
 fli *-• ^ (rt ^ tS 
 
 .a E 
 
 > £ 
 
 > o 
 o 
 
 00 
 
 «8 
 g S 5 
 
 « X £ -. 
 
 c: i" c ^^ 
 "* c •.-. 
 
 S 
 
 
 £ o o 
 
 u z z 
 
 It 
 
 E 
 
 O 60— C -C 
 
 « .2 5 t- "> Sj 
 
 Q «^ ;_ Q rd ro 4^ 
 
 « C 
 
 ZUO gH^SO 
 
 »- J= O CO 
 
 I- l_ -O — 
 
 O <u <u •- 
 
 -J 00 Z ^ 
 
 
 C 14- 
 
 Q> O 
 
 E a> 
 
 S « -^ S 
 
 -•-< j_t 4-1 re 
 
 C *- « J= 
 
 LU < Q^ (J 
 
 it tj 
 
 U I- 
 
 O O 
 
 z z 
 
 tj o 
 
 V 00 
 
 Si « 
 
 2 c 
 
 O V 
 
 r- <" 
 
 >. " 
 
 re Oj 
 
 E Q. 
 
 o </> 
 
 ^_3 
 
26 
 
 1970-74. This conclusion applies regardless of whether one uses the 1970 or 
 1 980 definitions of metropolitan and nonmetropolitan. 
 
 The momentum to population growth in the nonmetropolitan sector results 
 from the Interaction of net migration and natural increase. Between 1970 and 
 1978 the crude rate of natural increase (crude because it is not standardized for 
 age) fell somewhat less in the nonmetropolitan than in the metropolitan sector, 
 and it probably did so in part because of the tendency of net Inmigration to 
 bring in persons in their reproductive years. Through such a process, net in- 
 migration In one period of time can boost rates of natural increase in later 
 periods. 
 
 Because of this kind of "automatic" mechanism, nonmetropolitan areas in the 
 aggregate are likely to continue, for a while anyway, to have higher rates of 
 population growth than metropolitan areas. An abrupt reversal to the previous 
 pattern of a higher growth rate in metropolitan areas is therefore unlikely. 
 
 Other considerations also suggest a momentum behind nonmetropolitan 
 growth. Past trends toward decentralization of employment (see, e.g., U.S. 
 Bureau of the Census, 1972) have resulted in extensive movement of jobs away 
 from large cities, implying that a rapid recentralization of population would 
 not necessarily reduce commuting distances to an appreciable degree in the 
 short run. Moreover, the convergence of metropolitan and nonmetropolitan 
 income levels (Zuiches and Brown, 1978) suggests that nonmetropolitan 
 residents may be better able now than in the past to absorb higher commuting 
 costs or make other adjustments (e.g., buying smaller cars or better home „. 
 insulation) to rising energy costs. 
 
 To a very large extent, the duration of the faster rate of population growth in 
 the nonmetropolitan sector depends upon how metropolitan and nonmetro- 
 politan are conceptualized and how these concepts are put into practice. One 
 of the ironies of the present is that the Nation is still metropolitanizing but 
 only because of population growth in nonmetropolitan areas. This paradox 
 results from the fact that all of the increase between 1970 and 1978 in the 
 percent of the population defined as metropolitan is attributed to the growth 
 of nonmetropolitan cities and towns into metropolitan areas and to fusion of 
 other nonmetropolitan counties with existing metropolitan areas. When the 
 1980 census results become available, more nonmetropolitan counties will be 
 redefined as metropolitan, and the result is that the momentum to nonmetro-l 
 politan population growth ultimately leads to metropolitan population growth.j 
 
 REFERENCES 
 
 Beale, Calvin L. 1975. 77?^ Revival of Population Growth in Nonmetropolitan 
 America. Washington, D.C.: Economic Research Service, U.S. Department of 
 Agriculture. 
 
 1976. "A Further Look at Nonmetropolitan Population Growth Since 
 
 1970." American Journal of Agricultural Economics 58: 953-958. 
 
27 
 
 1977. "The Recent Shift of United States Population to Nonmetro- 
 
 politan Areas, 1970-75." International Regional Science Review 2: 113-122. 
 
 Blundell, William E. July 3, 1980. "New Rural Migration Overburdens and 
 Alters Once-Sleepy Hamlets." The Wall Street Journal. 
 
 Bowles, Gladys K. 1978. "Contributions of Recent Metro/Nonmetro Migrants 
 to the Nonmetro Population and Labor Force." Agricultural Economics 
 Research 30 (October): 15-21. 
 
 Carpenter, Edwin H. 1977. "The Potential for Population Dispersal: A Closer 
 Look at Residential Locational Preferences." Rural Sociology 42: 352-370. 
 t^DeJong, Gordon P., and Kenneth G. Keppel. 1979. Urban Migrants to the 
 Countryside. University Park, Pa.: Agricultural Experiment Station, Pennsyl- 
 vania State University. 
 
 Goldstein, Sidney. 1976. "Facets of Redistribution: Research Challenges and 
 Opportunities." Demography 1 3: 423-434. 
 
 Goodman, John L. 1979. "Reasons for Moves Out Of and Into Large Cities." 
 Journal of the American Planning Association 45 : 407-41 6. 
 
 Kain, John F. 1975. "Implications of Declining Metropolitan Population on 
 Housing Markets." In George Sternlieb and James W. Hughes, eds., Post- 
 Industrial America: Metropolitan Decline and Inter-Regional Job Shifts, pp. 
 221-227. New Brunswick, N.J.: The Center for Urban Policy Research, 
 Rutgers University. 
 
 Lansing, John B., and Eva Mueller. 1967. 77?^ Geographic Mobility of Labor. 
 Ann Arbor: Survey Research Center, University of Michigan. 
 
 Long, John F. Forthcoming. Population Deconcentration in the United States. 
 U.S. Bureau of the Census. 
 
 Long, Larr-y H., and Donald C. Dahmann. 1980. 77?^ City-Suburb Income Gap: 
 Is It Being Narrowed by a Back-to-the-City Movement? Special Demo- 
 graphic Analyses, CDS-80-1, U.S. Bureau of the Census. Washington, D.C.: 
 U.S. Government Printing Office. 
 
 Long, Larry H., and Paul C. Click. 1976. "Family Patterns in Suburban Areas: 
 Recent Trends." In Barry Schwartz, ed., 77?^ Changing Face of the Suburbs, 
 pp. 39-67. Chicago: University of Chicago Press. 
 
 Long, Larry H., and Kristin A. Hansen. 1979. Reasons for Interstate Migration: 
 Jobs, Retirement, Climate, and Other Influences. Current Population 
 Reports, series P-23, no. 81, U.S. Bureau of the Census. Washington, D.C.: 
 U.S. Government Printing Office. 
 
 McCarthy, Kevin F., and Peter A. Morrison. 1978. "Post-1970 Demographic 
 Change in Nonmetropolitan Areas: Insights from the 10-Percent CWHS." in 
 U.S. Department of Health, Education, and Welfare, Policy Analysis with 
 Social Security Research Files, pp. 635-660. Washington, D.C.: U.S. Govern- 
 ment Printing Office. 
 
 Morrill, Richard L. 1978. "Population Redistribution, ^965-l 5." Growth and 
 Change 9 (April): 35-43. 
 
 Morrison, Peter A. 1976. Rural Renaissance in America? The Revival of Popula- 
 tion Growth in Remote Areas. Washington, D.C.: Population Reference 
 Bureau. 
 
n 
 
 Munick, Warren A., and Dennis Sullivan. 1977. "Race, Age, and Family Status 
 
 Differentials in Metropolitan Migration of Households." Rura/ Sociology 42: 
 
 536-543. 
 Phillips, Phillip D., and Stanley D. Brunn. 1978. "Slow Growth: A New Epoch 
 
 of American Metropolitan Evolution." Geographical Review 68: 274-292. 
 Ploch, Louis A. 1978. "The Reversal in Migration Patterns— Some Rural 
 
 Development Consequences." Rural Sociology 43: 293-303. 
 Regional Economic Analysis Division. 1978. "Work-Force Migration Patterns, 
 
 1960-73." Survey of Current Business 56 (October): 23-28. 
 Renshaw, Vernon, Howard Friedenberg, and Bruce Levine. 1978. "Work-Force 
 
 Migration Patterns, 1970-76." Survey of Current Business 58 (February): 
 
 1 7-20. 
 Roseman, Curtis C. 1977. Changing Migration Patterns Within the United 
 
 States. Washington, D.C.: Association of American Geographers. 
 Spain, Daphne. 1980. "Reasons for Intrametropolitan Mobility: Are Schools a 
 
 Key issue?" Review of Public Data Use 8: 59-67. 
 Taeuber, Irene B. 1972. "The Changing Distribution of the Population of the 
 
 United States in the Twentieth Century." In Sara Mills Mazie, ed.. Population 
 
 Distribution and Policy, pp. 29-107, Research Reports of the Commission on 
 
 Population Growth and the American Future, vol. 5. Washington, D.C.: U.S. 
 
 Government Printing Office. 
 U.S. Bureau of the Census, 1947. Postwar Migration and Its Causes in the 
 
 United States. Current Population Reports, series P-20, no. 4. Washington, 
 
 D.C.: U.S. Government Printing Office. 
 1966. Reasons for Moving: March 1 962 to March 1963. Current Popu- 
 lation Reports, series P-20, no. 154. Washington, D.C.: U.S. Government 
 
 Printing Office. 
 1972. "Employment and Population Changes— Standard Metropolitan 
 
 Statistical Areas and Central Cities." Special Economic Reports, series 
 ES20(72)-1. Washington, D.C.: U.S. Government Printing Office. 
 1973. "Census Bureau Reports Migration Loss in Nation's Metropolitan 
 
 Areas During Last Three Years." Press release CB73-338. 
 
 .. 1 975, Mobility of the Population of the United States: March 1970 to 
 
 ^ March 1975. Current Population Reports, series P-20, no. 285. Washington, 
 D.C.: U.S. Government Printing Office. 
 
 1977a. General Housing Characteristics for the United States and 
 
 Regions. Current Housing Reports, H-1 50-75, part A. Washington, D.C.: U.S. 
 Government Printing Office. 
 1977b. Geographical Mobility: March 1975 to March 1976. Current 
 
 Population Reports, series P-20, no. 305. Washington, D.C.: U.S. Government 
 Printing Office. 
 
 .. 1 980. Estimates of the Population of Counties and Metropolitan Areas: 
 
 July 1, 1977 and 1978. Current Population Reports, series P-25, no. 873, 
 Washington, D.C.: U.S. Government Printing Office. 
 U.S. Department of Housing and Urban Development. 1978. 77?^ President's 
 
29 
 
 National Urban Policy Report. Washington, D.C.: U.S. Government Printing 
 
 Office. 
 Voss, Paul R., and Glenn V. Fuguitt. 1979. Turnaround Migration in the Upper 
 
 Great Lakes Region. Madison, Wis.: University of Wisconsin-Extension. 
 Wardv^ell, John M., 1977. "Equilibrium and Change in Nonmetropolitan 
 
 Growth." Rural Sociology 42: 156-179. 
 Wardwell, John M., and C. Jack Gilchrist. 1980. "Employment Deconcen- 
 
 tration in the Nonmetropolitan Migration Turnaround." Demography 17: 
 
 145-158. 
 Williams, James D., and Andrew J. Sofranko. 1979. "Motivations for the 
 
 Inmigration Component of Population Turnaround in Nonmetropolitan 
 
 Areas." Demography 1 6: 239-255. 
 Zuiches, James J., and David Brown. 1978. "The Changing Character of the 
 
 Nonmetropolitan Population, 1950-1975." In Thomas R. Ford, ed.. Rural 
 
 U.S.A.: Persistence and Change, pp. 55-72. Ames, Iowa: Iowa State Uni- 
 versity Press. 
 
 6 U. S. GOVERNMENT PRINTING OFFICE : 1980 340-997/1313 
 
PENN STATE UNIVERSITY LIBRARIES 
 
 Superintendent of Documents 
 U.S. Government Printing Office 
 Washington, D.C. 20402 
 
 Official Business 
 
 Penalty for Private Use, $300 
 
 ADDDD7mtlin 
 
 ui i^urntnerce ■ ^^K^ 
 
 I U.S.MAIL 
 
 COM-202 V^— — "^ 
 
 Special Fourth-Class 
 — Rate-Book