Internal Migration and County Population Dynamics for Michigan Counties

JRAP 42(1): 61-86. 2012 MCRSA. All rights reserved. Internal Migration and County Population Dynamics for Michigan Counties David J. Sorenson # and Ronald J. Gunderson* Augustana College (SD) #, Northern Arizona University* USA Abstract. During the latter half of the 2000-2010 decade the State of Michigan experienced a large and growing net outmigration of citizens. This research uses Internal Revenue Service migration data to examine recent trends in county-to-county net internal migration from one county to another within Michigan as well as migration from Michigan counties to counties in other states. The IRS data indicate that all but six Michigan s 83 counties experienced net outmigration outside Michigan in 2007-2008, including 14 counties with net outmigration to other states in 2007-2008 at a rate greater than one percent of their populations. The slowdown in general economic activity from 2007-2010, likely was responsible for a portion of the movement of persons away from Michigan and has changed the economic landscape of Michigan s counties; however, further research is necessary to determine whether or not Michigan may have entered into a long-term pattern of chronic high outmigration. 1. Introduction The release of the 2010 Census of Population data shows that the State of Michigan lost almost 55,000 people or 0.6 percent of its population over the 2000-2010 decade. Furthermore, Michigan was the only state that experienced an overall population decline over this period. The population decline was not confined to a single region; however, the state s largest city, Detroit, experienced a drop of over 237,000 residents, or 25 percent of its 2000 population, over the decade. Among major U.S. cities, Detroit s loss of population was exceeded only by New Orleans during this 10 year period. Michigan actually gained population during the first half of the previous decade; however, its economy was impacted more severely than most other states during the recent economic downturn due to substantial losses in manufacturing, and any turnaround had not occurred in time to be reflected in the 2010 Census. A report on the economic impacts of changes in county population in Michigan released by the Land Policy Institute at Michigan State University (Adelaja et al., 2009) points out that the State s population declines began to surface between 2005 and 2006, and the ripple effects of the losses had spread beyond the manufacturing sector to the increasingly important service sector. A substantial amount of consumer spending for services disappeared when families left the state. The downward cycle continues to be complicated as cities and counties find it difficult to maintain service levels in light of the decline in local tax revenues as families continue to leave the state. The decline of the Michigan economy drove an outmigration, especially pronounced in the latter half of the decade, which was large enough to lead to a decrease in total population. According to U.S. Census Bureau estimates for the 2000-2009 period, Michigan had the nation s fourth largest net domestic outmigration, trailing only New York, California, and Illinois. This loss more than negated the nation s ninth largest natural increase and fifteenth largest immigration flow. Michigan State Demographer Kenneth Darga has acknowledged that the state has experienced

62 Sorenson and Gunderson substantial net outmigration of population since 2005; however, over a longer time period, Michigan has not seen chronic outmigration as has been popularized in the headlines. In the twenty years prior to 2005, levels of inmigration and outmigration from Michigan were roughly in balance over much of the period (Darga, 2008). Darga also puts to rest additional fallacies that Michigan has experienced a chronic problem of high outmigration of its young people as well as a chronic brain drain, and chronically high unemployment rates. Although these movements have become apparent in the most recent five or six years, Darga points out that a careful review of the data for the years leading up to the middle portion of the latest decade shows that Michigan s population problems are relatively recent and that popular accounts have done more to misinform our thinking about the state s population and its economy. Nevertheless, the problems facing Michigan have become very real over the past few years. In this paper we provide an in-depth descriptive analysis of the domestic migration patterns of Michigan s counties using Internal Revenue Service county migration data to examine migration flows occurring over three separate periods beginning in 1995-1996. After a brief literature review, our research focuses on trends emerging from patterns in recent domestic migration in and out of Michigan s counties, examining net migration as a percentage of county population for all Michigan counties along with a breakdown of the net migration percentage changes occurring across counties within Michigan as well as the net percentage leaving Michigan. Later in the paper, we present an analysis of population movement within the urban hierarchy as these persons relocate to other regions both within and outside of Michigan. Finally, we discuss case studies of three counties and a summary of the changing distances moved by outmigrants. 2. Literature review Migration research has long been a popular topic with researchers. This is, in part, due to an evolving set of behaviors that provides the impetus for people to move, as well as a number of other factors that may be specific or unique to a region. The migration literature contains numerous studies that address the factors which underlie the decision to migrate based on economic and noneconomic considerations in both origins and destinations. Factors examined in these studies have included demographic characteristics such as the ages of the migrants, personal attachments to a specific location as a result of friends and family who already live in the area, and specific amenities that generate a special place in the hearts of individual movers (Muth, 1971; Greenwood, 1975, 1985; Partridge and Rickman, 2006). Climate and quality of life factors were recognized as important indicators of population movement as early as the 1950s (Ullman, 1954). Migration patterns reflecting differences in attitudes to incorporate environmental and rural-oriented amenities, as opposed to improving economic opportunities, were investigated more recently by Vias (1999). Quality of life and other location-specific issues were recognized early on by Graves (1973, 1979, 1980) and more recently by numerous researchers including Cushing (1987), Cebula (2005), Cebula and Payne (2005), Deller et al. (2001), Green (2001), Gunderson and Ng (2005), Plane and Jurjevich (2009), and (Whisler et al., 2008). Specific sets of considerations may emerge when changes in economic conditions within a state or region serve to push people out versus the pull of favorable environmental factors in destination areas (Roback, 1982; Blanchflower and Oswald, 1994). More recently Cebula and Alexander (2006) examined the impacts in net interstate migration arising from both economic and non-economic factors and found that income, climate and higher per pupil government expenditures for public education each had a positive effect on migration, and cost of living, hazardous wastes and higher state income tax burdens each contributed negatively to interstate migration. Finally, in relation to movement along the urban hierarchy, papers by Plane and Jurjevich (2009) and Plane, Henrie, and Perry (2005) focus on agearticulated migration to examine the movement of persons up or down the urban hierarchy in the U.S. and the associated repercussions of these moves. Gunderson and Sorenson (2010) examines such movements for California outmigrants, noting the generally expected movement toward larger urban areas, less urban-destination emphasis for less urbanized origin counties, and pronounced variation among counties in similar urbanization classifications. 3. Net migration in Michigan counties in three recent periods The following maps present information on overall net migration changes as a percentage of

Internal Migration for Michigan Counties 63 county population along with separate breakouts to show county-to-county net migration rates to counties within Michigan as well as net migration rates from Michigan counties to counties located outside Michigan. In each instance, the net migration is expressed as a percentage of population living in each county-of-origin at the beginning of the period. The data are compiled from the Internal Revenue Service Statistics on Income County-to-County Migration. The migration data appear in Tables A1 and A2 in the Appendix. Migration flows are presented for three separate periods of time -- 1995-1996, 2002-2003, and 2007-2008. These time frames were selected based on changes in specific trends occurring in Michigan s counties during these years. The last year in which migration flows into and out of Michigan were roughly equal was 1995-1996. Outmigration began to exceed inmigration in many of the years immediately after 1996; however, the pattern of consistent and increasing outmigration was only established in 2002-2003, which is the second period analyzed here. Finally, the 2007-2008 time frame represents the last year before the impacts of the great economic recession became widespread. It is the intent in this paper to focus on migration trends that were occurring prior to the additional impacts that accompany any substantial downturn in economic activity. Therefore, 2007-2008 represents the last year relevant to the objectives of our analysis. In order to better understand the migration flows, it is useful to examine the urban geography of Michigan. Michigan contains 83 counties including 15 metropolitan statistical areas and 18 micropolitan statistical areas. These are shown in Figure A1 in the Appendix. A total of 48 of the 83 counties are either part of a metropolitan or micropolitan area while the remaining 35 are classified in more rural categories. Table 1 summarizes the data from the accompanying maps shown as Figures 1-3. Figure 1 shows the net migration percentages for 1995-1996. During this timeframe, only two Michigan counties ( and, both in the Upper Peninsula) experienced rates of net outmigration greater than two percent. Two more counties saw changes between -1 and -2 percent and 14 additional counties experienced smaller, but still negative, migration flows. However, for this time period, 65 of the 83 counties experienced positive rates of net migration. Table 1. Number of counties by change in overall net migration. Net Migration Percentage 1995-96 2002-03 2007-08 Less than-2.0% 2 0 1-2.0% to -1.0% 2 3 17-1.0% to 0% 14 22 48 0% to 1.0% 36 44 16 1.0% to 2.0% 15 11 0 2.0% or higher 14 3 1 Net Migration Percentage -5.12 - -2-2 - -1-1 - 0 0-1 1-2 2-7.76 Net Migration Percentage, 1995-1996 Figure 1. Net migration, 1995-1996. These overall net migration rates were already changing by 2002-2003, when 25 counties experienced negative net migration and 58 saw positive flows, but the majority of counties in both periods were still experiencing positive net migration flows. However, by 2007-2008, the rush of migrants to leave the state left only 17 counties with positive net migration flows and, of these, only, at the northern tip of the Upper Peninsula, had an increase in net migration exceeding one percent (+2.1%). By 2007-2008, sixty-five counties were losing population through net outmigration, which is almost a complete reverse of the situation that existed in 1995-1996. Most of the declines were between zero and one percent, although 17 counties lost between one and two percent to outmigration,

64 Sorenson and Gunderson and one county,, which includes Detroit, underwent a -2.2 percent decline. Figure 2. Net migration, 2003-2004. Net Migration Percentage -3 - -2-2 - -1-1 - 0 0-1 1-2 2-3 Net Migration Percentage, 2002-2003 Net Migration Percentage, 2007-2008 within Michigan). The instate migration data indicate that County was the only county to experience a rate of net outmigration to other Michigan counties at a rate which exceeded one percent of its population during all three periods. The most notable change, however, was that the number of counties losing between 0 and 1 percent of their populations to other Michigan counties almost doubled from 18 in 1995-1996 to 33 in 2007-2008. This suggests that an increasing number of counties in the state are losing population via migration to other counties within Michigan, effectively bringing about a shifting distribution of population within the state. Table 2. Number of counties by change in net migration within the state. Net Migration Percentage 1995-96* 2002-03 2007-08 less than -2.0% 0 0 0-2.0% to -1.0% 1 1 1-1.0% to 0% 18 14 33 0% to 1.0% 39 51 47 1.0% to 2.0% 15 15 2 2.0% or higher 8 2 0 *Total = 81 since and Counties did not report data for migration within Michigan Counties for 1995-1996. Net Migration Percentage Within Michigan, 1995-1996 Net Migration Percentage -3 - -2-2 - -1-1 - 0 0-1 1-2 2-3 Figure 3. Net migration, 2007-2008. A somewhat different pattern emerges when we observe population movement based upon whether the migrants moved from one county in Michigan to another instate or moved out of the state altogether. Table 2 summarizes the data from the maps shown in Figures 4-6 (migration flows among counties Net Migration Percentage -3 - -2-2 - -1-1 - 0 0-1 1-2 2-3.38 Figure 4. In-state migration, 1995-1996.

Internal Migration for Michigan Counties 65 Figure 5. In-state migration, 2002-2003. Net Migration Percentage -3 - -2-2 - -1-1 - 0 0-1 1-2 2-3 Net Migration Percentage -3 - -2-2 - -1-1 - 0 0-1 1-2 2-3 Net Migration Percentage Within Michigan, 2002-2003 Figure 6. In-state migration, 2007-2008. Net Migration Percentage Within Michigan, 2007-2008 An entirely different migration picture emerges when migration to counties out-of-state is measured. Table 3 summarizes the data from the maps shown in Figures 7-9 (migration flows to counties outside Michigan). County lost 4.57 percent of its residents to outstate migration in 1995-1996. This can, in part, be attributed to the closing of a U.S. Air Force base in the county during that period. In addition, lost 1.5 percent of its population during that year. Twenty-one other counties lost between zero and one percent to outmigration to other states in 1995-1996. However, by 2002-2003, sixty-four of the 83 counties were losing population to other states, and by 2007-2008, seventy-seven counties experienced net outflows to other states while only six counties were gaining residents from other states. The state s two most populous counties recorded some of the largest losses (, -1.14%, and -1.33%). County, adjacent to County, and home to Ann Arbor and the University of Michigan, had the highest rate of net outmigration to other states during 2007-08 (-2.2%). Furthermore, County, which includes East Lansing, home of Michigan State University had the third highest rate of outmigration to other states, losing 1.4 percent of its population during 2007-2008. Table 3. Number of counties by change in net migration out of state. Net Migration Percentage 1995-96* 2002-03 2007-08 less than -2.0% 1 0 1-2.0% to -1.0% 1 0 13-1.0% to 0% 29 64 63 0% to 1.0% 50 19 6 1.0% to 2.0% 0 0 0 2.0% or higher 0 0 0 *Total = 81 since and Counties did not report data for migration Outside Michigan Counties for 1995-1996. Overall, fourteen Michigan counties experienced net outmigration to other states during 2007-2008 at rates exceeding one percent of their populations. The number of counties losing between zero and

66 Sorenson and Gunderson percent of their populations to other states increased from 29 in the initial year to 63 in the final year a period of only twelve 12 years. What occurred in many cases was that counties which were losing small amounts of their populations to out-of-state migration in 1995-1996 were experiencing larger losses by 2007-2008. Over the same time frame, the number of Michigan counties gaining between zero and one percent from other states fell from fifity to six. A more detailed breakout of the positive migration flows from other states to Michigan counties in each time period indicates that the majority of the counties that did experience inflows from other states had only minor increases in populations of less than 0.5 percent. Only nine counties increased between 0.5 and one percent of their populations in 1995-1996, and this number dropped to only one county by 2002-2003. By 2007-2008, no counties increased by over 0.5 percent and only three increased between 0.25 and 0.50 percent (, 0.47;, 0.36; and, 0.31). Net Migration Percentage -3 - -2-2 - -1-1 - 0 0-1 1-2 2-3 Net Migration Percentage Outside Michigan, 2002-2003 Figure 8. Out-of-state migration, 2002-2003. Net Migration Percentage Outside Michigan, 2007-2008 Net Migration Percentage Outside Michigan, 1995-1996 Net Migration Percentage -4.57 - -2-2 - -1-1 - 0 0-1 1-2 2-3 Net Migration Percentage -3 - -2-2 - -1-1 - 0 0-1 1-2 2-3 Figure 9. Out-of-state migration, 2007-2008. Figure 7. Out-of-state migration, 1995-1996. 4. Percentage of county outmigrants leaving the state As an alternative perspective on the changes in Michigan county outmigration, we also examined the percentage of outmigrants leaving the state, focusing on changes from 2002-2003 to 2007-2008.

Internal Migration for Michigan Counties 67 The early 1995-1996 year was dropped due to a combination of the exceptional County case and data limitations for two other counties ( and ) which led to suppression even of total in-state and out-of-state migration. In addition, 2002-2003 represented a break to a faster growth of net outmigration, and 1995-1996 data was quite similar for a number of counties. Looking first at the baseline 2002-2003 outmigration, the border counties immediately jump out, with seven Upper Peninsula (U.P.) and three Indiana-bordering counties experiencing more than fifty percent of their outmigration out-of-state, as shown in Figure 10. Given the proximity of bordering urban centers, this is not surprising, although it should be noted in the case of the Indiana-bordering counties that significant Michigan metropolitan areas don t fare well in attracting outmigrants. Two additional U.P. counties have out-of-state percentages above fifty percent, and an additional four have percentages above forty percent. Two additional Indiana-bordering counties and County (home to Ann Arbor and the University of Michigan) also sent more than forty percent of outmigrants out-of-state. and counties, the two largest metropolitan Detroit counties, and County, the central county of Rapids, sent more than thirty percent of outmigrants out-of-state. The twelve counties sending the lowest percentage of outmigrants out-of-state, fourteen to twenty percent, are all in lower Michigan, and spread over a wide range of populations, including a few with slightly more than 100,000 residents. Many of the characteristics evident in 2002-2003 persisted into 2007-2008, especially the border county and Upper Peninsula concentrations. Two additional counties (from nine to eleven),, in the U.P., and, sent more than half of their outmigrants out-of-state. An additional fourteen counties (tripling from seven to twenty-one) sent more than forty percent out-of-state. Of the twelve counties sending less than twenty percent out-of-state, only and still sent that low a percentage out-of-state. Individual county changes in the percentage of outmigrants moving out-of-state are shown in Figure 11. Only seven of Michigan s 83 counties had a decrease in the out-of-state percentage between 2002-2003 and 2007-2008. Four of those counties were in the U.P., two were in the northern part of lower Michigan, and the remaining county borders Indiana. Only one of those counties had a change of over ten percentage points. In comparison, fifteen counties experienced increases of ten percentage points or more, including all of the suburban Detroit counties. County, conspicuous with its low percentage in 2002-2003, had the largest change in the state, an increase of 16.1 percentage points. The second largest change was in County, which increased from 44.3 to 59.5 percent of outmigrants moving out-of-state. Despite sending approximately 10,000 more outmigrants out-of-state in 2007-2008 than it did in 2002-2003, this constituted an increase of only nine percentage points. Percent of Outmigrants 14-20 20-30 30-40 40-50 50-80 Figure 10a. Percent leaving state, 2002-03. Percent of Outmigrants 14-20 20-30 30-40 40-50 50-80 Percentage of Outmigrants Leaving State 2002-2003 Figure 10b. Percent leaving state, 2007-08. Percentage of Outmigrants Leaving State 2007-2008

68 Sorenson and Gunderson Change in % of Outmigrants Leaving State 2002-2003 to 2007-2008 Table 4. Rural-Urban Continuum Code descriptions. 2003 Code Description Metro Counties: 1 metro area with 1 million pop. or more 2 metro area of 250,000 to 1 million pop. Change in Percentage -20 - -10-10 - -5-5 - 0 0-5 5-10 10-20 Figure 11. Change in percent leaving state. 5. Urban hierarchy effects Recently, considerable literature has focused on migration flows among levels of the urban hierarchy. While numerous ways of defining the hierarchy exist, we have chosen to use the Rural-Urban Continuum Codes (RUCC), also known as Beale Codes, developed by the Economic Research Service of the USDA. Table 4 provides a summary of the nine codes, which range from a top category of counties in metropolitan areas of one million or more people to a bottom category of completely rural or less than 2,500 urban population, not adjacent to a metropolitan area. The RUCC codes of Michigan s counties are mapped in Figure 12. The north/south split in urbanization is even starker in this figure than in the earlier metropolitan map. All of the counties in the three most rural categories are in the Upper Peninsula or the northern half of lower Michigan, and the distribution of metropolitan areas is such that even the least urbanized counties in the southern half of lower Michigan are adjacent to metropolitan counties. 3 metro area of fewer than 250,000 pop. Nonmetro Counties: 4 5 6 7 8 Urban population of 20,000 or more, adjacent to a metro area Urban population of 20,000 or more, not adjacent to a metro area Urban population of 2,500-19,999, adjacent to a metro area Urban population of 2,500-19,999, not adjacent to a metro area Completely rural or less than 2,500 urban population, adjacent to metro area Completely rural or less than 2,500 urban 9 population, not adjacent to metro area Source: Economic Research Service, U.S. Department of Agriculture. Rural-Urban Continuum Code 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-9 Rural-Urban Continuum Codes Figure 12. Rural-Urban code map.

Internal Migration for Michigan Counties 69 Migration to the differing county types from Michigan counties is shown in Table A3, organized by the RUCC of the sending county. In addition, the percentage migrating to each RUCC level is summarized for the group. As one would expect, outmigrants from the large metropolitan counties tend to migrate toward other metropolitan counties, with more than two-thirds of the 2002-2003 relocation to large metropolitan areas, much of it simply migration within the Detroit metropolitan area, and another 17 percent to smaller metropolitan areas. Only five percent moved to counties without an urban center of more than 20,000. About two-thirds of outmigrants from counties in mid-sized metropolitan areas had similar percentages, fairly evenly split between other counties in the same level and counties in the largest metropolitan areas. Only seven percent of outmigrants from these counties moved to counties without an urban center of more than 20,000, with the remaining one quarter of outmigrants moving to intermediate urban counties. Outmigrants from smaller metropolitan areas (RUCC 3) also concentrated in metropolitan destinations, with almost two-thirds moving to larger metropolitan areas. However, only nineteen percent moved to the largest metropolitan areas, while more than forty percent moved to mid-sized metropolitan areas. Another sixteen percent migrated to other smaller metropolitan areas. Even for this group, though, a mere ten percent moved to areas without an urban center of 20,000 or more. Outmigrants from urbanized counties adjacent to metro areas (RUCC 4) overwhelmingly focused on metropolitan destinations (about 85 percent), but for this group only about ten percent moved to the largest metropolitan areas, while about three-quarters moved to smaller metro areas. Even in this group only about ten percent moved to the least urbanized counties. More migration to nonmetropolitan locations is seen among outmigrants of less urbanized counties, but even in those counties significant percentages moved to metropolitan areas. In the similarly urbanized counties not located adjacent to a metropolitan area (RUCC 5), we see an expected drop in metropolitan destination, but the size of the drop, to less than forty percent of outmigrants, is surprising. In fact, a higher percentage of outmigrants move to the least urbanized counties. Outmigration patterns for the smaller counties adjacent to metropolitan areas (RUCC 6 and 8) are similar in that 60-70 percent migrate to metropolitan areas, but the RUCC 6 counties send larger percentages to the largest metropolitan areas, while RUCC 8 sends relatively more to smaller metropolitan areas. RUCC 8 counties send more than a quarter of outmigrants to the lowest three levels of the urban hierarchy, while RUCC 6 counties send less than ten percent to the smaller counties. In the two lowest nonmetropolitan categories not adjacent to metropolitan areas, less than one-third of outmigrants move to metropolitan areas, while more than forty percent move to counties in the three lowest RUCC categories. The location pattern percentages described for the groups are far from uniform, as shown in Table A4. Among the large metro area counties, for example, and outmigrants were much more evenly split between large and midsized metropolitan areas than were the other large metro counties, whose outmigrants were much more concentrated in other large metropolitan counties. Among the mid-sized metropolitan counties, 72 percent of and 50 percent of outmigrants moved to large metropolitan destinations, and a few other counties sent around a quarter of outmigrants to large metropolitan counties, but many other RUCC 2 counties sent less than ten percent of their outmigrants to larger metropolitan counties. Among the smaller metropolitan counties, several sent two-thirds or more of their outmigrants to larger metropolitan areas, while others sent less than half, with a low of only one quarter from County. While differences exist at other levels, they are perhaps most pronounced among the RUCC 6 counties. Around seventy percent of outmigrants went to the largest metropolitan counties, and a similar percentage of outmigrants went to mid-size metropolitan areas, but much lower numbers prevailed among the other counties. In three of those counties, less than a quarter of outmigrants went to the largest and mid-sized metropolitan areas combined. While the basic patterns of relocation within the urban hierarchy are interesting, changes over time are more instructive for examining the effect of disruption to migration from the economic problems of the state. Table 5 summarizes the urban status of receiving counties in 2002-2003 and 2007-2008. Inspection of these numbers shows remarkable consistency in the profile of percentages to different destination types. The largest deviation was the six percentage point increase in those moving from RUCC 2 to RUCC1 counties. Only five of the other eighty changes exceeded three percentage points. Change for individual counties was evaluated using an average absolute percentage point change

70 Sorenson and Gunderson measure across the nine destination county types, similar to a coefficient of specialization where the 2002-2003 profile provided the baseline distribution. The numbers are shown in Table A4 and are mapped in Figure 13. Only five counties, all in the Upper Peninsula or northern portion of lower Michigan, had changes over 25 percent. These typically involve smaller migration flows and are more sensitive to suppression. The counties in the 10-25 percent range were also typically northern counties with smaller flows. The majority of counties, 53 of 83, had changes of less than ten percentage points. Table 5. Urban Status of Receiving Counties, 2002-2003 and 2007-2008. Receiving County Rural-Urban Continuum Code 1 2 3 4 5 6 7 8 9 RUCC1 Counties: 2002-2003 Outmigrants 97,591 23,996 5,609 1,927 1,169 2,538 3,185 157 1,179 2002-2003 % to RUCC 71.1% 17.5% 4.1% 1.4% 0.9% 1.8% 2.3% 0.1% 0.9% 2007-2008 Outmigrants 112,617 25,182 5,937 1,839 1,036 2,064 2,228 109 746 2007-2008 % to RUCC 74.2% 16.6% 3.9% 1.2% 0.7% 1.4% 1.5% 0.1% 0.5% RUCC2 Counties: 2002-2003 Outmigrants 25,477 27,704 12,563 5,838 1,060 3,319 1,682 364 499 2002-2003 % to RUCC 32.5% 35.3% 16.0% 7.4% 1.4% 4.2% 2.1% 0.5% 0.6% 2007-2008 Outmigrants 32,275 28,450 12,036 4,936 1,075 3,050 1,281 378 323 2007-2008 % to RUCC 38.5% 33.9% 14.4% 5.9% 1.3% 3.6% 1.5% 0.5% 0.4% RUCC3 Counties: 2002-2003 Outmigrants 5,806 13,087 4,977 3,187 398 1,664 481 722 163 2002-2003 % to RUCC 19.0% 42.9% 16.3% 10.5% 1.3% 5.5% 1.6% 2.4% 0.5% 2007-2008 Outmigrants 7,111 13,569 5,083 3,002 328 1,553 477 601 55 2007-2008 % to RUCC 22.4% 42.7% 16.0% 9.4% 1.0% 4.9% 1.5% 1.9% 0.2% RUCC4 Counties: 2002-2003 Outmigrants 1,224 5,380 3,048 63 299 1,056 165 17 39 2002-2003 % to RUCC 10.8% 47.6% 27.0% 0.6% 2.6% 9.4% 1.5% 0.2% 0.3% 2007-2008 Outmigrants 1,432 5,234 3,354 36 260 834 182 - - 2007-2008 % to RUCC 12.6% 46.2% 29.6% 0.3% 2.3% 7.4% 1.6% 0.0% 0.0% RUCC5 Counties: 2002-2003 Outmigrants 1,013 1,202 326 325 676 535 1,387-1,221 2002-2003 % to RUCC 15.2% 18.0% 4.9% 4.9% 10.1% 8.0% 20.7% 0.0% 18.3% 2007-2008 Outmigrants 1,293 1,243 464 249 736 641 1,259-1,048 2007-2008 % to RUCC 18.6% 17.9% 6.7% 3.6% 10.6% 9.2% 18.2% 0.0% 15.1% RUCC6 Counties: 2002-2003 Outmigrants 1,632 3,088 1,606 736 529 1,553 900 36 42 2002-2003 % to RUCC 16.1% 30.5% 15.9% 7.3% 5.2% 15.3% 8.9% 0.4% 0.4% 2007-2008 Outmigrants 1,670 3,087 1,749 852 602 1,344 761 47 75 2007-2008 % to RUCC 16.4% 30.3% 17.2% 8.4% 5.9% 13.2% 7.5% 0.5% 0.7% RUCC7 Counties: 2002-2003 Outmigrants 1,386 1,191 434 120 1,587 890 2,729 289 1,227 2002-2003 % to RUCC 14.1% 12.1% 4.4% 1.2% 16.1% 9.0% 27.7% 2.9% 12.5% 2007-2008 Outmigrants 1,361 1,304 609 162 1,676 848 2,849 273 1,246 2007-2008 % to RUCC 13.2% 12.6% 5.9% 1.6% 16.2% 8.2% 27.6% 2.6% 12.1% RUCC8 Counties: 2002-2003 Outmigrants 50 265 612 27-40 301-76 2002-2003 % to RUCC 3.6% 19.3% 44.6% 2.0% 0.0% 2.9% 22.0% 0.0% 5.5% 2007-2008 Outmigrants 88 292 646 32-43 391-61 2007-2008 % to RUCC 5.7% 18.8% 41.6% 2.1% 0.0% 2.8% 25.2% 0.0% 3.9% RUCC9 Counties: 2002-2003 Outmigrants 287 252 111-1,182 37 1,148 42 105 2002-2003 % to RUCC 9.1% 8.0% 3.5% 0.0% 37.4% 1.2% 36.3% 1.3% 3.3% 2007-2008 Outmigrants 285 244 88 24 1,261 23 1,221 84 150 2007-2008 % to RUCC 8.4% 7.2% 2.6% 0.7% 37.3% 0.7% 36.1% 2.5% 4.4%

Internal Migration for Michigan Counties 71 Outmigration Orientation Change Based on Absolute % Changes to RUCCs for the flows. However, Table 6 illustrates the changes. Although two counties experienced decreases and another had too few inmigrants to avoid suppression in 2007-2008, these are far outweighed by the seven counties with 2007-2008 data but no 2002-2003 data, the increase in Cook County, Illinois, and increases in migration to every other region after the explicitly reported flows were registered. Index Value (0 to 100) 0-5 5-10 10-25 25-50 50-80 Outmigrants 2002-2003 Figure 13. Outmigration orientation change. 6. County case studies 6.1. To better appreciate changes in migration, we also conducted a case-study evaluation of three particular counties. We initially consider, a county of about 80,000 residents in the northern part of lower Michigan with an RUCC code of 5, i.e., a non-metropolitan county having an urban center of at least 20,000 which is not adjacent to a metropolitan county. Among non-metropolitan counties, it is of sufficient size to generate large enough flows to analyze. Net migration was positive in all three years discussed above, but it declined over time from 1.36% to 0.28%, largely due to a decrease in the out-of-state net migration rate from 0.23 to -0.65. Instate outmigration was fairly steady over the years, but out-of-state outmigration increased from 1128 to 1624 between 2002-2003 and 2007-2008. The percentage of outmigrants leaving the state increased from about thirty percent to forty percent. The largest flows of outmigrants from County moved to adjacent counties in both years (Figure 14), but the total flow was lower in 2007-2008. Major metropolitan counties in Michigan were other major recipients of outmigrants. Percentage changes in the number of outmigrants from are shown in Figure 15. Here the decline to neighboring counties is clear, as is the decline to,, and counties, but other counties, most notably and, had sizable increases. Given the smaller number of out-of-state destinations, no map is provided Number of Exemptions 15-50 50-100 100-200 200-300 300-400 Figure 14a. outmigration, 2002-03. Number of Exemptions 15-50 50-100 100-200 200-300 300-400 Fig. 14b. Gr. outmigration, 2007-08. Outmigrants 2007-2008

72 Sorenson and Gunderson % Change in Exemptions -100 - -50-50 - -25-25 - 0 0-25 25-50 50-130 % Change in Outmigrants 2002-2003 to 2007-2008 Fig. 15. Change in Gr. outmigration. Table 6. Leading out-of-state destinations for County outmigrants. Counties: State 2002-2003 2007-2008 % Change Maricopa AZ 35 27-22.9 Los Angeles CA. 20. San Diego CA 16 14-12.5 Broward FL. 22. Marion FL 25.. Pinellas FL. 21. Cook IL 46 69 50.0 Clark NV. 26. Buncombe NC. 20. King WA. 17. Regions (Other Flows): 2002-2003 2007-2008 % Change Northeast 80 101 26.3 Midwest 243 289 18.9 South 444 669 50.7 West 239 329 37.7 6.2 The second case-study county is, a metropolitan area of over 300,000 and part of the consolidated Detroit metropolitan area. As noted above, County includes Ann Arbor and the University of Michigan. Net migration has been close to zero in-state, but out-of-state net migration changed from -0.33% in 2002-2003 to -2.21% in 2007-2008. These percentages correspond to an out-ofstate outmigration increase of more than 4,000 residents, from 7,995 to 12,365. The percentage of outmigrants leaving the state increased from 44.3% in 2002-2003 to a staggering 59.5% in 2007-2008. County outmigrants within Michigan are heavily concentrated in the southeastern part of the state, as shown in Figure 16. Both time periods look similar in the mapping shading schemes, but the actual percentage changes indicate declines to all major receiving counties with the exception of, including declines of more than 25 percent to six Michigan counties as shown in Figure 17. The largest decline in numbers occurred in County, where 400 fewer outmigrants relocated. Number of Exemptions 9-50 50-100 100-500 500-1500 1500-4000 Outmigrants 2002-2003 Fig. 16a. outmigration, 2002-03.

Internal Migration for Michigan Counties 73 Outmigrants 2007-2008 exceeded fifty percent, including a large increase in outmigrants to the Chicago area and to several California counties, in a state itself known as a net loser of domestic migrants. In addition to these numbers, there is a sizable increase in flows to counties in Southern states where individual county flows were not reported. Finally, 41 out-of-state counties were reported in only 2007-2008, while only 13 were reported only in 2002-2003. Number of Exemptions Fig. 16b. outmigration, 2007-08. 9-50 50-100 100-500 500-1500 1500-3510 % Change in Exemptions -100 - -50-50 - -25-25 - 0 0-25 25-50 50-135 Fig. 17. Change in outmigration. Out-of-state migration flows were reported for a large number of destination counties. Those with more than 100 outmigrants in at least one of the two years are reported in Table 7. In every single one of these cases, outmigration increased between 2002-2003 and 2007-2008. The increase frequently % Change in Outmigrants 2002-2003 to 2007-2008 Table 7. Leading out-of-state destinations for County outmigrants. Counties: State: 2002-2003 2007-2008 % Change Maricopa AZ 132 234 77.3 Los Angeles CA 177 247 39.6 San Diego CA 107 247 130.8 San Francisco CA 64 108 68.8 Santa Clara CA 56 204 264.3 New London CT 27 656 2329.6 Cook IL 400 593 48.3 Dupage IL 48 119 147.9 Middlesex MA 83 199 139.8 Hennepin MN 75 142 89.3 St Louis MO 54 144 166.7 New York NY 165 255 54.5 Cuyahoga OH 93 103 10.8 Franklin OH 60 119 98.3 Lucas OH 128 149 16.4 Harris TX 63 100 58.7 Travis TX 55 100 81.8 King WA 118 183 55.1 Regions (Other Flows): 2002-2003 2007-2008 % Change Northeast 499 477-4.4 Midwest 763 820 7.5 South 1192 1661 39.3 West 532 538 1.1 The flow of County outmigrants can be better appreciated in Figures 18-20. Figure 18 shows outmigration flows for 2002-2003, illustrating the broad national spread of outmigrants throughout the United States. Figure 19 shows the same map for 2007-2008. The number of flows makes changes difficult to appreciate, but one can see a

74 Sorenson and Gunderson clear increase in flows to Western locations, an increase in density to the east, and a decline in flows northward in Michigan. Figure 20, which details percentage changes in migration levels for those counties whose figures were reported in both years, provides a striking picture of the changes that have occurred. Southern Florida, the northeastern megalopolis, and southern California/Arizona/Las Vegas stand out as areas with significant increases, while most of the shades representing decline are found within the state of Michigan. Outmigrants from 11 1956.5 3902 Figure 18. Outmigrants from County, 2002-2003. Outmigrants from 10 1758.5 3507 Figure 19. Outmigrants from County, 2007-2008.

Internal Migration for Michigan Counties 75 Outmigrants from -100 - -50-50 - -25-25 - 0 0-25 25-50 50-2400 Figure 20. Percentage change in outmigrants from County, 2002-03 to 2007-08. 6.3 The final county is, the state s most populous and home to Detroit. s net migration has been negative since 1995-1996, initially due to losses relative to the rest of the state, but those have now been more than matched by out-of-state losses, bringing the overall net migration rate to -2.23% in 2007-2008. In terms of actual outmigrants, County sent 10,000 more outmigrants out-of-state in 2007-2008 than it did in 2002-2003. The percentage of outmigrants leaving the state increased from 31.9% in 2002-2003 to 40.9% in 2007-2008. County outmigrants within Michigan are heavily concentrated in the southeastern part of the state, especially the neighboring suburban and counties, as shown in Figure 21. Both time periods look similar in the mapping shading schemes, but one can see declines in some of the secondary destinations. The percentage change map for Michigan (Figure 22) shows that most counties did experience declines, many of them over fifty percent, but that a significant number of counties, including and, received increasing numbers of outmigrants. Among the metropolitan area counties, saw a decline of almost 900 inmigrants from, while and saw increases of two thousand and almost three thousand, respectively. Out-of-state migration flows were reported for large number of destination counties. Those with more than 200 outmigrants in either one of the two years are reported in Table 8. In all but one case with a small decrease, outmigration increased between 2002-2003 and 2007-2008. The increase frequently exceeded fifty percent, and it more than doubled in nine cases, including Maricopa County (Phoenix), Clark County (Las Vegas), Harris County (Houston), and Fulton County (Atlanta). In addition to these numbers, it should be noted that 103 counties were reported in only 2007-2008, while only 23 were reported only in 2002-2003. The flows of County outmigrants are mapped in Figures 23-25. The outmigration flows for 2002-2003 show the large number of destinations from County, with an even broader national spread than. Figure 24, for 2007-2008, again shows the broad spread, but an initial glance shows both a greater density of lines emitted from County and a much large number of destinations in all regions. Figure 25, for percentage changes between the years, again provide an excellent summary of the changes that have occurred. Southern Florida, the northeastern megalopolis, and southern California/Arizona/Las Vegas once more stand out as areas with significant increases, although with some decreasing counties, but are

76 Sorenson and Gunderson joined by clusters around Seattle, Denver, Dallas/Ft. Worth, Houston, Atlanta, and Nashville. The Chicago metropolitan area also stands out in this map. As with, the distinct clustering of declining counties is found within the state of Michigan. % Change in Outmigrants 2002-2003 to 2007-2008 Outmigrants 2002-2003 % Change in Exemptions -100 - -50-50 - -25-25 - 0 0-25 25-50 50-120 Number of Exemptions 15-100 100-500 500-1000 1000-5000 5000-16000 Fig. 21a. outmigration, 2002-03. Number of Exemptions 15-100 100-500 500-1000 1000-5000 5000-18000 Fig. 21b. outmigration, 2007-08. Outmigrants 2007-2008 Fig. 22. Change in outmigration. Table 8. Leading out-of-state destinations for County outmigrants. Counties: State: 2002-03 2007-08 %Change Maricopa AZ 448 1156 158.0 Cook IL 564 836 48.2 Clark NV 345 768 122.6 Harris TX 209 456 118.2 Los Angeles CA 365 410 12.3 Franklin OH 211 348 64.9 Fulton GA 158 342 116.5 Lucas OH 341 332-2.6 Dupage IL 134 327 144.0 Tarrant TX 123 300 143.9 Cobb GA 172 297 72.7 Mecklenburg NC 111 294 164.9 Shelby TN 146 269 84.2 Dekalb GA 146 254 74.0 Dallas TX 129 254 96.9 Gwinnett GA 99 248 150.5 San Diego CA 221 229 3.6 Cuyahoga OH 166 226 36.1 Orange FL 151 215 42.4 Davidson TN 135 213 57.8 Wake NC 75 206 174.7 Other Flows: 2002-03 2007-08 %Change Northeast 586 716 22.2 Midwest 679 598-11.9 South 1273 1534 20.5 West 3015 3595 19.2

Internal Migration for Michigan Counties 77 Outmigrants from 10 7889 15768 Figure 23. Outmigrants from County, 2002-2003. Outmigrants from 14 8915 17816 Figure 24. Outmigrants from County, 2007-2008.

78 Sorenson and Gunderson Outmigrants from -100 - -50-50 - -25-25 - 0 0-25 25-50 50-300 Figure 25. Change in outmigrants from County, 2002-03 to 2007-08. 7. Average distance travelled The stark contrasts in migration patterns invited the question of just how far, on average, outmigrants were moving when they left their Michigan homes. In order to capture the moving distances, county-tocounty migration flows were merged with data on highway mile differences between all pairs of counties within the United States. For all Michigan counties and the state, the distances travelled are summarized in Table A5. We found that all three of the case-study counties had much higher average distances travelled in 2007-2008 compared to 2002-2003. County saw an increase from 367 to 540 miles, and saw an increase of almost one hundred miles, changing from 282 to 377. Even smaller County had an increase from 144 to 218 miles. The average distances travelled for all counties are mapped in Figure 26. The longest average distances, over 250 miles, were travelled from a number of U.P. counties and larger metropolitan counties. Much smaller distances, less than fifty miles, were travelled in quite a few counties, also including two U.P. cases. We must regard these numbers with some caution, however, as smaller counties are more likely to have reported county destinations for nearby migration, whereas more distant migration may not survive the suppression process. Given the suppression problems, it may be more informative to examine the changes between the two time periods, which are mapped in Figure 27. Here we see quite a wide variety of experiences. Although the statewide average went up by 98 miles, twenty-five, or almost a third of the counties, experienced decreases in average distance travelled. Decreasing-distance counties were concentrated in the U.P. and northern part of lower Michigan, but a number of cases were found further south. Most of the decreases were small, less than twenty-five miles, but a few were quite large. There were large increases, over fifty miles, in almost as many counties as experienced any decline. These were also found throughout the state, but a clear metropolitan Detroit cluster, the major college counties, and Rapids/ stand out. s increase of 173 was the largest, but nine other counties saw increases over 100 miles, including 161 miles for County. Much smaller increases, under 25 miles, were found in a comparable number of counties (24), while few counties fell in to the 25-50 mile change category.

Internal Migration for Michigan Counties 79 Average Distance Migrated 2002-2003 Change in Avg. Distance Migrated 2002-2003 to 2007-2008 Average Miles Migrated 20-50 50-75 75-100 100-150 150-250 250-380 Change in Avg. Miles Migrated -100 - -50-50 - -25-25 - 0 0-25 25-50 50-200 Figure 26a. Migration distance, 2002-2003. Average Miles Migrated 20-50 50-75 75-100 100-150 150-250 250-600 Figure 26b. Migration distance, 2007-2008. Average Distance Migrated 2007-2008 Figure 27. Change in migration distance. 8. Conclusions The State of Michigan experienced a large and growing net outmigration of citizens during the latter half of the 2000-2010 decade. Using Internal Revenue Service county migration data, we have examined recent trends in county-to-county net migration activity from one county to another within Michigan as well as from Michigan counties to counties in other states. From 1995-1996 to 2007-2008, the number of counties losing between zero and one percent of their populations to other Michigan counties increased from 18 to 33, which is an indicator that the population distribution is spatially shifting within the state of Michigan. However, the more important finding may be that the number of counties losing between zero and one percent of their populations to counties in other states increased from 29 in 1995-1996 to 63 in 2007-2008, while another 14 Michigan counties experienced net outmigration to other states in 2007-2008 at a rate greater one percent of their populations. Seventy-seven of Michigan s 83 counties experienced net outmigration to counties outside Michigan in 2007-2008. In addition, the average distance the outmigrants moved when they left Michigan also was increasing. Over the period from 2002-2003 to 2007-2008, the average distance travelled by the outmigrants from all Michigan counties increased by 98 miles. While

80 Sorenson and Gunderson almost a third of Michigan counties experienced a decrease in average distance travelled, most of the declines were small, less than twenty-five miles. Almost as many counties experienced large increases exceeding fifty miles in the average distance traveled by the outmigrants; primarily from the larger counties in metropolitan Detroit,, Rapids/ and in, where the increase in the average distance moved was the highest at173 miles. Although the slowdown in overall economic activity during the latter half of the previous decade was particularly detrimental to the State of Michigan, the movement of persons away from Michigan has changed the economic landscape of Michigan s counties and of the residents remaining in those counties. Further research will determine whether or not there are valid concerns that Michigan may have entered into a long-term pattern of chronic high outmigration, and in particular if this is concentrated among its youth and its better educated and high-income earners. References Adelaja, S., J.G. Hailu, and M.A. Gibson. 2009. The Economic Impacts of County Population Changes in Michigan - Full Report. New Economy Report Series. The Land Policy Institute at Michigan State University. Blanchflower, D., and A. Oswald. 1994. The Wage Curve. Cambridge, MA: MIT Press. Cebula, R.J. 2005. Internal Migration Determinants: Recent Evidence. International Advances in Economic Research 11:267-274. Cebula, R. J., and J.E. Payne. 2005. Net Migration, Economic Opportunity and the Quality of Life, 1999-2000. International Review of Economics and Business 52:245-254. Cebula, R. J., and G. M. Alexander. 2006. Determinants of net Interstate Migration, 2000-2004. Journal of Regional Analysis and Policy 36:2:116-123. Cushing, B. J. 1987. Location-Specific Amenities, Topography, and Population Migration. Annals of Regional Science 21(2):74-85. Darga, K. 2008. Fallacies That Misinform Our Thinking About Michigan s Population and Economy. Presentation at Michigan Revenue Estimating Conference. Deller, S.C., T.S. Tsai, D.W. Marcouiller, and D.B.K. English. 2001. The Roles of Amenities and Quality-of-Life in Rural Economic Growth. American Journal of Agricultural Economics 83:352-365. Graves, P. E. 1973. A Reexamination of Migration, Economic Opportunity and Quality of Life. Journal of Regional Science 13:205-211. Graves, P. E. 1979. Income and Migration Reconsidered. Journal of Human Resources 14:112-121. Graves, P.E. 1980. Migration and Climate. Journal of Regional Science 20:227-238. Green, G.P. 2001. Amenities and Community Development. Journal of Regional Analysis and Policy 31:2:61-76. Greenwood, M.J. 1975. A Simultaneous-Equations Model of Urban Growth and Migration. Journal of the American Statistical Association 70:797-810. Greenwood, M.J. 1985. Human Migration: Theory Models, and Empirical Studies. Journal of Regional Science 25:521-544. Gunderson, R.J., and P. Ng. 2005. Analyzing the Effects of Amenities, Quality of Life Attributes and Tourism on Regional Economic Performance Using Regression Quantiles. Journal of Regional Analysis and Policy 35(1):1-22. Gunderson, R.J., and D.J. Sorenson. 2010. An Examination of Domestic Migration from California Counties. Journal of Regional Analysis and Policy 40(1):34-52. Muth, R.F. 1971. Migration: Chicken or Egg? Southern Economic Journal 57:295-306. Partridge, M.D., and D.S. Rickman. 2006. An SVAR Model of Fluctuations in U.S. Migration Flows and State Labor Market Dynamics. Southern Economic Journal 72:958-980. Plane, D.A., C.J. Henrie, and M.J. Perry. 2005. Migration Up and Down the Urban Hierarchy and Across the Life Course. Proceedings of The National Academy of Sciences - PNAS 102(43):15,313-15,318. Plane, D.A., and J.R. Jurjevich. 2009. Ties That No Longer Bind? The Patterns and Repercussions of Age-Articulated Migration. The Professional Geographer 61(1):4-20. Roback, J. 1982. Wages, Rents and Quality of Life. Journal of Political Economy 90:1257-1277. Ullman, E.L. 1954. Amenities as a Factor in Regional Growth. Geographical Review 44(1):119-132. Vias, A.C. 1999. Jobs Follow People in the Rural Rocky Mountain West. Rural Development Perspectives 14(2): 14-23. Whisler, R.L., B.S. Waldorf, G.F. Mulligan, and D.A. Plane. 2008. Quality of Life and the Migration of the College Educated: A Life-Course Approach. Growth and Change (39):58 94.

Internal Migration for Michigan Counties 81 Appendix. Figure A1. Michigan Metropolitan and Micropolitan Areas, 2004.