With growth in demand for highly skilled workers expected to continue for the long term, many states are wondering what they can do to retain and attract recent college graduates. Some business leaders and policy makers in New England have expressed particular concern, citing the region’s high housing costs and noting that the southern and western sections of the country have gained population over the past twenty years at the expense of the North and East.

Using the National Longitudinal Survey of Youth (NLSY), I have been able to track the migration patterns of young people between 1979 and 1996, and make some calculated guesses as to what factors influenced these moves. My research suggests that New England may well face some disadvantages, in part because many of its residents destined for college were mobile as children or leave the area to attend college. Both factors increase the likelihood of graduates living outside the region after finishing their studies.

But New England also has some strong advantages, in particular, a vibrant economy with high pay for college grads. College graduates also appear to be attracted to states with a seacoast and the amenities that tend to locate in those areas. Perhaps surprising, high housing costs do not appear, at least in this sample, to be a major determinant of where recent college graduates move.

Have education will travel

While the highly skilled are highly prized by policy makers and business firms, they are also relatively hard to hold onto; more highly educated American workers typically exhibit substantially greater mobility. About 19 percent of NLSY respondents with just a high school diploma changed their state of residence at least once between 1979 and 1996; the figure was almost 37 percent for those completing four years of college, and 45 percent for those with even more schooling.

The greater mobility of the college-educated can also be expressed in terms of their frequency of movement. For young adult high school graduates, about 5 percent of sample members (on average) moved to a different state in any given year; for those completing college, the rate was about double.

Some of these moves were indeed the result of attending college out of state; about 27 percent of the sample graduated from college outside the state where they attended high school. But the highly educated continued to be mobile even after graduation. By 1996, 35 percent lived outside their state of high school attendance; the figure rose to 44 percent for those with education beyond college.

Most moves took place fairly shortly after college graduation. One year after college graduation, about 30 percent of grads were living in a state other than where they attended high school. This proportion rose to about 40 percent by year six, but then stayed roughly unchanged through year ten.

Regional differences

Although U.S. college grads had a high propensity to move, they didn’t move around the country evenly. Individual regions of the country experienced different net migration rates. Consider first the movement of college grads relative to their state of high school graduation. Recent college graduates were far more likely to move into the South Atlantic, Mountain, and Pacific divisions than other regions of the country. The South Atlantic and Pacific also saw below-average proportions of college graduates leaving those regions. As a result, net in-migration rates were highest for the South Atlantic, Mountain, and Pacific census divisions.

By contrast, New England gained about 20 percent more college graduates through in-migration (relative to the number who had finished high school here), but lost about 30 percent more to out-migration, for a net migration rate of -10 percent. Yet, this does not necessarily indicate a “brain drain,” since the NLSY sample doesn’t include migration from foreign countries, inflows more than five years after college graduation, or other potential sources of college graduates, such as an increase in the fraction of high school students from the region who obtain a college degree.

For the most part, the broad conclusions about migration patterns remain unchanged after comparing the location of recent college graduates to their college state, with the South Atlantic, Mountain, and Pacific states still big net gainers. But, some areas look relatively weaker or stronger by this measure. New England, for example, continues to show net domestic out-migration but at a more modest rate. This suggests that New England is better at retaining students who attended college in the region than in retaining its high school students who left the region to attend college elsewhere.

Is it the economy?

Not surprisingly, the economy played a significant role in regional migration patterns. Recent college graduates whose home states offered relatively high employment growth (compared to the national average) were less likely to move to other states. Six of the ten states with the highest employment growth during the period were in the South Atlantic and Pacific regions (Florida, Georgia, Virginia, Washington, Delaware, and North Carolina)—regions that also experienced among the lowest out-migration rates among college graduates. Recent college graduates were also less likely to move from their home state if it offered high average pay. Pay for college graduates tended to be highest in heavily urban states in the East—New Jersey, Maryland, Connecticut, the District of Columbia, New York, Massachusetts, and Virginia were all in the top ten (as were Alaska, California, and Michigan)—and lowest in heavily rural states.

One might also expect that regional differences in the cost of living would impact migration decisions, particularly high housing costs. But, the effect of high housing costs is difficult to assess. First, the states with the highest house prices changed over time. In 1980, seven of the top ten were in the west; by 1990, although Hawaii and California continued to rank one and two, the remaining top ten were all on the East Coast—with four in New England: Connecticut, Massachusetts, Rhode Island, and New Hampshire. New England and the Middle Atlantic states (as well as Hawaii and California) remained far above the national average in housing costs through 1996.

So young college graduates may have been discouraged from migrating into New England. Movers into New England, for example, faced housing prices about 40 percent above their high school state—second only to the differential faced by those moving to the Pacific region. Nonetheless, analysis using the NLSY sample fails to show that high house prices led to out-migration. This is because states with high house prices also tend to have other characteristics such as high pay, a low unemployment rate, and a coastal location, which are valued by recent college grads. Once these other variables are taken into account, house prices by themselves have an insignificant effect, although perhaps further research would show that housing costs become more important as individuals get older.

Past moving experience is the strongest predictor

Despite widespread attention paid to economic motivations for migration, personal experiences also turn out to play an important role. Someone who has moved in the past is more likely to do so again.

Attending college out-of-state was especially important. A person who moved to attend college was 54 percent more likely to live out-of-state five years after graduation than someone who went to college in-state. Someone who moved between birth and high school was 17 percent more likely to change states again than someone who had not done so. A past move was more important than either economic factors or amenities in determining who relocated after finishing school, explaining about 20 percent of the variation in individual migration patterns.

Yet, different parts of the country differ with respect to how geographically mobile young adults were during childhood. Those growing up in the East North Central and Middle Atlantic regions were the most likely to have been born in the state where they attended high school. By contrast, childhood mobility for college-bound New Englanders was very high, with about one-half graduating from high school in a different state than the one where they were born.

Regions also differ in the propensity for individuals to attend college out-of-state. The Pacific states had the lowest proportion leaving for college and the lowest proportion living out-of-state five years after college. College-bound students from New England were almost three times as likely to go out-of-state for college and to remain out-of-state after college.

This suggests that regional differences in the structure of higher education may influence migration patterns. The Pacific states—notably California—have an extensive, relatively low-cost, public higher education system, so most college-bound students choose to remain in the region. In California, more than three-quarters of all bachelors’ degrees in 1997 were granted by public institutions. In Massachusetts, by contrast, their share was less than one-third. This, along with high childhood mobility, contributed to the likelihood that young college graduates who went to high school in New England ended up locating elsewhere after completing their studies.

What can we learn?

What are the pertinent lessons for employers and policy makers?

The majority of young college graduates in a state are likely to be people who went to high school or college in that state. However, recent graduates also show a high degree of willingness to undertake long-distance moves.

The majority of moves are to states offering improved conditions along several dimensions—high job growth, lower unemployment, higher pay, lower housing costs, or better amenities. Young adults moving to the South Atlantic and Mountain states enjoyed greater-than-average increases in employment growth and greater decreases in unemployment than movers to most other regions. The clearest attraction of the Pacific region seems to be its seacoast location. College grads have moved there despite a mixed economic picture with housing costs, in particular, far higher than most other parts of the nation.

Although state economies are influential in determining where new college graduates move, it is important to keep in mind that over 40 percent of the observed moves were to states with lower employment growth. This suggests that location preferences vary from individual to individual, and circumstances unobservable to researchers (such as specific job offers or personal relationships) also help determine location decisions. An employer in a slow-growing or otherwise “unattractive” state that is trying to fill a relatively small number of slots may not be at a marked disadvantage. Recruiting success depends largely on matching particular individuals to particular jobs, rather than on being located in a particular part of the nation.

*Sample members were between the ages of 14 and 22 in 1979 Source: Author’s calculations using the National Longitudinal Survey of Youth

Note: Based on place of residence five years after college graduation. Does not include migration from outside the United States. Net migration may not exactly equal “moved into” minus “left” because of rounding. Source: Author’s calculations using the National Longitudinal Survey of Youth, 1979 to 1991