Investigating Millennial Brain Drain with Six Region Types

Regional “brain drain” in the United States is a common point of concern for economic developers, educators, and government officials. Are highly skilled or well-educated individuals leaving one region in pursuit of opportunities in other areas of the country? Do some areas experience what is called “brain gain”, where the area attracts more than its fair share of these high-skilled, well-educated individuals? Rather than focus on a binary determination of “brain gainers” and “brain drainers”, in this article Impact DataSource defines six region types based on regional education characteristics and observed population changes among the millennial generation, people born between the early 1980s and early 2000s. The regions defined in this article and this interactive map include three basic categories, which are combined with an education qualifier to make a total of six region types.

  1. Gainer  – Region that sees a uniform positive increase in the entire college and post-college population (ages 15 to 29).
  2. Drainer  – Region that sees a uniform decrease in the entire college and post-college population (ages 15 to 29).
  3. Exporter  – Region that sees an increase in college-aged population (ages 15 to 24), but a decrease in the post-college population (ages 25 to 29).
  4. Educated Gainer  – A more specific category of a Gainer region where the percentage of population having obtained a college degree exceeds the percentage observed at the national level.
  5. Educated Drainer  – A more specific category of a Drainer region where the percentage of population having obtained a college degree exceeds the percentage observed at the national level.
  6. Graduate Exporter  – A more specific category of an Exporter region where the percentage of population enrolled in a college degree program exceeds the percentage observed at the national level.

The above region types were based on an analysis of American Community Survey (“ACS”) data regarding population, educational attainment, and school enrollment status. This article investigates regional population changes for individuals from 15 to 29 years old, which traditionally includes a period of human capital investment and the first few years of working life. Understanding regional population shifts among this group of individuals and the phenomena of “brain drain” and “brain gain” can help local, state, and federal government officials develop policies that can build economically strong and stable regions.

Educational Attainment
According to the 2015 ACS (5 Year Estimates), 29.8% of the U.S. population over age 25 has a college degree (a bachelor’s degree or higher). Across each of the 50 states and District of Columbia, the percent of state population with a college degree varies from 19.2% in West Virginia to a whopping 54.6% in the District of Columbia.

College Enrollment Status
The percentage of population having obtained a college degree speaks to an attribute of the adult population of that state. To investigate this idea of brain gain we also need to take a step backward in the education process and investigate the proportion of the overall population identified as currently enrolled in a college degree program.

If there was no geographic mobility in the United States and educational attainment and enrollment did not fluctuate over time, the states with the highest proportion of college-educated individuals would necessarily be those states with the highest proportion of students enrolled in a college degree program. Of course, the U.S. population is geographically mobile, and educational enrollment and attainment are not static figures – so it is expected that these two lists will exhibit some variation. However, despite geographic mobility and educational trends over time, we can see that many of the most highly educated states also rank highly in terms of the proportion of the population enrolled in college degree program. In short, states that rank highly in college enrollment percentage tend to rank highly in educational attainment percentage.
The District of Columbia is at the top of both lists, with the highest percentage of both population with a college degree and f population enrolled in college. However, Colorado ranks third in the college degree attainment percentage, but it is 16th in the college degree enrollment percentage. Crudely analyzed, this result implies that Colorado attracts individuals that are already college educated to the state.
In order to observe which areas are losing or gaining population during the traditional college enrollment years, we looked at population data by state and county from the ACS.

Population Change
For a given state or county, we can observe the population aged 10-14 in 2010 and compare that to the population aged 15-19 in 2015. If there were no deaths and zero net migration, we would expect the population of individuals aged 10-14 in 2010 to be identical to the population aged 15-19 in 2015. Assuming mortality for this age group is negligible, any observed change in this population would be attributable entirely to individuals moving in to, or out of, the region. Not surprisingly, the data shows that most migration occurs in the early years, roughly between ages 15 and 30, which is where we focus our analysis.

Although not all young adults attend college, it is helpful to rename the age group categories to aid comprehension and simplify terminology. Subsequently, the Age 15-19 Group is called “Early College Age”, the Age 20-24 Group is referred to as “Late College Age”, and the Age 25-29 Group is “Post College Age”.

Age-Specific Migration in Iowa, Colorado, & Michigan
Let’s look first at the State of Iowa. The table below shows the populations of each of the three main age groups in 2010 and 2015. In 2010, there were an estimated 199,074 individuals between ages 10 and 14 (shown in orange). Five years later, these individuals would be in the Early College Age, between 15 and 19 years old. In 2015, there were in fact 213,453 Early College individuals, showing a 7.2% increase over the baseline from five years prior.
The fact that there was an increase implies there was migration into Iowa among individuals in their late teenage years during this time period. Positive migration to Iowa is also seen for Late College Age in 2015 (yellow). But for the Post College Age in 2015 (blue), we see a negative change from five years prior. This negative change implies there was migration out of Iowa among those individuals aging into their late 20s.

Using ACS data for the five-year period beginning in 2009, the same analysis was conducted to track the percentage change in age groups between 2009 and 2014 as well. The resulting percentage changes were averaged with those from the 2010-2015 period, shown in the table above, to determine the general trend and magnitude of these age-related migration patterns. The change in population for Iowa is shown below, covering all age groups.

In the case of Iowa, small increases are observed for the age groups below 14 years old. A strong spike is observed for the Early College Age population (15-19) followed by an additional small increase for the Late College Age (20-24) category – following the trends shown in the table. A significant out-migration is observed in the Post College Age group (25-29), also as indicated in the table. After the Post College Age, there follows modest increases for individuals between ages 30 and 39, before showing an increasing negative trend afterward.
As mentioned, it is assumed that mortality does not play a significant role in the analysis of school- and college-aged students, but this assumption is much less likely to hold true when the analysis is aimed at older age groups. You can be fairly confident that the large negative change for 80- to 84-year-olds is not due to some sudden urge to move to another state. It would be dangerous to draw too much conclusion about migration among older age groups due to the increase in mortality rate therein.

Overall, the trends for Iowa seem to indicate that young adults have been moving into the state during their Early and Late College years, then moving out at a significant rate during their Post College years. The migration patterns of other states can be used to verify these findings.

Colorado and Michigan also exhibit distinctly interesting trends, which are notably different from those of Iowa and from each other. These trends are illustrated in the graph.

Compared to the whipsaw of inflows and outflows of young adults seen in Iowa (shown in yellow), Colorado sees a crescendo of growth among the same population (green), while Michigan sees a uniform exodus from the state (pink).

While all three states exhibit distinctly different trends among the ages of interest for this analysis, some general trends are clearly consistent. As noted, areas appear to experience the greatest gains or losses of population in the young adult age groups, roughly between ages 15 and 30 – the groups of interest in our study. Similarly, all three states show a consistent negative trend in population for individuals over 40 years of age, as the issue of mortality comes more strongly into play.

These three states are clear examples of three of the six region types defined for this analysis.

Exporter
Iowa appears to lose population just at the age it is common to graduate college, a possible indication that the state “exports” college-educated individuals. One can feel more confident about that observation after adding an analysis of college enrollment in the state as well.

First, in an isolated graph of the three age groups of particular interest, it can be seen that Iowa gains population over the Early and Late College Age groups and loses population in the Post College Age group. Therefore, Iowa is clearly an exporter of young adults.

Additionally, the proportion of the population enrolled in a college degree program in Iowa exceeds that of the United States overall (7.5% in Iowa vs. 7.4% nationally).

With these two sets of facts, we categorize Iowa as a Graduate Exporter.

Gainer
For comparison, let’s consider Colorado, which, based on enrollment and attainment percentages, appears to attract college-educated individuals. Looking at population changes by age categories in the graph below, Colorado exhibits growth in all age categories below 44 years old, with particularly strong growth in the Post College Age category. Overall, Colorado clearly gains population throughout the student and young adult age categories.

In addition to this finding, we have seen that the percentage of the Colorado population having attained a college degree exceeds that of the United States.

Given this set of findings, Colorado can be classified as an Educated Gainer.

Drainer
For the final comparison, let’s consider Michigan, which will round out the range of community types we have defined in our study of “brain gain”. In the graph below, it is seen that Michigan has exhibited a loss in population across all age categories.

Additionally, the percentage of the Michigan population having obtained a college degree is below the percentage nationwide and therefore Michigan is classified simply as a Drainer.

Six Region Types
Impact DataSource defined six region types based on observed population changes for individuals aged 15 to 30 and based on regional education characteristics. The regions include three main types, subject to education qualifiers for six regions total.

  1. Gainer  – Region that sees a uniform positive increase in the entire college and post-college population (ages 15 to 29).
  2. Drainer  – Region that sees a uniform decrease in the entire college and post-college population (ages 15 to 29).
  3. Exporter  – Region that sees an increase in college-aged population (ages 15 to 24), but a decrease in the post-college population (ages 25 to 29).
  4. Educated Gainer  – A more specific category of a Gainer region where the percentage of population having obtained a college degree exceeds the percentage observed at the national level.
  5. Educated Drainer  – A more specific category of a Drainer region where the percentage of population having obtained a college degree exceeds the percentage observed at the national level.
  6. Graduate Exporter  – A more specific category of an Exporter region where the percentage of population enrolled in a college degree program exceeds the percentage observed at the national level.

Nationwide Millennial Population Shifts
These regional definitions were applied to all 50 states and 3,000+ counties in the United States. Approximately 70% of the states and 45% of the counties could be classified into one of the six region types. Regions not classified into one of the six types simply do not exhibit a discernible trend in population and education for the 15- to 29-year-old age categories.

It is tempting but difficult to rank these region types in an order of “best” to “worst”. Clearly, Educated Gainer regions, those that have a highly-educated population that is growing among young adults, are in an enviable position. Conversely, Drainer regions and Educated Drainer regions are facing a difficult future. However, it is hard to determine if a region with below-average educational attainment that is losing its young adult population is necessarily in a worse position than a region that is also losing young adult population but has a relatively high educational attainment level.

One may be inclined to identify Graduate Exporter regions as the epitome of “brain drain”, however that view is incomplete. The Graduate Exporter regions play a significant role in producing the highly-educated individuals that contribute to creating economic growth and prosperity across the nation. These regions see a significant influx of young students and produce a significant outflow of graduates. At the county-level, a Graduate Exporter can generate great gains for a wider region, and therefore the Graduate Exporter county is not necessarily something to disparage. However, at the higher level, a Graduate Exporter state may be cause for concern.

Investigate your region with this map.