Category Archives: Demographics

Map of the Day: Changes in Probability of Death

Map source: Wall Street Journal

It’s not new news anymore that gains in life expectancy have leveled off in the United States, driven by startling and unexpected declines among young and middle-aged whites. The so-called “deaths of despair,” including drug overdoses, are on the rise. So are liver disease (associated with alcoholism) and suicides. Chronic diseases associated with obesity such as diabetes, heart disease, and stroke are up, too.

The map above shows changes in the probability of death among 20- to 50-year-olds in the 50 states between 1990 and 2016. There is a remarkable divergence — health for this age group has improved significantly for some states, including Virginia, and gotten worse for others.

A breakdown by county in Virginia would be revealing. I hypothesize that western Virginia, especially the far Southwest, would show patterns similar to neighboring West Virginia and Kentucky. Although a more granular look at the data might reveal a different pattern, it appears that Central Appalachia is ground zero for deaths of despair.

Chart of the Day: Household Income Distribution by Region

Our friends over at the Demographics Research Group at the University of Virginia continue to display data in interesting ways. The chart above breaks the state into eight demographic regions and then plots the median household income by locality. As can be seen, there is significant variability within regions — particularly in Northern Virginia and Central Virginia, both of which in this schema include some outlying rural counties. But even the poorest Northern Virginia locality has a higher median income than the richest localities in Southside and Southwest Virginia.

As with all income comparisons, this does not adjust for the considerable differences in the cost of living, particularly housing, so it does not accurately reflect disparities in living standards. If the Demographics Group could adjust for cost of living, we’d really have something interesting to examine.

Why Are Asians and Hispanics So Healthy?

City/county ranking of Virginia health outcomes based on potential years of life lost before age 75. Source: Robert Wood Johnson Foundation

The Robert Wood Johnson Foundation has issued its annual Healthy Community report for the United States. As usual, the information is packaged in such a way as to highlight the health disparities between racial/ethnic groups. But the findings for Virginia, which the state-level report largely overlooks, do not fit the dominant institutional-racism narrative. It turns out that Asians are the healthiest racial/ethnic group by far. It also turns out that, despite lower incomes and education levels, Hispanics are healthier than whites. The only finding that conforms to the narrative is the blacks are the least healthy of any group.

The info-graphic to the right shows differences in health outcomes (potential years of life lost before age 75) by place and by race/ethnicity. The “place” metric compares the differences in health outcomes by city or county. There is a wide disparity (as also seen in the map above) between localities with high incomes and high levels of education and localities with low incomes and education. The worst pockets of unhealth are in far Southwest Virginia, Southside, the Eastern Shore, and older cities. No surprises there.

Far more interesting is the disparity between racial/ethnic groups, which many researchers and commentators persist in defining as a gap between whites on the one hand and blacks and Hispanics on the other — a gap matching the socio-economic divide and consistent with the paradigm of America as a nation afflicted with institutional racism and discrimination.

Yet of all major racial/ethnic groups, Asians are the healthiest. By far. Here in Virginia, according to the study, Asians experienced the lowest level of “premature deaths,” measured by years lost per 100,000 — only 2,600. Hispanics fared next best, with 3,100 years lost, whites with 6,200, and blacks with 8,700.

Another remarkable finding: Whites reported the highest incidence of poor mental health days: 1.6 for Asians, 2.7 for Hispanics, 3.5 for blacks, and 3.8 for whites.

Results conformed to stereotype for poor or fair health, while self-reported “poor health days” showed almost no difference between whites, blacks, and Hispanics. Asians reported the fewest poor health days.

The comparative good health of Hispanics in Virginia is all the more remarkable given that, as the report documents but takes little note of, Hispanics have lower high school graduation rates, have less health insurance, and have a higher rate of teen births than any other group.

Asians and Hispanics do not fit the dominant narrative of the relationship between race and health in the United States. It strikes me that these anomalies are worth exploring. Persuading public health researchers to dig deeper may be a hard thing to do, however. The received wisdom, once established, is a hard thing to dislodge.

Update: And then there’s this headline from the Roanoke Times: “Report finds death rates rise for white, middle-class Virginians.”

Chesterfield County Leaking Affluent Households

Which better represents the future of Chesterfield County? Rudd’s Trailer Park……. (Photo credit: Richmond Magazine

This column was published originally in the Chesterfield Observer. While the details of migration trends in and out of Chesterfield is unlikely to prove of great interest to anyone outside of Chesterfield, the analysis shows how citizens can use IRS migration data to gauge the health of their home locality. 

Hobbled by the sequestration-driven budget squeeze of defense spending, Virginia experienced its fourth consecutive year in 2016 of out-migration, the University of Virginia’s Demographic Research group reported late last year. While 301,000 income tax-filing households moved into the state, 315,800 moved out, for a net loss of 14,800 households. The last four years are quite a comedown for a state that previously had seen healthy population inflows every year since the Internal Revenue Service began compiling the statistics in 1978.

The picture looks somewhat better for Chesterfield County, which saw a net gain of 687 households from people moving in and out of the county in 2016. Some 10,300 households entered the county while 9,600 left and another 123,400 stayed put.

….or this McMansion?

People move from one locale to another for a multitude of reasons, but it’s normally a good sign when more people move in than out. Insofar as people follow jobs when they move, a net gain in migrants could mean that more jobs are being created. An inflow of residents also pumps up demand for housing, retail and services, thus stimulating local economic activity.

On the flip side, an influx of households places greater burden on the county to provide education, public safety, streets and roads, and other basic government services. In an ideal world, the newcomers bring in more taxable income and spending power to help pay for those services than those who leave. Unfortunately, that’s not what’s happened in Chesterfield. Households that moved here in 2016 reported an average income of $56,200; those that left reported $58,500 – for a total net loss of $17 million in countywide income. Admittedly, that’s a drop in the bucket compared to the $10.1 billion in total income reported by all non-migrants. But if this becomes a trend and persists for years and decades, it could fundamentally change the nature of the county.

A large percentage of the coming and going consists of people moving to and from neighboring jurisdictions in the Richmond metropolitan area. In 2016, Chesterfield experienced a net gain of 401 residents from Henrico County, 229 from the City of Richmond, and 56 from the city of Petersburg. However, the county lost a net of 126 households to Powhatan County.

The good news for Chesterfield is that it is importing more affluent households from Richmond, Henrico and Petersburg than it is exporting.

Newcomers from Richmond earned on average $46,100, while those moving from Chesterfield to the city reported only $39,500 in income.

Similarly, Henrico immigrants to Chesterfield earned $60,200 on average while households going the other way earned only $49,900.

The differential for Petersburg was $37,200 on average for households heading from the city to the county compared to a lowly $29,800 for households heading in the reverse direction.

However, Chesterfield lost significant income to Powhatan County in 2016. While the number of migrants is relatively small, the income differential is vast. Households moving from Powhatan to Chesterfield made $48,900 on average while those leaving Chesterfield earned $87,200, a differential of $37,300.

The largest sources of in-migrants from outside the region are Fairfax County, Virginia Beach, and Wake County, North Carolina (in the Raleigh metropolitan area).

The wrong conclusion to draw from this data is that Chesterfield taxpayers might benefit from crafting policies and ordinances that make the county less attractive to the poor, say, by blocking real estate projects developed for lower-income households. Aside from the ethical issues raised by discriminating against the poor, that’s not even good policy. Poor people will gravitate toward the cheapest, least desirable housing stock available in the metro area, whether it’s public housing projects in Richmond or aging cul-de-sac neighborhoods of small, rundown 1950s and ’60s era ranch houses in Chesterfield, regardless of any policies the county pursues.

A better strategy is to make carefully considered investments that help build a more prosperous, livable and sustainable community for all. Tracking the IRS migration data is a good way to tell how well county leaders are doing to create a desirable place for everyone to live, work and play.

Another Look at Virginia’s Lagging Population Growth

Image credit: StatChat blog

After decades as one of the nation’s fast-growth states, Virginia’s population now is growing line with national averages, according to data found in Hamilton Lombard’s latest post on the University of Virginia Demographics Research Group’s StatChat blog.

Lombard attributes the lagging population growth to out-migration resulting from the impact of federal budget austerity on Northern Virginia’s defense-oriented technology sector. Despite this slowdown, NoVa is adding to the state’s population faster than any other region in Virginia because its population skews younger — more women are in the child-bearing age. Indeed, since 2010, NoVa has added more people than the rest of the state combined.

The Richmond region has added the second largest number of people since 2000. And despite its lagging economy, Hampton Roads has contributed substantially to population growth. However, due to emigration of young people, an aging population, and fewer births, population growth in Virginia outside the urban crescent has collapsed since 2010.

Writes Lombard: “Without a surge in population growth, by 2020 Virginia could have close to 200,000 fewer residents than would have been expected based on past population growth trends. Meanwhile, Virginia’s aging population will likely cause the number of counties with more deaths than births to continue to increase, slowing population growth throughout the Commonwealth, even in Northern Virginia.”

Retirement, Not Jobs, Pushing Virginians Out of State

Virginia has been losing population to domestic out-migration for the past five years. Most people (including me) have assumed that the reason for the exodus (well, not really an exodus, more of a drip… drip… drip… leakage) can be attributed to sub-par economic growth. In other words, more people are leaving than coming because more jobs are being created elsewhere than here.

But the latest data from United Van Lines calls that assumption into question. United’s data roughly tracks that of the Internal Revenue Service taxpayer change-of-address data in noting that for every 100 moves in and out of Virginia 53% were outbound compared to only 47% being inbound.

But get this: Two-thirds of the reasons cited for moving into Virginia were jobs, while only a little more than half were so cited for moving out. The widest outbound-over-inbound gap was for retirement, the second widest for family. Virginia also suffered smaller gaps for health and lifestyle.

Why would there be such a large retirement gap? Our 5.75% top income tax bracket? Hellish traffic in Northern Virginia? Too many polar vortexes? Perhaps readers can chime in with their speculations.

Fewer Young People, More Geezers Working These Days

Source: StatChat blog

The percentage of young people (ages 16 to 25) participating in Virginia’s workforce has been sliding over the past 15 years, while the percentage of old-timers (65 years or older) in the workforce has been increasing.

The percentage change of geezers in the workforce increased 59% between 2001 and 2016, according to data published by the Demographics Research Group of the University of Virginia on the StatChat blog. By contrast, the percentage of the working 20- to 24-year-olds slid 11%, while the percentage of teenagers (ages 16 to 19) tumbled 24%.

The percentage of young people (ages 16 to 24) enrolled in school has been rising steadily since 1960. The big change in Virginia has been a marked decline in the percentage of students who are working while in school, as seen in the graph below. StatChat sees this as a good thing:

A greater proportion of teens are now choosing not to work in order to dedicate more of their time to school studies and to non-work-related extracurricular activities, perhaps to gain a competitive advantage in admission to postsecondary educational programs.

That’s one explanation. Another is that young people today don’t have the same work ethic as their elders. Members of the most affluent generation in American history would prefer to spend more time finding themselves, “giving back,” or following their bliss than submitting to the grind of a regular job. That’s not true of all young people, some of whom work very hard, but I have a hunch that it’s fair as a generalization.

The Demographics group did not proffer an explanation of why more seniors are working, but I’ve got a hunch. Baby Boomers may have been a hard-working generation, but they weren’t very frugal. The cohort now reaching retirement age did not do a very good job of saving for retirement. Now Boomers are finding they have to work longer to make ends meet. Just a hunch.

Virginia’s Income Drain: $1.5 Billion Last Year

Last week I noted that more people left Virginian between 2015 and 2016 than moved into the state — the fourth year in a row the Old Dominion suffered more out-migration than in-migration. From a taxpayer’s perspective, that wouldn’t be so bad if poor people were leaving and rich people were coming in. Sadly, that doesn’t seem to be the case.

The IRS data is based on the change in addresses of people who file income tax returns. The “returns” column in the table above shows the net gain or loss in the number of returns filed; the “exemptions” includes taxpayers plus other family members claimed as exemptions. Tax filers leaving Virginia reported an average income per filer of $73,900, while those who entered the state reported $66,100.

Not only did Virginia lose a net 9,000 taxpayers in 2016, we lost nearly $1.6 billion in income. Assuming an effective income tax rate of about 5%, that represents a loss of about $76 million in tax dollars.

One year’s loss of $1.6 billion in income is hardly a disaster when those who stayed behind reported $250 billion in income. But the steady erosion of the population and tax base over four years does add up, and it will continue to add up if Virginia can’t turn things around. Our collective tax and debt obligations looms a little bit larger when there are 9,000 fewer tax payers each year to shoulder the burden.

Not surprisingly, Virginia lost the most taxpayers (3,900) and the most income ($645 million) to income tax-free Florida. We’re hardly alone in that regard. Wealthy retirees of many other states do the same thing. But how does Virginia explain the net loss of more than 3,000 residents, along with $320 million in income, to high-tax Maryland?

Virginia was a net exporter of income to 40 other states and an importer of income from only 10 states (plus Washington, D.C.). We can console ourselves that our reversal of fortune from 35 years as an importer of people and wealth is temporary, driven by cutbacks in federal funding for the state’s military-industrial complex. But what if there’s more to the story? It would be helpful to take a closer look at which cities, counties and metropolitan areas are winners and losers… which I will do if I can find the time.

More People Still Leaving Virginia than Moving In

Source: StatChat blog

Virginia has experienced its fourth consecutive year of domestic out-migration, reports Hamilton Lombard with the University of Virginia’s demographic research group in its StatChat blog. Prior to 2013, Virginia had never experienced a year of out-migration since the Internal Revenue Service began collecting data in 1978. (The data is based on address changes for households filing income taxes.) Virginia’s population is still growing thanks to a surplus of births over deaths, but the growth rate has slowed from 80,000 a year in the 2000s to 50,000 for the past four years.

Lombard attributes the out-migration largely to the impact of sequestration-related cutbacks to defense spending on the economy in Northern Virginia and Hampton Roads. He also suggests that the high cost of Northern Virginia housing may play a factor in the exodus.

While Virginia still experiences net in-migration from traditional feeder states in the Northeast, it is exporting population to fast-growth Sunbelt cities. The largest cohort of immigrants is in the 26- to 35-year-old age range, although Lombard expects to see a growing number of retirees leaving the state for lower-tax climes.

The New Look of Virginia High School Grads, Circa 2030

Lots of good data coming out of the retreats of the House and Senate appropriations committees yesterday and today… The chart above appeared in a presentation by April Kees, legislative fiscal analyst, to the Senate Finance Committee.

By 2030, whites will constitute a bare majority of high school graduates in Virginia. The percentage of blacks will shrink slightly, while percentages of Asians and Hispanics will soar.

Bacon’s bottom line: This is what college administrators are talking about when they allude to a challenging demographic future. The percentage of Hispanic students graduating from high school and populating the potentially college-bound population will grow by six percentage points, offsetting the seven-point decline in the percentage of whites. Insofar as whites tend to come from more affluent families, to attend better schools and to be better academically prepared than Hispanics, colleges are bracing for student bodies that need more remedial work and financial assistance.

On the other hand, Asian students tend to come from more affluent households and to be better prepared academically than all other ethnic groups, including whites. They could prove to be a mother lode for institutions looking for students with high SAT scores and no need of financial assistance.

Is a Washington-Baltimore-Richmond Mega-Region in Our Future?

The Boston-Washington corridor

In 2008 economic geographer Richard Florida argued in his book, “Who’s Your City?”, that the economic units that matter in understanding economic growth and development aren’t nation states, or states, or even metropolitan statistical areas. They are mega-regions — conglomerations of metropolitan areas that are increasingly bound to one another through business interactions. By Florida’s reckoning, the mega-region biggest in the United States and the second largest in the world is the Boston-Washington corridor, which extends as far south as Richmond and Hampton Roads.

I long thought of the idea of a mega-region as a meaningless abstraction — an academic concoction rather than a reflection of economic reality. Metropolitan areas, which describe definable labor markets, are the primary units of economic development. But two news stories today have forced me to consider the possibility that MSAs are not immutable if the will exists to transcend them.

First, the Greater Washington Partnership, created last year, has issued a vision statement for “the Capital Region” encompassing the Baltimore, Washington, and Richmond metropolitan statistical regions. Admittedly, that’s a far cry from a megalopolis stretching all the way to New York and Boston, but it’s a bigger than anything that exists now in Virginia or Maryland. The economy of the Capital Region, proclaims the organization’s website, is the third-largest in the United States and seventh largest in the world.

States the website: “By acting together, and focusing on super-regional solutions, we can overcome jurisdictional impediments, achieve solutions at a scale that is equal to the problems we face, and deliver new sources and engines of growth to achieve economic well-being and prosperity.”

In the Richmond Times-Dispatch today, Michael Martz quotes Dominion CEO Thomas Farrell, one of 21 corporate CEOs on the partnership’s board, as saying,  “The Greater Washington Partnership can make an impact on such pressing issues as transportation and talent, if those issues are addressed regionally.”

The overarching goal of the CEOs is to attract talent and promote innovation. A law of knowledge-economy economics, known as the agglomeration effect, is that larger regions exert greater gravitational pull on talent and corporate investment than smaller regions. The implication: Washington, Baltimore and Richmond are all stronger if they function as a single big region rather than three smaller regions. The incredible power of the agglomeration effect drives the growth of mega-regions, and it is the primary justification for building ties between neighboring regions.

Now, it’s one thing to proclaim a common identity, and another to achieve it. One can easily envision Washington and Baltimore as a single MSA because the entire swath of land between the two core cities has been filled in and developed. As a result, the labor markets of the two regions overlap to a significant degree. The same cannot be said of Washington and Richmond. But ties between Richmond and Washington, though tenuous, are emerging.

That brings me to the second news item. The Stephen Fuller Institute has just published a study, “Migration in the Washington Region: Trends between 2000 and 2015 and Characteristics of Recent Migrants.” The Washington region has a problem. While its population continues to grow as a result of foreign immigration and a surplus of births over deaths, the region has been leaking native-born citizens.

Between 2000 and 2015, Washington has experienced a net domestic migration to the Baltimore area of 77,000, and to the Hagerstown-Martinsburg area of 35,000. The number three and four recipients of Washington out-migration were Winchester (16,000) and Richmond (14,000). Charlottesville (4,000) was 15th largest recipient of domestic out-migrants. While downstate Virginia’s ties to the Washington region aren’t as strong as Maryland’s, they are still substantial. (Interestingly, Hampton Roads shipped a net 14,000 population to Washington over the same period, a pattern no doubt influenced by military ties between the two regions.)

When Washingtonians leave the metro area, by and large, they aren’t moving to New York, Boston or Philadelphia. Some are moving to retirement areas in Florida or the Eastern Shore, and a few to Charlotte and Raleigh. But the overwhelming majority are settling nearby — in the Baltimore, Hagerstown, Winchester and Richmond regions.

In other words, while the business CEOs speak grandiosely about pulling the three regions together, they aren’t trying to make something out of nothing. Below the radar screen, thousands of households making decisions of where to live and work implicitly recognize a commonality not reflected in government statistics.

If the political class buys in to the idea of a Baltimore-Washington-Richmond mega-region, the single-most important thing it can do is to knit the regions together with better transportation infrastructure. Saying this goes against my grain because I am suspicious of infrastructure mega-projects of all kinds, which invariably turn out to be boondoggles. But adopting the view of economic strategist rather than fiscal scold, I would say that top priorities would be: fixing the Washington heavy rail system, creating a higher-speed rail system from Richmond to Washington, and completing the extension of the Interstate 95 tolled express lanes to south of Fredericksburg. If we want to make a mega-region a reality, then we must invest in transportation infrastructure that enables people to move easily between the component regions.

One more thing. If Virginians want to become part of an economically competitive mega-region, they need to cast aside traditional resentments between Northern Virginia and the Rest of Virginia, NoVa and RoVa. Legislators must transcend their parochialism and prioritize projects of regional value, even if it means deferring local needs, in the expectation of everyone gaining something greater in return.

Second Chart of the Day: Unemployment

Source: Commonwealth Institute

Another chart from the Commonwealth Institute based on the latest U.S. Census data: poverty rates across Virginia metro areas.

Here’s what leaped out at me: Every single metro area, from Harrisonburg to Winchester, had a poverty rate below the statewide average of 11%. How high must the poverty rate for non-metro (aka rural) Virginia be to skew the numbers in such a way? As Augie Wallmeyer says, there are two Virginias.

Chart of the Day: Median Household Income

Source: Commonwealth Institute

Here’s the latest data on median household income from the Census Bureau, courtesy of the Commonwealth Institute. Nothing much new here: Residents of the Washington Metropolitan Statistical Area make 50% or more than inhabitants of Virginia’s other metros, more than $90,000, just like they always have.

There is a well-defined second tier: Hampton Roads (Virginia Beach-Norfolk), Richmond, Charlottesville and Winchester. I haven’t looked into it, but my porky sense tells me  (if you don’t get the feeble joke, porky sense is Bacon’s analogy to spidey sense) that Charlottesville is on the rise. All those horse country gentry and handsomely paid University of Virginia administrators may be pulling up median incomes.

The smaller metros — Roanoke, Lynchburg, Harrisonburg, Blacksburg, Staunton/Waynesboro — constitute a third income tier. And then there’s non-metro Virginia, which is not included in this chart, which I expect constitutes a fourth tier.

Overall, Virginians’ median income rose 1.8% last year. While incomes in the Old Dominion are relatively high — $68,100 statewide compared to the national median of $57,600 — the growth in income lagged the national average of 2.4%. Sequestration still haunts the commonwealth. Incomes in the Washington metro, only 1.5%, dragged down the state average. Once the highest-income metro in the United States, Washington now lags San Jose and San Francisco.

As an aside… the Commonwealth Institute notes that “communities of color” — African-Americans and Hispanics — tend to have much lower incomes on average, citing “structural barriers” such as poor schools, housing discrimination and employment discrimination. Given the fact that it was citing Census data on household income, the think tank appears to have missed an excellent opportunity to examine the contribution of household size and structure on income levels.

One of the biggest contributors to household income is the number of bread winners in the household. If African-American and Hispanic households are more likely than communities of pallor to consist of single-income households — as, in fact, they are — the breakdown of the family contributes in a direct and measurable way to reduced median household income.

Americans Increasingly Skeptical of Value of Four-Year Degrees

Graphic credit: Wall Street Journal

Americans are finally getting wise to the value of a four-year college degree. A Wall Street Journal/NBC News poll finds a significant growth in skepticism over the past four years, especially among Americans who haven’t graduated from a four-year college.

Overall, 49% of Americans believe that earning a four-year degree will lead to a good job and higher lifetime earnings, compared to 47% who don’t — a two-percentage point gap. Four years ago, that gap was 13 points.

Skeptics number in the majority — 57% to 37% — among Americans 18 to 34 years old. That should come as no surprise, as that age group has taken on a disproportionate share of the $1.3 trillion in outstanding student debt and is having the greatest trouble repaying it.

A majority of women still have faith in the four-year degree, reports the Wall Street Journal, but men’s attitude has undergone a dramatic reversal. Four years ago, men saw college as worth the cost by a 12-point margin; today they say its not, by a 10-point margin.

Many observers pushed college attendance on the astonishing superficial grounds that college graduates on average earn higher salaries and experience a lower unemployment rate than those who never went to college. What such analysis ignores is that the average earnings and unemployment for all college grads is not necessarily typical of earnings and unemployment of college grads on the margins, who were less academically prepared, received lower grades, attended less prestigious institutions. It also ignores the ugly reality of millions of Americans who racked up large debts attending college but failed to graduate.

Awareness is spreading that people can earn solid middle-class wages with a couple of years of technical training, without losing two years of earnings attending a four-year college or spending tens of thousands of tuition, fees, room, and board. The WSJ gave a great example:

Jeff McKenna, a 32-year-old from Loveland, Colo. said he doesn’t believe college is worth the cost. Mr. McKenna went to a trade school, earning a certificate as a mechanic and how earns a base salary of $50,000 a year. He said he has never gone three weeks without a job, including during the recession.

“I have friends from high school that are making half what I’m making, and they went and got a four-year degree or better, and they’re still $50, $60, $70,000 dollars in debt,” Mr. McKenna said. “There’s a huge need for skilled labor in this country.”

Indeed there is. As more people — young men, mostly — think like Jeff McKenna, there will be a growing demand for community colleges and trade schools that teach marketable blue-collar skills. Skepticism runs greatest in the college-age population, making it likely that four-year colleges will find it increasingly difficult to maintain their enrollments. Those at greatest risk are institutions that appeal to precisely those demographics — rural, lower-income, male — where skepticism runs the deepest.

As Virginia Inches Toward Becoming a Majority Minority State, What Do Racial/Ethnic Classifications Mean Anymore?

Whites will comprise less than a majority of Virginia’s population by 2040 — 47.4% — according to recent projections by the Demographics Research Group at the University of Virginia. That’s down from a forecast 58.6% in 2020.

The percentage of non-Hispanic whites and blacks in the state’s population will shrink by 5.7% and 8.3% respectively, while the percentage of Asians and Hispanics will increase by 96.0% and 114.3% respectively.

To some degree, demographic projections reflect underlying demographic reality. But they also are influenced by politics and culture, as Hamilton Lombard points out in a post yesterday on the StatChat blog. “It can be easy to read too much into very long term population projections,” he warns. “All the racial/ethnic projections only make sense if you understand the haphazard way we categorize and track race in the U.S.”

For example, a large majority of Hispanic Americans self-identify as white, but the Census Bureau categorizes them as a “non-white” minority because they also identify as Hispanic. Before 1970, they were categorized as white.

But after the Civil Rights Acts of the 1960s, the National Council of La Raza successfully lobbied to have anyone with a “Spanish origin” counted as a separate ethnic population in the 1970. Armed with data for the newly categorized Hispanic population which the 1970 census supplied, organizations could apply for various grants and develop policies specifically for Hispanic Americans. Other groups, after seeing the success of La Raza, have lobbied to have various U.S. ethnic populations counted separately. As a result, the number of race/ethnic categories on the census has risen from four in 1960 to possibly nine by 2020.

Another example of how the Census Bureau shapes perceptions of ethnicity and race: Since the Census began allowing respondents to identify as more than one race in 2000, the U.S. “mixed race” population, 86% of whom select white as one of their races, has grown from zero to nearly 10 million.

Yet another example: Census has proposed counting Middle-Eastern and North-African Americans as a separate race. Because most self-identify as white, the new classification would accelerate the decline of the “white” population and increase the “non-white” population.

Bacon’s bottom line: Two mega-trends are colliding here. On the one hand, the Great American Assimilation Machine continues to do its work, eroding ethnic and racial identities. On the other hand, by creating a racial spoils system (dispensing funds and perks to non-whites), government policy creates material incentives for people to nurture separate ethnic identities.

A century ago, white ethnic identities such as English, Scotch-Irish, Irish, Italian, German, Polish, Swedish, Jewish, etc. were as strong as racial identities today. Over time, intermarriage and the dissolution of ethnic enclaves merged white Americans into the melting pot. Today, white Americans are less likely than ever to define themselves by the national origin of their ancestors and more likely than ever to simply think of themselves as generically “white.”

The Ancestry.com ads running on cable TV are a striking illustration of this trend: There would be no need to utilize DNA to identify peoples’ ethnic origins unless most people had forgotten those origins. I thought I was Hispanic and found out I was half Italian! I thought I was German and found out I’m a mutt!

Meanwhile, the rise of “multi-racial” populations is proceeding apace. According to a Pew Research Center analysis, one-in-seven U.S. infants (14%) were multiracial or multi-ethnic in 2015, nearly triple the share in 1980. This is not just a matter of “light skinned” ethnicities intermarrying. Increasingly, Americans are broaching the color line.

One would think that all but the racial purists among us would welcome this trend. But political forces are driving the population in the opposite direction. Many politicians believe that the path to political power lies in the cultivation of racial grievances. These politicians (I won’t mention names) exist in both parties. By enabling the doling out of government spoils (usually at the behest of the political party that favors activist government — but, again, I won’t mention names), the Census Bureau’s ethnic/racial classifications perpetuate the sense of separateness.

It is impossible to predict which force — assimilation or the urge to political power — will win out in the end. But we can count on one thing: Changes in politics and culture undoubtedly will influence which races and ethnicities the Census Bureau Sam counts, and, consequently, how Americans perceive themselves. In the meantime, readers should understand Census ethnic and racial classifications for what they are: increasingly meaningless distinctions imposed and maintained for political reasons.