Are demographic shifts the key to American voting patterns? The particular election results selected for testing this hypothesis matter.
The last time Republicans won the popular vote decisively in a presidential election was in 1988 when George H.W. Bush defeated Michael Dukakis by eight percent. Approximately 85 percent of voters in 1988 were white (non-Hispanic white), and 15 percent were African American, Hispanic, and Asian. Bush won the two-party vote among these white voters by approximately 20 percent and Dukakis won among minority groups by 60 percent.
In 2016, the Republican candidate Donald Trump again won the two-party vote among whites by 20 percent and lost to Hillary Clinton among the minority groups by the same 60 percent. But this time around, whites only comprised about 72 percent of the two-party vote, and Democrat Hillary Clinton won a narrow popular vote victory of just over two percent.
In other words, the vote share by party within major voting blocs stayed the same over 28 years, but a shift in the share of the total vote for the two groups of 13 percent from whites to minorities, significantly altered the popular vote result by over 10 percent in total. Donald Trump, despite losing the popular vote, won an Electoral College majority by winning states that had gone Democratic in six or seven straight elections – Michigan, Pennsylvania and Wisconsin – all by tiny margins of one percent or less. He also won two other Midwestern states – Iowa and Ohio – which had gone for Democrats in 2008 and 2012 by healthier margins, as well as shifting Florida to the GOP column with a margin of just over one percent.
Shifting Population
The shift in the shares of the American electorate among whites and minorities has occurred due to a shift in the shares of the population, though the population shift has been more rapid. This reflects the fact that the fastest growing groups – Hispanics and Asians – include significant numbers of people who are either not citizens or are not registered to vote, or don’t show up to vote in the same degree as whites and blacks. In any case, the shift in both the population shares, and the voting group shares continues at a steady rate. Whites now make up just over 60 percent of the population, and just half of live births. Almost 19 percent of the population is Hispanic, close to 13 percent are African American and six percent are Asian American. Most estimates are that non-Hispanic whites will be a minority of the population sometime in the 2040s. This trend-line reflects several factors including:
Immigration – barely 10 percent of legal immigrants come from predominantly white countries in Europe, Australia, or Canada;
Fertility rates – births per woman of child-bearing age (ages 15-44) are dropping for all groups, but are higher for minority groups than for whites; and
The current age distribution of the population reveals that whites are the oldest group, Hispanics the youngest.
Democratic Majority?
Given that Republicans do best among whites, and Democrats much better among minority groups, an argument can be made that the country is headed toward a permanent Democratic majority, resembling California, the population of which has included more than 50 percent minority group members for decades, and is increasingly Democratic with every election cycle. The number of states in which minority groups form a majority of the population is increasing.
Not surprisingly, there have been attempts to quantify voting shifts that will accompany the anticipated future population changes. In 2002, Ruy Texeira and John Judis wrote a book titled The Emerging Democratic Majority. After the 2002 midterms, Republicans controlled the White House, and both houses of Congress. After the 2016 elections, Republicans controlled the White House and both houses of Congress, the House by a bigger margin than in 2002, the Senate by a smaller margin. In both the 2000 and 2016 presidential elections, Republican candidates had won Electoral College majorities despite losing the popular vote. In 2016, the margins in both the Electoral College for Republicans and the popular vote for Democrats were bigger than in 2000.
Looking at the 2016 and 2002 scoreboards, one might question whether the thesis of the emerging Democratic majority was sound. The relative strengths of the two parties in 2016 at the federal level, with Republicans in the stronger position, also showed up at the state level, where Republicans controlled roughly two-thirds of governors’ mansions and held majorities in two-thirds of state legislative bodies.
Texeira and Judis added a factor to the trend-line on group shares in their book. They divided the white population into those with a college education and those without one. While the white population share overall was declining, the share of college-educated whites was growing relative to the share of non-college educated whites. The authors provided data suggesting that non-college educated whites were more likely to vote Republican than college-educated whites, and college-educated whites were expected to be a battleground in the future with rising support for Democrats over time.
The Mistake
In 2017, Texeira admitted in an article that he had been mistaken, and political support for the two parties was not moving in a straight line towards Democratic dominance. It is possible Texeira would have moved back to his original thesis after the 2018 midterms when Democrats won control of the House of Representatives, gaining 40 net seats, picked up seven governors and majorities in six state legislative bodies.
The better argument may be that things are not so certain when it comes to politics. This starts with acknowledging that vote shares within groups are unstable and more complicated than simply whites and minorities. Minorities consist of African Americans who tend to vote around 90 percent for Democrats, and Hispanics and Asians who also favor Democrats, but not as overwhelmingly – more like 70 percent to 30 percent in recent cycles.
Among Hispanics, more than half identify as whites on the census form, suggesting that the neat division between whites and minorities is a bit artificial. Cuban Americans tend to vote at higher rates for Republicans than do Mexicans, who vote at higher rates for Republicans than do Puerto Ricans. Mexicans in California are more steadfastly Democrats than Mexicans in Texas, just as white voters are more often Democrats in California than in Texas. Relative sizes of the various population groups within the states does not explain why California consistently provides Democrats with 60 percent or greater support, and Texas still leans Republican, though less so today than in the recent past.
A Moving Population
A second factor relates to movement of Americans among states. About four percent of households in the United States change their principal domiciled state each year. Americans are moving less frequently among states than in the past. But there are distinct patterns. African Americans are abandoning Eastern and Midwestern states in significant numbers to move to locations with lower crime rates, better economic opportunities, and better climates, primarily in the South (Georgia, North Carolina, Florida, South Carolina, Texas).
Hispanics are moving in the same general direction, though also to farm belt states, and whites are also moving South and Southwest for economic reasons or better retirement climates in Florida and Texas. The absence of state income taxes have made Florida, Texas, Tennessee, and Nevada big draws for all population groups. The movement of blacks out of several Midwestern states, and the lower turnout and enthusiasm among black voters for Hillary Clinton in 2016 than for Barack Obama in his two presidential runs, is a major reason for Donald Trump’s narrow victories in states once considered part of an impregnable Democratic wall.
A third factor relates to the share of immigrants to total population within the country. Current census estimates are that more than 43 million people of those here legally were born outside the United States. The immigrant share of all Americans has been rising, now over 13 percent, just a bit below the all-time high recorded almost a century ago. Legal immigration has averaged more than one million people a year for more than a decade.
But the pace may be slowing due to changes in immigration policy at the federal level. In the most recent year for which data is available, 42 percent of legal immigrants were from Latin America, 40 percent from Asian countries, 10 percent from Africa, and the remainder from Europe, Canada and Australia. In recent years, about half of legal immigrants were related to American citizens, another 20 percent were family sponsored, and more than 10 percent were refugees or asylum seekers. Other countries, such as Canada and Australia, to a much greater extent than America, choose the immigrants they want, focusing on skills. The norm in the United States, is for immigrants to choose the country, mostly tied to family connections. A more skills-based immigration system might shift spots among various population groups.
Illegal immigrants are estimated to number about 11-12 million, though one recent study by a Massachusetts Institute of Technology researcher concluded that twice that number may be here illegally. If a future Congress and president agreed on an immigration reform proposal that would grant large-scale amnesty and a path to citizenship for many or most of those here illegally, the total legal immigrant level would rise and the population share of both Hispanics and Asians would rise, with an impact on voting strength among the two parties over time. In any case, the actual population distribution reflects a far higher percentage of minorities than do the voter rolls, and the number who show up to vote. This lag is also one of the explanations for why the Texeira forecast has had a bumpier ride than anticipated.
The annual level of legal immigration to the United States has enabled the country’s population to increase as birth rates (the number of births each year per thousand women of childbearing age) have dropped to around 60 per 1,000 women. Half a million fewer babies are being born in the United States in recent years than 60 years ago, when the country was just over half as populated as today. The fertility rate (average number of births per woman of childbearing age over 30 years (15-44) has dropped below the replacement level of 2.1 children to around 1.8 or slightly below, compared to 3.55 in 1954.
The birth rate and fertility rate current in America is associated with anticipated population declines in some countries in Europe and Japan, which accept few immigrants. Birth rates in the United States have dropped for all major population groups, but fastest among Hispanics, among whom the birth rate is down almost 30 percent from levels a decade back (95 per 1,000 women per year to 68 per 1,000 women per year). Whites have the lowest birth rate (57), just below that of Asians (59), and African Americans (63) and all groups are now below replacement level (about 70 per 1,000 women per year). Multiplying the annual birth rate by 30 and dividing by 1,000 produces the current fertility rate.
Education Divide
The biggest voting shift in the 2016 race occurred among whites in communities outside metropolitan areas, primarily in the Midwest and also in Pennsylvania and Florida and particularly among non-college educated whites. Trump won in Southern white rural areas by enormous percentages, and by very large margins in similar counties in some of the Northern and Midwestern states. However, another shift that first surfaced in 2016 and was a much more significant factor in 2018, was the movement of college-educated suburban voters towards the Democrats, particularly among women.
How much of this movement was due to a personal distaste for President Trump or whether it represents a permanent shift of allegiance between the parties is not clear. But this has been a factor in making Texas, North Carolina, Georgia and Arizona more competitive, and has shifted Virginia, and Colorado to the Democratic column. The growth in the Hispanic share of the population is the major reason Nevada, and Colorado now consistently vote Democratic, and Arizona has become a battleground.
The conclusion that seems most defensible given the recent voting results and demographic shifts is that Republicans need to do better among minorities to remain competitive and need to quickly reverse the trend away from the party among suburban voters, the largest population group in the country.
Economic growth of the past few years has produced record low unemployment rates for blacks and Hispanics. Will Republicans get any credit for this in the presidential race, or congressional races in 2020? The gender gap between men and women is expanding among all population segments. There are some indications that Trump is making some inroads among African American male voters. So too, the controversy surrounding affirmative action and preferences for admissions to the most competitive colleges and universities has pitted the Asian community against their presumed minority group political allies. Will this be reflected in changes in election voting patterns?
There is also a commonly held assumption that Hispanics are generally supportive of very liberal immigration policies including the relaxation of border controls and amnesty for illegals already here. But those who work in lower paying jobs, which includes significant numbers among all minority groups, might resent illegal immigrants willing to work for still lower wages, thus competing for their current jobs.
Demographic changes – immigration levels, movement of people among the states, and birth rates will all change the relative voting shares of major groups nationally and in individual states. A party that can appeal to various groups will do better than a party reliant on a shrinking base of voters in one group only. In this sense, the results of one election can be a misleading signal on general competitiveness and appeal in the future.
Richard Baehr is the chief political correspondent for American Thinker and also writes regularly for Israel Hayom and PJ Media.
