I was born in 1949 into a pretty typical lower-middle-class American family. My father, like about half of all Americans (47%) at the time, had left school before getting a high-school diploma to go to work when his father died. My mother, twelve years his junior, had a high-school diploma. Together they ran a small business making awnings for homes and businesses. Few people with college degrees lived in our small, rather rural community south of Boston. Those that did were typically professionals like doctors and lawyers. Among my neighbors, only the elderly lady across the street who once taught school might have held a college degree.
Up until 1970, that was mainstream America, where four-out-of-five adults had a high-school diploma or less, and about a fifth had some college experience or a degree. Since then, the proportion of American adults who hold a college degree has more than doubled. In the two decades between 1970 and 1990, the proportion of adults with at most a high-school diploma fell by twenty percentage points. The next three decades saw an equal decline. By 2020 the share of Americans with a college degree equaled the share with high-school diploma or less.1
It is hard to imagine that a sociological change of this magnitude would not have profound political effects.
Most political scientists agree that better-educated people participate more often and more widely in political life. This finding spans decades and nations. As far back as 1970, Verba, Nie, and Kim published a path-breaking study of participation based on surveys taken in seven different countries. In all of them people with higher levels of “socio-economic status,” typically wealth and education, participated more often in politics.
In a nation like ours where there are so many institutional obstacles to participation, education probably matters more for voting than in other countries. Better-educated citizens will be better able to navigate the intricacies of the electoral process like personal registration, locating one’s polling place, and having enough free time to vote. The relationship between more education and higher turnouts has persisted for decades.
We can combine these figures to see how the composition of the electorate has changed. If we multiply the shares of the population by educational attainment and the share of each group that voted, we can estimate the share of the total electorate that each group represented.
Because better-educated people turn out at higher rates, the electorate has become more and more dominated by college-educated citizens over the past decades.
For most of the postwar period, Americans with college degrees cast their lots with the Republican Party, while those with a high-school diploma or less voted for the Democrats. This chart tracks the Presidential election choices of Americans with a college degree or more versus those of Americans with a high-school diploma or less, using the archive of data from the American National Election Studies.2
The bars represent the difference between the percent of college graduates who voted Democratic and the proportion with a high-school diploma or less. Up until 1972, the proportion of voters with a high-school education or less were more likely to vote for Democrats, while college-educated voters gave more support to the Republicans. This is hardly surprising given that only a bit over ten percent of Americans had a college degree even in 1970. College was still an elite institution, and the college-educated voted Republican in line with their interests.
Many fascinating historical tidbits appear in these results. Republicans turned out in especially disproportionate numbers for Barry Goldwater in 1964. That came to an abrupt end during Richard Nixon’s term in office. Degree-holders support for Nixon fell dramatically in 1968, and not because of George Wallace’s candidacy; he drew just three percent in this group. Four years later, as the War in Vietnam dragged on and Watergate had penetrated the consciousness of politically-aware voters, Nixon failed to win a majority of college-educated voters, the first time for a Republican candidate since World War II.
These well-educated Americans rallied behind Gerald Ford in 1976, and to a slightly lesser degree, Ronald Reagan in 1980, but they were surprisingly less enthusiastic about Reagan’s re-election bid four years later. After that, things got worse for the Republican Party. Their margins of support among college-educated voters shrank markedly starting with the 1988 election won by George H.W. Bush. In the three elections between 2000 and 2008 college-educated voters split their support evenly between the parties’ candidates.
Barack Obama’s re-election bid drew a higher proportion of college-educated voters to his side than during his initial run for the Presidency. In fact, Obama’s initial coalition included enough voters to swing the needle slightly back to the Republicans in 2008. That came to an abrupt halt in 2016. Donald Trump’s arrival on the scene exacerbated these educational differences to an extent not seen since World War II.
The contingent of college-educated citizens is even a greater share of Democratic votes than their share of the electorate as a whole. All of these trends bode poorly for Republicans in the long run.
1The uptick for “some college” probably represents the effects of the expansion of Federal student assistance in 1992. Nevertheless, the demand for a four-year degree continued its inexorable rise, while the proportion of adults with “some college” has remained pretty flat for decades.
2Data are weighted by VCF0009z as recommended in the documentation. The 1948 data do not appear to meet the same level of quality as the later samples and have been ignored.