One (admittedly crude) way of measuring the nation's leftward drift is to assign "left" and "right" labels to the presidential candidates in every quadrennial election (including candidates of third, fourth, and fifth parties), and then to determine the fraction of the total popular vote that went "left" or "right". I have done this for the 26 elections spanning 1920-2020. I chose 1920 as the starting point because that year marks the birth of the modern (post-Teddy Roosevelt) Republican Party -- the rhetoric (if not the practices) of which favors smaller government, reliance on free markets, toughness on crime, traditional (Judeo-Christian) morality, etc., etc., etc.
The resulting measure of leftness vs. rightness, to which I will come, is admittedly crude for at least these reasons:
Some candidates have in fact been less "right" or "left" than their predecessors and successors. That is to say, the core measures of "left" and "right" -- the percentages of votes going to Democrats and Republicans" -- aren't fixed. GOP candidates since Barry Goldwater, for example, have supported Social Security and its offshoots (Medicare and Medicaid) and one of them (George W. Bush) enlarged it by pushing for prescription-drug coverage when he was president.
Labelling a third (or fourth or fifth) party as "left" or "right" is sometimes iffy. Take, for example, H. Ross Perot's run for the presidency in 1992, which garnered him 19 percent of the popular vote. His appeal was broad and it is believed that he drew votes equally from those who would otherwise have gone for George H.W. Bush or Bill Clinton. But a lot of otherwise-GOP votes went to Perot because GHWB defaulted on his "no new taxes" pledge. Aside from taxes, Perot's platform was conservative (e.g., balancing the budget, reducing reliance on foreign trade) and his running-mate, James Stockdale, was a hard-core conservative.
With those caveats in mind, consider the leftward drift of the electorate since 1920. First is the election-to-election split:
Next is the election-to-election shift:
Finally, and most significantly, is the cumulative shift from 1920 to 2020:
The Great Depression and the Goldwater-scare election of 1964 were blips. The trend since 1972 is unmistakably leftward. What's behind the shift, which (for reasons discussed above) is really more leftward that the numbers suggest? I put it down to three things:
Women got the vote and their tendency to vote gradually overtook and surpassed that of men.
The black vote emerged from Southern suppression in the 1960s, and the tendency of blacks to vote gradually increased relative to the tendency of whites.
The young-adult vote got a big boost with the passage of Amendment XXVI, which lowered the voting age from 21 in most jurisdictions to 18 in all jurisdictions.
What's next? Illegal immigrants will be given the right to vote in more and more jurisdictions, and then they will be granted amnesty and full citizenship. The long-promised permanent Democratic majority will be upon us, and today's morally bankrupt regime will morph into overt Stalinism.
See also "The Course of the Mainstream", "What's Going On? A Stealth Revolution", "The Fickle Electorate", "The American Electorate's 'Squishy Center' vs. Liberty", and "Another Measure of Political Polarization: The Winner's Share of the Popular Vote in Presidential Elections".