Oxycodone pain pills prescribed for a patient with chronic pain. (John Moore/Getty Images)

IT IS a political cliche that President Trump owes his electoral victory to the extraordinary support he received from white voters without a college degree, two-thirds of whom voted for the Republican. Much less settled is the question of why these largely low-income voters, once reliable Democrats, cast their lot with a brash billionaire from New York.

The precise source of the discontent that produced this outburst of reactionary populism is hotly debated; some of Mr. Trump’s support reflects motives, such as xenophobia or racism, that can be neither comprehended nor respected. Yet there is an objective aspect to “white working-class” grievance. Anyone who doubts it need only read “Mortality and morbidity in the 21st century,” the new report by Nobel Prize-winning economist Angus Deaton of Princeton University and his colleague Anne Case, presented to a Brookings Institution conference Thursday. The gist is that death rates for non-Hispanic white men and women aged 25 to 64 rose steadily between 1999 and 2015, while death rates for the comparable age cohorts of all other demographic groups either held their own or continued to improve. The cumulative impact of these trends, the authors note, account for the stunning fact that overall life expectancy in the United States decreased slightly between 2014 and 2015, the first such decline since 1993, during the AIDS epidemic.

Even more troubling are the specific causes of rising mortality among non-college-educated white Americans: drug overdoses, suicides and alcohol-related liver disease, or, as Mr. Deaton and Ms. Case aptly call them, “deaths of despair.” If the despair could be cured by bringing back high-paying jobs that built the erstwhile “blue collar aristocracy,” as Mr. Trump promised during his campaign, then there might be cause for hope. However, the authors demonstrate that “deaths of despair” do not seem connected to falling income; otherwise, mortality would also have risen, not fallen, for Hispanics and African Americans, whose incomes fared no better than those for whites.

Mr. Deaton and Ms. Case blame “cumulative disadvantage,” whose components include not only job loss but also the breakdown of communities and the decline of marriage. That seems plausible. But it isn’t immediately clear how government could reverse those long-term trends, although it is clear that Mr. Trump’s policy agenda won’t do much for the people who gave him their votes, and might hurt them.

The study’s authors do emphasize one relatively controllable factor: mass prescription of opioids for chronic pain, which made the epidemic of despair-deaths “much worse than it otherwise would have been.” The rate of death from prescription opioid overdoses more than quadrupled between 1999 and 2015, according to the Centers for Disease Control and Prevention (CDC), killing a total of 183,000 people during that period. Prescriptions, though not yet overdose deaths, have started to ebb in recent years, a trend fostered by policy shifts at the CDC in the latter years of the Obama administration.

If he does nothing else to keep faith with the people who elected him, Mr. Trump must at least continue the more rational policy on opioid prescribing that began, belatedly, under his predecessor.