Wednesday, January 18, 2017

Choices, Choices

In The Week, Michael Brendan Dougherty reflects on the sudden escalation of concern many policy-makers have had about Russia in recent months.  I don't necessarily agree with all of Dougherty's claims, but he hits on a telling point here:
The problem is, America's NATO war guarantee is wrapped up in a larger ideological status quo across the West. Trade liberalization, political liberalization, increased migration, sexual and cultural liberation from Christian traditionalism, the further political integration of the E.U., and the expansion of the Western alliance to the borders of Russia are all wrapped together in the minds of policymakers. And so, every reversal for any part of that project is seen by the guardians of the policy consensus as a demoralizing reversal for the Western alliance and, consequently, a gain for revisionist Putinism.
The international institutions that have been left to us have evolved from a certain set of geopolitical circumstances and partake of certain policy presumptions.  It seems (to me at least) that many of these institutions have done considerable good, but the question before us isn't whether they've done good in the past but how to preserve and revise these institutions so that they can do good in the future.

That's one of the major reasons why ideological nostalgia has been so toxic for the enterprise of looking forward for both foreign and domestic policy.  This nostalgia has made many current leaders resistant to--and perhaps even ignorant of--the fact that we no longer live in 1989.  For instance, for years, public skepticism about the European Union has been simmering throughout Europe.  And yet policymakers nevertheless plunged ahead with an integration that became less and less tenable as time went on.  And now many of these policymakers are now shocked, shocked that some voters might be having second thoughts about the "European" project.  Taking a more moderate approach to integration (especially in terms of immigration and currency.) and responding more to immigration could have helped solidify the EU.  Instead, a reckless integration threatens breaking it apart.

The importance of attention to present-day realities has implications beyond the EU, of course.  In defending international institutions, policymakers need to think of the world not as it was or as they would like it to be but as it is.

We can have either a robust international order or ideological nostalgia--but not both.

Wednesday, January 11, 2017

Repeal and What?

In my NRO piece earlier this week, I warned about the political dangers of Republicans repealing the Affordable Care Act (well, actually only repealing the provisions of the ACA they can repeal without a filibuster-proof majority) without following up with a replacement relatively soon.

There have been a couple updates today on the fate of the repeal-and-delay strategy.  At his press conference today, Donald Trump argued that replacing the ACA should follow very soon after the repeal of the measure.  Meanwhile, some GOP senators who had been pushing for an amendment to delay a vote on health-care repeal have now said that they're going to drop this amendment.

Monday, January 9, 2017

Resisting Ideological Nostalgia

In NRO today, I argue that the GOP should learn from the last time a political party took full control of the federal government.  In 2009-10, Democrats made certain choices that set the stage for a multi-cycle liquidation of the part.  In order to avoid that fate, the GOP will need to respond to the demands of the moment and resist ideological nostalgia.

Thursday, January 5, 2017

Hacking the Vote

In reporting the hacking of the email messages of prominent Democrats during the 2016 election, the press has by and large settled on a rather intriguing shorthand: "election hacking."  (See examples here, here, and here.)  Because of its vagueness, this shorthand risks conflating the hacking of emails in order to influence the election in some way (which there is some evidence of) and the hacking of the election results themselves (which there is no evidence of).  "Email hacking" or "DNC hacking" or "Podesta hacking" or "hacking of Democrats" would seem more precise but also less likely to inspire paranoia in the American public.  "Election hacking" may offer a kind of argumentative figleaf for those who want to delegitimize the election results without openly saying that some nefarious entity actually went in and hacked the vote totals to swing the election to Donald Trump.

That's not to say that some folks won't try to argue that the election results themselves were hacked.  In the latest issue of the New York Review of Books, a publication I enjoy, Michael Tomasky argues that we will never really know whether a foreign entity (in his case, Russia) hacked the election:
But if their reports are accurate, what this amounts to at the very least is that Russia tried to influence the outcome of the election in Trump's favor. Whether it managed to determine the outcome by meddling directly in the actual voting is something we don't know and will likely never know.  To arrive at such a conclusion would require a thorough forensic investigation of vote tabulations in at least the three states where Trump's margin over Clinton was less than one percent--Pennsylvania, Michigan, Wisconsin--and other steps; but this is not going to happen.
A belief that Russia hacked the actual election results has some popularity on the left (according to one poll, 52 percent of Democrats believed that it did), but there is absolutely no evidence that any such vote hacking did take place.  U.S. officials have affirmed the accuracy of the voting results, and state recounts have uncovered no evidence of vote hacking.

While it might assuage Democratic feelings to keep alive the evidence-free narrative that Russia hacked the election results, it's damaging to our body politic in general.  If that election was hacked and if we can keep alive that hypothesis without any concrete evidence, why not say the same thing about other elections?  How do we know that any elected official actually won his or her office?  If Russia can hack the election results without any obvious evidence, how do we know that other entities also have not hacked this election or other elections?  Electoral paranoia destroys the public trust that our republic is built upon.

Moreover, if it were true that Russia hacked the election results (again, there's no evidence for this), that perhaps is the biggest indictment of the Obama administration imaginable: on its watch, its policies allowed a foreign entity to destroy the democratic process.  Thus, it would be hard simultaneously to support the notion that Russia hacked the election and to hold the belief that President Obama was anything other than the worst president since James Buchanan, maybe ever.  The left can have Barack Obama as a noble president or the proposition that Russia hacked the election results--but it can't have both.

We have more than enough real challenges at the moment.  We don't need paranoia to invent more.

Saturday, December 31, 2016

Fear Itself

In March 1933, Franklin Roosevelt spoke to his country at a time of great crisis:
So, first of all, let me assert my firm belief that the only thing we have to fear is...fear itself — nameless, unreasoning, unjustified terror which paralyzes needed efforts to convert retreat into advance. In every dark hour of our national life a leadership of frankness and of vigor has met with that understanding and support of the people themselves which is essential to victory.
In many respects, we seem in a moment of some crisis today.  Extended economic stagnation, proliferating debacles abroad, and cratering trust in public institutions are some of the signs of this crisis.  Roosevelt reminded his countrymen that the crisis of his time was as much one of sentiment as it was of exterior forces.

A similar point applies to our own time: Many of the challenges we face are internal--not external.  And even many of the external challenges (in foreign affairs, for instance, or international terrorism) will depend upon our internal constitution.  A fractious and divided nation will unable to confront many of the real exterior threats facing the United States.

So in the year ahead, we need to be vigorous in addressing those internal challenges honestly, fairly, and imaginatively.  Decadence can take many forms.  We face a governing caste that all too often confuses self-righteousness with virtue and cant with learning.  A vulgarized political and cultural discourse inhibits our ability to draw measured distinctions.  The withering of civil society threatens the maintenance of the norms upon which republican governance depends.  In part due to our elites' emphasis on no-choice politics, our public debates remain frozen in tired antagonisms.  The recourse of the powerful to shame politics risks empowering shamelessness, which itself can be troubling; courtesy and empathy--the quest to mediate difference--often prove crucial for defending the architecture of a free society.  As can be seen in the wake of November's election results, tantrum politics might amuse political partisans, but it usually only worsens our broader civic fever.  Indulgence in a paranoia fed by cryptic inference can be hardly afforded now, when we need a rejuvenation of civic faith.

These challenges are vast.  They cannot be traced to one man or women.  They cannot be settled by a single election.  They cannot be solved by one political actor.

Dealing with them will require virtue and judgment in our daily lives.  Thankfully, virtue can take many forms, too.  It can be witnessed at a volunteer at a charity soup kitchen, in a person seeing the best in another, in a parent teaching a child, in the good works and good words of countless religious organizations, and in a person sitting down and grappling with with the intricacies of Plato or Aquinas.  We can advance the restoration we so desperately need by rebuking partisan blindness, by cultivating a sober love of liberty, and by recognizing our own fallenness and what our nobler hopes depend upon.

The American people have inherited a great thing in the Republic.  The challenges that we face can be addressed.  We are not fated to decline and disunion--not yet, at least.  In 2017 and beyond, we need to put away the childish things of despair and vanity and instead take on our duties to our fellow man and to our own higher callings.

Monday, December 19, 2016

Apples and Oranges

A common bit of punditry over the past month has been to compare Donald Trump's margin of victory to that of congressional Republicans.  This type of analysis, however, might be a kind of apples-to-oranges comparison, as the dynamics of congressional races, especially House ones, differ substantially from the dynamic of the presidential race.

One of the key differences between those two kinds of races is the power of incumbency.  Outside of wave elections or significant personal misconduct, congressional incumbents are notoriously difficult to dislodge.  In 2016, about 97 percent of House incumbents on the general-election ballot won reelection.  In the wave year of 2010, the House reelection rate slipped down--to a mere 85 percent.  House incumbents can often blow out their challengers with higher name recognition, overflowing campaign coffers, and a seasoned political team.  This leads to a situation where, in a presidential election year, House incumbents often do better than their presidential counterparts in a district.  When a political party has a majority in the House, that advantage is compounded, leading to a situation where the majority party in the House gets more votes than its presidential standard-bearer.

It's hard to compare House and presidential electoral performance over an extended period of time because, in recent decades, House elections have become much more tied to presidential politics.  Ticket-splitting still lives, but it is much diminished from the middle of the 20th century.  That caveat aside, over the last twenty years, the majority party in the House usually performs better than its presidential nominee.

In the chart below, I give the raw popular-vote percentage followed by the margin in parentheses:
House Republicans: 48.2 (0)
Bob Dole: 40.7 (-8.5)
Difference between margin of majority party and that of nominee: +8.5
House Republicans: 47.6 (+0.5)
George W. Bush: 47.9 (-0.5)
Difference between margin of majority party and that of nominee: +1
House Republicans: 49.4 (+2.6)
George W. Bush: 50.7 (+2.4)
Difference between margin of majority party and that of nominee: +0.2
House Democrats: 53.2 (+10.6)
Barack Obama: 52.9 (+7.2)
Difference between margin of majority party and that of nominee: +3.4
House Republicans: 47.6 (-1.2)
Mitt Romney: 47.2 (-3.9)
Difference between margin of majority party and that of nominee: +2.7
House Republicans: 49.1 (+1.1)
Donald Trump: 46 (-2.1)
Difference between margin of majority party and that of nominee: +3.2
Now, if one wanted to be truly rigorous about this whole comparison, one would have to take some further numbers into account: the size of the House majority heading into the general election (the bigger the majority, the bigger the compounding advantage of incumbency), the number of vacant and non-contested House seats, and the rate of House-race nationalization.

Nevertheless, the back-of-the-envelope chart above suggests that, relative to House Republicans, Donald Trump performed better than Barack Obama did relative to House Democrats in 2008.  Democrats had a smaller majority heading into 2008 than Republicans did heading into 2016, which suggests that House Republicans in 2016 had a greater incumbency advantage than House Democrats did in 2008.  If you're looking for evidence that Trump was radically weaker than a generic Republican candidate, the overall House vote might not provide it.

As with Senate elections, House elections have a baked-in power of incumbency, which is one of the reasons why trying to compare the performance of incumbent members of Congress to presidential candidates is a tricky enterprise.

Friday, December 16, 2016

Restoring Public Debates

At NRO today, I argue that the attempt to impose arbitrarily narrow bounds on public debates has helped cause some of the current political crisis.  On immigration, flag-desecration, entitlements, and other issues, there's often a significant gap between public opinion what pundits view as acceptable bounds of debate. Rather than trying to cast mainstream views as belonging to a fringe, we would be better off engaging in good-faith efforts of persuasion.

Tuesday, December 13, 2016

Just Say What?

Monitoring the Future, a project that tracks teen drug use, released some new data for 2015.  Fortunately, there has been a decline in many areas of teen use of illicit drugs over the past year.  The numbers swing a bit from year to year, but the overall trends can be revealing.

It's an assumption in many contemporary political discussions that the "Just Say No" campaign and, more broadly, Nancy Reagan's campaign against teen drug use were failures.  However, the broader trends of Monitoring the Future show a substantial and long-term decline in teen drug use during the 1980s.  Media headlines to the contrary, teens today are less likely to use illegal drugs than they were in the early 1980s (this is not to discount the recent explosion of opioid-related deaths, a serious concern but one distinct from teen drug use).  During the sustained anti-drug campaigns of the 1980s and early 1990s, teen drug use plummeted.

In 1979, over 54 percent of high school seniors (see Tables 5-1 to 5-4) had used an illicit drug within the past 12 months; by 1992, that had been cut in half, to 27 percent.  In 1979, a majority of seniors had used marijuana in the past year, but only 22 percent had in 1992.  Similar declines can be seen in the number of high-school seniors who had used illegal drugs in the past month, and, in certain categories, huge declines were seen in the daily use of drugs.  For instance, in 1979, almost 10 percent of seniors reported being daily pot users, a number that collapsed to 2 percent by 1991.  In many areas, the use of drugs has increased from the nadirs of the early-to-mid-90s (alcohol is an exception to this; it has fallen since the mid-90s).  For example, the use of illicit drugs over the past year hovered between the low 30s and high 20s in the early 90s but reached over 40 percent in 2015; the use of illicit drugs other than marijuana has not shown the same increase.  By and large, similar trends play out for college students.  (An interesting side note is the collapse in cigarette smoking among college students.  While the percentage of college students who are daily cigarette smokers did dip in the late 80s and early 90s, it rebounded in the late 90s only to fall through the floor in recent years.  Whereas about 18 percent of college students were smokers in 2000, only about 4 percent were in 2015.)

Correlation is, of course, not causation, and there are no doubt a variety of factors that contributed to declining drug use among America's youth.  But it seems possible that the sustained White House-led anti-drug campaign played a role in a broader public turn against drug use.

Tuesday, December 6, 2016

Rethinking Fusionism

A couple links on the enterprise of mediating between conservatives and populists:

Robert VerBruggen argues that Trump's Cabinet picks represent a blend of populist and conservative tendencies.

On Twitter, Henry Olsen says that the Carrier deal underlines the need for conservatives to reinvigorate their thinking about economic policy.

Thursday, December 1, 2016

Carrying Shibboleths

At the moment, I'm not going to get into a deeper discussion of the mechanics of the Carrier deal (though I will say in passing that announcing a deal to save 1000 jobs is better optics than standing around chanting "You did build that").  But it might be worth looking at a few of the attacks against this measure--and see how these attacks may fundamentally misunderstand some of the dynamics and history of manufacturing in the U.S.

Some critics note that the Carrier deal is just a drop in the bucket of the manufacturing jobs lost since 2000.  That's quite true, but small, incremental changes can be both a foundation for further change and a narrative rallying point.  Political messaging is in part about making the small example a sign of something bigger, so it's not clear why the smallness of this deal should invalidate any worth it might have.

But an even bigger argument made against the Carrier deal is that it's fruitless--any jobs saved will soon be eaten up by automation.  It's certainly true that automation is changing employment patterns and that some jobs have been lost--and will continue to be lost--to automation.  However, we should not turn automation into a shibboleth that freezes all thinking or that offers a comprehensive catch-all for explanations of the economy.

First of all, the huge U.S. trade deficit is not caused by automation.  For years now, the United States has supported a regime that incentivizes (often artificially) cheap imports from foreign nations.  Maybe this incentivization is in the broader interests of the American economy; maybe it isn't.  But, whatever its ultimate utility, this trade regime has caused some products that might be produced domestically to be produced abroad.  A reformed trade policy might cause some of these goods to be produced in the United States, which would in turn increase manufacturing employment.

Second, it's not clear how quickly--if at all--automation will ultimately devour all manufacturing jobs.  A chart in this Brookings Institution report is revealing: it shows that manufacturing employment (in terms of raw numbers of workers) was basically stable from 1980 to 2000.  The United States only shed millions of manufacturing jobs after 2000.  Interestingly, manufacturing gained far more in productivity in the twenty years between 1980 and 2000 (about doubling over that period) than it has in the years since 2000 (only increasing by about 25 percent).  Thus, the period between 1980 and 2000 saw huge gains in technology and productivity while not losing that many manufacturing jobs; the years since have seen a sluggish growth in productivity while also hemorrhaging manufacturing jobs.  This suggests that automation may not be a sufficient explanation for the loss of American jobs in manufacturing.

Third, even if automation will destroy many manufacturing jobs, there is no reason why policy-makers shouldn't take what responsible steps they can to ensure the survival of the manufacturing jobs that remain.  If more responsible trade, tax, and regulatory policies can help some manufacturing jobs stay even temporarily, that could be a not negligible win.  Human beings are not just economic inputs or outputs, so, if major economic disruption could be put off for even a decade, that could give older workers a chance to finish their careers before retiring while also providing space for younger workers to retrain.

The past eighteen months have shredded many Beltway political truisms.  There are some policy truisms that could also use reexamination.

Tuesday, November 29, 2016

Finding Middle Ground

The history of movement conservatism as a major political force reveals the electoral alliance between conservatism and vigorous populism. In recent decades, leaving aside the most recent example of Donald Trump, Republicans have relied on populist energies to power major electoral victories: Ronald Reagan, the Gingrich revolution of 1994, and the tea-party wave all depended on populist energies (even if these energies were less than they now are). While he didn’t deliver on many populist policy aims, even George W. Bush relied upon populist optics in 2000 and 2004. This conservative-populist alliance might not always be healthy, and there is no reason for conservatives to surrender their deeper principles in order to cozy up to a populist insurgency. However, conservatives might be wise to locate areas of sympathy between conservatism and populism and work to address the broader causes of this latest populist disruption.
For those interested in the theme of populism, Noah Millman has an interesting case for the necessity of populist energies as a way of informing the preferences of those in power:
Populists may be the only ones who truly understand what democracy really is for, and that is, fundamentally, for expressing dissatisfaction. Elections force leaders to turn to the people and say: How am I doing? — and to accept the people's verdict if the answer is: Not so great.
For a large swath of the country, the answer has been "not so great" for quite some time. This year, they rendered their verdict.
And I am thankful that they did. In the absence of populism, democracy becomes a competition between groups of elites to divide the people up with maximum efficiency, such as to lower the economic cost of bidding for a majority that will deliver power. Populist revolts of the left- or right-wing variety are the primary mechanism by which the electorate can punish elites for that strategy, and force them to consider the alarming possibility of losing control of the political economy entirely.

Wednesday, November 23, 2016

The Virtue of Gratitude

Thanksgiving is tomorrow, so I thought I'd take this opportunity to outline briefly some of the reasons why gratitude is something very much worth celebrating.

Gratitude calls upon us to think about what is good in our lives.  Reflecting on the good is good for the heart, but it's also good for the head.  Attending to the good helps us better understand what is actually beneficial and really worth striving for.  Inside and outside politics, the continued inquiry into the good is essential.

Gratitude also helps us realize our own partiality and our dependence upon others.  For many religious believers, the ultimate entity toward which we should have gratitude is, of course, God.  However, there are other figures deserving of gratitude as well: our families, our friends, heroes past and present, kindly strangers, and so forth.  We can feel particularly keen gratitude for gifts that are unearned--that are given out of the fullness of the heart regardless of our own shortcomings.  We can feel gratitude knowing how much we depend on others and with hope that, despite our many flaws, charity and grace can still come upon us.

Happy Thanksgiving!

Friday, November 18, 2016

The 14-Year Rule Lives

An unacknowledged winner of 2016 has been John McConnell, a veteran Republican speechwriter who served in the George W. Bush White House.  According to Jonathan Rauch, McConnell outlined the following rule in the early 2000s: "No one gets elected president who needs longer than 14 years to get from his or her first gubernatorial or Senate victory to either the presidency or the vice presidency."  As Jeffrey Anderson noted in The Weekly Standard last year, that 14-year rule holds for every presidential election since 1860.  Thus, once a presidential aspirant gets elected to the Senate or the governor's mansion, he or she has 14 years to make it to a winning presidential ticket (either as VP or president) if he or she hopes to be president at some point.  Obama had 4 years between being elected to the Senate and winning the White House, and George W. Bush and Bill Clinton had 6 and 14 years, respectively, between winning their first gubernatorial race and the presidency.  George HW Bush had never been elected to the Senate or a governorship, so the 14-year rule did not apply to him.

There's obviously no metaphysical reason why this 14-year rule has to be the case, but Rauch and Anderson have some interesting reflections on why it is a trend.  It seems to suggest the American people's preference that presidents have some experience but also represent something politically fresh.

In 2016, Donald Trump (never previously elected to any government office) faced off against Hillary Clinton, who was first elected to the Senate in 2000.  Secretary Clinton was thus 2 years past the 14-year rule.  (Interestingly for Democrats, Joe Biden would also have been well past the 14-year rule; first elected to the Senate in 1972, it took him 36 years to make it to the vice-presidency.  Bernie Sanders, however, would not have run afoul of the 14-year rule; he was first elected to the Senate in 2006.  Nor would Hillary Clinton have been past the 14-year rule if she had been the Democratic nominee in 2008 or if Obama had picked her to be vice-president in 2008.)

When Donald Trump beat Hillary Clinton, the 14-year rule won again.

Do Not Press the Button

At NRO, I mount an old hobbyhorse and criticize yet again the nuclear option on the filibuster:

The nuclear option would lead to a situation where senators — the legislative voices of their states — would lose their independence. The challenges facing the nation are complex indeed, and the nation needs a multiplicity of voices and brains working to solve those problems. In recent years, the Senate has benefited from a variety of reformers — from Jeff Sessions to Mike Lee to Marco Rubio — working to propose solutions to both new and old problems. Detonating the nuclear option would help centralize and stiffen the Senate when the moment calls for decentralization and flexibility.
The results of last Tuesday remind us that seemingly permanent majorities can be anything but. Both the Republican and the Democratic parties will face some time in the electoral wilderness in the years ahead, and the minority protections of the Senate should be there for both sides. The Obama years have led the Democratic party into one political box canyon after the next, and Senate Democrats may yet come to rue Harry Reid’s use of the nuclear option for executive appointees (because of Senator Reid’s decision, removing the filibuster for Supreme Court nominees would be far less of a shock to the congressional system than removing it for legislation).

Read the rest here.

Tuesday, November 15, 2016

Still Counting

As I mentioned last week, it takes a long time for the votes for a presidential election to be counted.  This fact didn't stop many pundits from using early totals to say that Donald Trump got fewer votes than Mitt Romney or John McCain.

Dave Wasserman has a handy running vote total, and his figures now show Donald Trump with 60,961,967 votes (this number will almost certainly continue to increase).  According to figures I've seen, Mitt Romney got fewer than 60,940,000 votes.  This implies that Trump has now passed Romney's popular vote total.

The first task of coping with political disruption is to see where one's feet stand, even if (especially if) that can be a tricky enterprise at times.

(Update: If you're interested in votes, you might also enjoy this analysis of the shake-up of electoral coalitions in 2016.)

The Party of Sam's Club, Maybe?

Perhaps because of an indoctrination in identity politics, many in the media are obsessed with viewing the results of the 2016 election through a racial lens.  Exit polls (I know, I know) suggest Donald Trump improved on Mitt Romney’s margins with all ethnic groups.  In fact, he seems to have received a smaller absolute percentage of the “white” vote than Mitt Romney.

Even as pundits fixate on racial dynamics, the changes in income and education for partisan preferences might be far more revealing.  According to national exit polls, there was a decisive swing of voters without a college degree toward the Republican nominee and an appreciable movement away from him by college-degreed voters.  In 2012, Mitt Romney lost high-school dropouts by 29 points and high-school graduates by 3; those with some college but without a college degree he lost by a single point.  Trump, however, won those who had a high school degree or less by 6 points, and those with some college he won by 9 points.  While Trump substantially improved with those without a college degree, he lost ground with college graduates, with a margin that was 8 points worse than Romney’s with college graduates and those who held postgraduate degrees (he lost the latter group by 21 points).

Income tells a similar story: Trump still lost those making under $50,000 a year, and he only fought Clinton to an essential draw among the middle and upper classes.  However, compared to Mitt Romney, he did much better with the poor and working class and worse with the wealthy.  Romney lost those making under $30,000 by 28 points; Trump lost them only by 12.  Trump did 6 points better than Romney among those making between $30,000 and $50,000, losing them by only 9 instead of 15.  Conversely, he lost about 9 points relative to Romney among those making over $100,000 annually.

Polling has had some errors this year, and the early rounds of exit polling presented a mistaken picture of the electorate.  Still, county-level electoral data confirm these trends.  Working-class counties across the country--especially in the Midwest--swung toward the GOP relative to 2012.  Meanwhile, many wealthy inner-ring suburban counties trended more Democratic.  This trend was noticeable both around coastal metropolises like Washington, DC and in middle America; Romney won Johnson County, Kansas, the wealthiest county in the state and part of the Kansas City metropolitan area, by 17 points in 2012, but Trump only eked it out by 2.  The New York Times has a handy map showing the swing of each county in the Democratic or Republican direction; most counties across the country swung more Republican, but, often, the counties that swung more Democratic are among the wealthiest in a given state.  In Minnesota, for example, the only counties that swung more Democratic were those in Minneapolis-St. Paul area.

In many of the counties on the southern border, Trump received more votes than Romney.  For example, he improved upon Romney’s margins and absolute vote total (getting 19 percent of the vote rather than 13) in Texas’s Starr County, where Hispanics constitute over 95 percent of the population.  That may be evidence that Trump did indeed improve upon Romney’s numbers with non-”white” voters.

State results accentuate and at times exaggerate this national trend.  The Rust Belt was crucial for Trump’s victory, and huge swings can be seen in these states.  In Wisconsin, Trump did over 20 points better than Romney’s margin with those who had a high school degree or less and 15 points better with those with some college, but his margin with postgraduate voters fell by 18 points compared to Romney; he lost that demographic 26-69.  He did much better than Romney with workers making under $50,000 annually, losing them only by 4 points instead of 25.  However, he lost those making over $100,000 annually by 2 points rather than win them by 20, as Romney did.  (See this interesting thread by Jeff Blehar on the county-level results of Wisconsin.  That county-level analysis tracks in many ways with what the exit polls suggest.)

In Pennsylvania in 2012, Romney lost high-school graduates by 21 points, and those with some college by 4 points.  Trump won them by about 13 points and 3 points, respectively.  College graduates and those with graduate degrees swung against Trump relative to Romney, who won college grads by 16 points and lost postgraduates by 8 points.  Trump, however, only won college graduates by 4 points and lost postgraduates by 22 points. By massively improving his standing with voters without a college degree, Trump more than made up for his slippage with college grads and advanced degree holders.  Trump also considerably improved upon Romney’s performance with mid- and lower-income voters.  While Romney lost voters making under $50,000 annually by 36 points, Trump only lost them by 12 points.

Michigan in many ways follows the trends of Pennsylvania and Wisconsin: Trump gained with lower-income voters and with voters with less formal schooling but also lost ground with upper-income and degreed voters.  In Ohio, he did better overall than Romney in many categories.  Tracking union voters sheds light on this dynamic.  In Michigan, Ohio, and Wisconsin, Trump did at least 20 points better than Romney among union voters, which helped him secure those states.

These trends could also be seen in states that Trump did not win.  For instance, in Virginia, Trump did better in 2016 than Romney did among lower-income voters, but slipped with the middle and upper classes.  The educational swing in New Hampshire caused the Granite State to become more educationally polarized than racially polarized; the gap between “white” and “non-white” voters was smaller than the gap between high-school-only voters, who backed Trump, and postgraduates, who overwhelmingly didn’t.

Again, exit polls are imperfect vehicles, but the magnitude of those swings suggests that there was likely some shift in the electorate based on income and formal schooling.  Peggy Noonan has proposed that one way to view the 2016 election is a clash between the “protected” and the “unprotected.”  In Coming Apart, the social scientist Charles Murray noted a similar divide between the secure Belmont, a fictional embodiment of white-collar wealth, and the more precarious Fishtown, an imagined locus of blue-collar struggles.  It seems as though Trump drew some of the traditionally Democratic “unprotected” to him while repelling some “protected” voters who usually lean Republican.

These results suggest great opportunities and equally great dangers for Republicans. If Donald Trump and the GOP Congress are able to govern well (by avoiding international crises, behaving ethically, and successfully responding to economic anxiety), it would seem possible that they could expand their majorities in 2018 and 2020 by winning back the college-educated and upper-income voters who used to favor them more.  Trump’s 47 percent, then, could be a foundation for a majority in 2020.  In addition to locking down the states he has already won, a marginally better performance with Belmont voters could help Maine, New Hampshire, Minnesota (where Trump won voters making under $100,000 a year), and even potentially Colorado  and Nevada tip into the Republican column.

If, however, Republicans disappoint the working class and prove the fears of many white-collar voters to be well-founded, they could be in for a massive electoral repudiation.  Blue-collar voters could desert them while college-educated voters continue to shake their heads in disgust.  Trump’s Electoral College majority seems robust, but, absent continued working-class enthusiasm, it could also prove tenuous.

A further difficulty might face Republicans hoping to build on Trump’s coalition: There might be such tensions between Belmont and Fishtown that any gains with one group would be offset by losses with the other.  If that’s the case, expanding the current Republican majority could be troublesome.  However, while there are conflicting interests (and ideologies) here, statesman-like behavior in the White House combined with prudence in legislation, which meets at least some of the demands of both groups, could establish some common ground.

With a base that is more economically anxious and more blue-collar than in the past, Republicans will need to keep worker-oriented policies at the front and center. Trade could prove an especially tricky topic; many embattled Republican senators campaigned against or were conflicted about TPP, and a majority or plurality of voters in key Rust Belt states agreed with the idea that international trade “takes away U.S. jobs.” The GOP will need to be very cautious about introducing new trade compacts and might need to embrace a more proactive case for trade reform. On immigration, Republicans could increase federal enforcement while working to implement changes to the legal immigration system so that it better encourages opportunity and fosters national belonging. Tax reform would be wise to focus on improving the condition of working families. The Affordable Care Act’s many deleterious effects could be rolled back and a more inclusive and flexible health-care reform could be passed. Regulatory reform could provide more opportunity for workers while also improving business vitality. As president, Donald Trump would be able to put a halt to the administrative culture war launched by the federal bureaucracy during the Obama years. Trump’s judicial appointments could also shore up his support with the traditional Republican base and with voters who view governing through judicial fiats as another embodiment of unaccountable political power.

In the aftermath of 2012, it seemed clear that Republicans needed to improve their standing with the working class.  Trump's election may be a sign that they have made some headway in that goal.  But it will take success in governing to solidify and to expand that coalition.