« Ruy Teixeira and Ken Mehlman Offer Opposing Views in Washington Post Analysis | Main | Bush's Hispanic Support Headed Downwards »

Comparing the Bush and Reagan Eras

Ron Brownstein had an interesting column, "GOP's Future Sits Precariously on Small Cushion of Victory" in the LA Times last Monday that put Bush's re-election victory in some much-needed historical context. He pointed out:

Measured as a share of the popular vote, Bush beat Kerry by just 2.9 percentage points [actually now down to 2.7 points--RT]: ....That's the smallest margin of victory for a reelected president since 1828.

The only previous incumbent who won a second term nearly so narrowly was Democrat Woodrow Wilson: In 1916, he beat Republican Charles E. Hughes by 3.1 percentage points. Apart from Truman in 1948 (whose winning margin was 4.5 percentage points), every other president elected to a second term since 1832 has at least doubled the margin that Bush had over Kerry.

In that 1916 election, Wilson won only 277 out of 531 electoral college votes. That makes Wilson the only reelected president in the past century who won with fewer electoral college votes than Bush's 286.

Measured another way, Bush won 53% of the 538 electoral college votes available this year. Of all the chief executives reelected since the 12th Amendment separated the vote for president and vice president a group that stretches back to Thomas Jefferson in 1804 only Wilson (at 52%) won a smaller share of the available electoral college votes.

But, even more interesting to me, since I've been pondering the comparison between the Bush era and the Reagan era, is the following point he makes about what a re-election victory has usually meant to the incumbent party and what typically has followed that reelection victory:

Throughout American history, the reelection of a president has usually been a high-water mark for the president's party [emphasis added]. In almost every case, the party that won reelection has lost ground in the next presidential election, both in the popular vote and in the electoral college.

The decline has been especially severe in the past half century. Since 1952 there have been six presidential elections immediately following a president's reelection. In those six races, the candidate from the incumbent's party has fallen short of the reelection numbers by an average of 207 electoral college votes and 8.4 percentage points in the popular vote.

Because his margin was so tight, Bush didn't leave the GOP with enough of a cushion to survive even a fraction of that erosion in four years. Even if the GOP in 2008 matches the smallest electoral college fall-off in the past half century the 99-vote decline between Reagan in 1984 and George H.W. Bush in 1988 that would still leave the party well short of a majority.

Very interesting stuff indeed. And it suggests that comparing the GOP's previous re-election victory and the current one is an exercise with more than academic implications.

Start with the obvious: Reagan got 58.8 percent of the popular vote in 1984, besting his Democratic opponent by 18.2 percentage points, compared to Bush's 2.7 point victory margin, and carried 98 percent of the electoral vote, compared to Bush's 53 percent. Indeed, if you put the two elections of the Reagan era together, we find Reagan averaging 54.8 percent of the popular vote with a 14 point victory margin and 94 percent of the electoral vote, compared to Bush's average of 49.4 of the popular vote, a 1.1 point victory margin and 52 percent of the electoral vote.

Quite a difference and, as Brownstein emphasizes, essentially no cushion against incumbent party third term slippage.

It's also fascinating to compare that 1984 GOP high water mark to the current one in terms of how the GOP is faring in different types of counties. Take, for example, those 100 fastest-growing counties (since 2000) where Bush did so well in 2004. I've pointed out the less-than-earthshaking nature of this trend elsewhere. But it's interesting to note that in those very same counties in 1984, Reagan did even better: he carried them by 36 points, compared to Bush's 25 point victory this year.

Or, if you prefer, take the 100 fastest-growing counties from the 1990's: Bush carried them by 27 points this year; Reagan carried them by 38 points in 1984.

Another interesting point of comparison is to look at large metropolitan areas. In the exurban or fringe counties of these areas, Bush beat Kerry by 24 points....but Reagan beat Mondale in these same counties by 29 points. So even in this area of particular strength for Bush, he he still lagged somewhat behind Reagan's 1984 performance.

And Bush lagged way, way behind Reagan in the most consequential part of large metro areas, their central counties. Reagan carried these counties by 11 points in 1984, while Bush lost them by an identical margin this year. That's a huge anti-Republican swing of 22 points in a group of counties that are much more consequential than exurbia to GOP electoral fortunes. In the 2004 election, these central counties still cast 43 percent of the overall vote, compared to just 5 percent for the exurban counties.

Yet another way to look at the Bush '04/Reagan '84 comparison is to compare Bush's strength in his best areas--rural and exurban counties combined--with Reagan in the same counties in 1984. That comparison shows that Bush carried these counties by a healthy 21 points (60-39) this year...but Reagan did even better in 1984, carrying them by a 25 point margin (62-37).

Moreover, despite the fact that exurban areas have been growing fairly rapidly, they start from a small enough base that their share of all US voters has increased only modestly over the last twenty years. In fact, once you combine these exurban areas with the rural areas, which have been declining slightly, the share of the US vote cast by the combined group of counties has held rock steady at 25 percent between 1984 and 2004.

In short, when you compare the Bush era to the Reagan era, even Bush's strongest areas don't look so strong and there's less real growth going on in his coalition than generally supposed. If this is a contemporary high water mark for the GOP--and there are good historical reasons for supposing it is--they could be in real trouble.


This is all fair and accurate (and mandate discussion is something I've done myself, not just in terms of re-elected incumbents but also war presidents [see http://www.samueljohnson.com/blog/archives/0411a.html#5a], so I quickly understood where you were coming from).

What wasn't clear to me from either the piece here or the original Brownstein article is if this is purely an academic exercise or something in response to actual comparisons going on. Aside from mandate claims, is anyone really comparing Bush to Reagan, or is is just the coincidence of Reagan having been the last reelected Republican president? (IE, is the Reagan contrast a straw man?)

There are certainly opportunities for the Democrats, however. GOP membership is in decline, according to http://www.harrisinteractive.com/harris_poll/index.asp?PID=444. There's a lot of jockeying for position already going on, with hard righters pressuring the moderates and so on... I'm in total agreement that the potential vulnerability of 3rd parties is high.

I think historical perspective is worth having. However, using it as historical wisdom to look ahead can lead to predictions that turn out (in hindsight) to be plain wrong.

In that past, a foundering job market, a worsening war overseas, rising deficits, etc. have undone past presidential re-election bids. Large numbers of undecided voters have broken for the challenger, and high turnout has favored Democrats. Despite all that, things turned out a bit differently this time around.

I appreciate comparisons to the past. It may be that those who don't study history are doomed to repeat it. But as they say in the stock market, past performance is no guarantee of future returns.

I severely doubt this analysis. There are several factors that didn't come into play with Reagan the first that springs to mind is that Reagan didn't have the chokehold on the media that the GOP now does.

I think that your thoughts need to take into account not only party loyalties, but the performance of George W. Bush. As one that follows current affairs, and pointed out by most media outlets during the campaign, we know that things did not go especially well for GWB and the US during his first term. Based on this, I think that a comparison with Reagan is misleading. I am unsure that so much of the US voting public aligns along party lines as your piece suggests. Given a more viable candidate in '04, I am sure that Bush would be out of office in January.
In surfing channels on TV this weekend I heard a part of a speech that Limbaugh gave a few weeks ago. He made fun of Kerry's camo outing saying that this was pandering in the most transparent form. I think he was right, the left is out of touch (as is the far-right) and if a more in-touch candidate was presented during the election I think that things may be quite different.

While this piece gave me a bit of a lift, I have to wonder if it is entirely fair to compare W to the hugely (though inexplicably) popular Reagan-- would it not be more fair to compare him to other two-termers like Eisenhower, Nixon, Clinton?

On one thing we need to agree, and that's what's expressed here: the fundamentally miniscule nature of the Bush victory margin, both in the electoral and popular vote. The press and GOP are overstating it, both for reasons of ideology, and in comparison to the non-margin of 2000 (as well as, I believe, the initially reported early-day exit poll numbers). But Democrats are also inflating the margin, because they look at this presidency they fear and loathe, and feel that finishing second in such a circumstance is a humiliation far beyond the numbers.

That the GOP/media slant is silly is obvious. But Dems need to look at this dispassionately and realize that their take is also overbaked. Bush had a less-than-triumphant term, but he also had a less-than-disastrous one. Whatever the over-the-term economic numbers were, the election year economy, no matter how you slant it, was NOT recessionary (had it been, Bush would have been defeated, period). Iraq was problematic at best, and will almost certainly grow far worse, but on Election Day, voters were split on it as success or failure. Compare this to the wide perception of Vietnam as failure by November 1968 and you start to see why the administration was not turned out.

I bought Ruy's book upon publication and embraced its thesis. I still -- despite three thisclosebutnotquitethere outcomes -- believe he sees the future more clearly than most pundits. But I also believe his theory (like Kevin Phillips' before it) is descriptive of a trend, not an iron-clad predictor of each quadrennial contest. I believe, a la Alan Lichtman's Keys to th Presidency, that each individual presidential election is decided circumstantially -- and Bush's circumstances were, like Truman in '48, not quite bad enough for outright rejection. By locking up his own party so early, and by winning a few seats in the mid-terms, Bush was able to enter 2004 with all four of Lichtman's political keys in his favor. This meant it would take a disastrous 3-for-9 in the remaining keys for Bush to be defeated -- a historically unlikely event (though Bush almost pulled it off, losing at least four, maybe five by my reckoning).

This, however, didn't lead to any easy re-election road for Bush. In fact, he damn near achieved Grover Cleveland II-ism -- a shift of a bare 65,000 votes in Ohio would have left him a popular vote winner and EC loser. Consider if Kerry had been the one with the 3% margin: he'd had almost certainly have had a College score of 310-320. That's the emerging majority, and I think Dems are still (presidentially) in better shape than the GOP right now.

Alot of us had hoped that 2000 was the 1976 of its era -- the last gasp of a once-strong-but-fading presidential coalition, one that barely holds on, at that. Such turned out not to be the case this year, but to assert we're seeing the opposite -- a surge toward the old coalition -- is to me a ludicrous claim. It's just as possible that 2004 will turn out to be the '76 of the cycle -- that, minus the party unity over Bush, the GOP will engage in fratricide, while the problems of the economy and Iraq become apparent even to the most casual observers. We can never predict the circumstances of an election four years off, but I think the chances of a negative incumbent environment in 2008 are quite strong. Given the substantial strength of the Democratic base vote (EVs of at least the mid-200s for four straight elections), the party's in an excellent position to reap major benefits from that.

Historical comparisons are meaningless as predictors of the future.

Instead of proposing that the Republican Party is at a highwater mark, one should more aptly say they'll steal every future election until they're exposed and punished.

Not only do they have a choke hold on the media, they also have a choke hold on the electronic voting machines. It seems to me that they can't possibly lose an important election they don't want to.

I have been away for a week and have just caught up on the posts for the past week. As always there was certainly a lot of interesting analysis. If we are to keep a perspective on this past election, we must remember that the Republicans planned for a close election. They had a mult-point plan to win the ppuluar vote and the electoral vote. We should also keep in mind some data from a couple of weeks ago. That 50% of Bush's popular vote margin came from Texas, Mississippi, Georgia, and Alabama. Also that the vote was not as high in California precincts in 2004 as in 2000. It is obvious there was a plan on the part of the Republicans to turn out the vote in their non-battleground states. The comments by Repulbicans and media on the lack of importance to vote in non-battleground states, along with lack of attention by the Democratic candidates led to lower voter turnout in strong Democratic states. Therefore,
a 3 million popular vote margin is not the true sentiment of American voters.

The key to winning elections is to frame the message and work on winning state legislative and council of state elections over the next couple of years. Simply being a majority party will not win elections. Nixon and Reagan both won by landslides when the Democrats were the majority party.

I think it's obvious to all but the GOP triumphalists that Bush had a narrow win this year. The comment that a re-election is a high-water mark for a party is interesting as far as it goes, but I tend to look more at the popular vote trends than the electoral vote trends to gauge a party's broad level of support. The electoral margin is heavily affected by regional voting patterns, which change a lot over time.

In the 70s and 80s, voting patterns were much more homogeneous than they are now: this allowed Reagan to pile up 44- and 49-state wins with popular vote margins of 10% and 18%. With a similar popular vote sprad, electoral vote margins today would be much smaller (no Republican could win Massachusetts, and no Democrat Mississippi). Even by 1996 things had shifted enough to where Clinton's 9-point popular vote win (similar margin to Reagan in 1980) only gave him 379 EV's to Reagan's 489. The obvious corollary is that the chance of the popular and electoral votes misaligning is much greater nowadays.

Look at the 1976 map to see must how much things have changed. The big swing states in 1976 were Ohio and ... New York! California and Illinois went Republican. So did Washington and New Jersey. It's just not possible to compare electoral margins between then and now.

Interesting stuff Ruy. I've often thought W was the last dance... and kind of wished we'd just stuck Dean at him to really make a blow-out prom, but what worries me are the following: (a) democrats don't seem to be able to build electoral majorities with a specific platform, they are often just the choice when people have had enough of republican rule, (b) the distinction between exurban and suburban is starting to become one of fancy, in my mind, and i'm beginning to see very real inroads made in what we call "suburbs" or even "urban suburbs" by the GOP, primarily through the church, (c) i don't see how this sort of analysis points us in the right direction for '06, it's decidedly national in prospective, can we start putting your excellent skills to use in targeting districts, crafting platforms and testing messages?

we have what we believe is a building block for a platform for '06 and beyond. our own independent tests have confirmed both the positive economic impact as well as anecdotal evidence of swing-voter empathy. feel free to take a look if you like.


The environment has been favorable for Democrats to win the presidency since 1992. This was supposed to be an era dominated by Democratic presidents and Republican control of Congress. The problem is that we keep nominating duds.

"The decline has been especially severe in the past half century. Since 1952 there have been six presidential elections immediately following a president's reelection. In those six races, the candidate from the incumbent's party has fallen short of the reelection numbers by an average of 207 electoral college votes and 8.4 percentage points in the popular vote."

This is all well-and-good, but I think this is a case of a correlation that does not suggest causation. My guess is that the drop-off has more to do with two extraneous facts that may or may not be present in 2008. First, after an incumbent is re-elected, there is usually a vice-president running in the subsequent election. Vice-presidents are notoriously difficult to elect, for reasons spelled out in great detail elsewhere. There will likely not be a vice-President running in 2008 (save perhaps Gore). Second, just like with any race, House or Senate, there are usually advantages to incumbency that disappear in an open race. This time around, it was a difficult re-election for Bush. It's by no means likely that this time a non-incumbent will necessarily start out from a more difficult point than did Bush this time around. Overall its an interesting tidbit, but I'd put it in a category with "the president's party always loses in midterms" and other "Election Laws" that emphasize the correlation and ignore the causation.

A footnote in the interest of historical accuracy: in 1828 the incumbent president, John Quincy Adams, LOST to Andrew Jackson.

I agree w/you Sean, this is simply not an analysis unless you can go beyond formal comparisons and give some meaningful interpretation of events. Anyone can make correlations, from the astrological to the meteorological, and it hardly provides some law of history. Perhaps every time an incumbent was defeated Mars was in the house of aquarius and it was raining in mongolia. If these conditions hold again should the Democrats start breaking out the champagne?

If we're determined to make hasty and near-empty generalizations, however, my guess is that if parties are often defeated after two terms in the presidency it's simply because both major parties have very similar policies and the public simply gives the other guy his turn on the throne.


It's better to be an incumbent President in almost all circumstances, than not. The exceptions are pretty rare (you've clearly lost a big war - LBJ - or the economy has gone into depression - Hoover). Other than that, it's better to be an incumbent than not. Non-incumbents always start from a more difficult position than incumbents. And that was the case in the 2004 elections as well.

The problem that the Republicans face in 2008 is harder than they would like to admit. Jeb is the obvious choice, but has repeatedly indicated that he would not run in 2008. Cheney is clearly unelectable. Most of their Washington-level leadership is also clearly unelectable (Delay, Hastert, Hyde, etc.). McCain is too old. Frist is not popular and not charismatic. There is nobody else really charismatic in the Senate.

Their only choice is the governors. Here, there isn't much fertile ground as it might seem. The only nationally recognized figures are Schwarzenegger (invalid), Pataki (not charismatic), Perry (homosexual and therefore verbotten to the Republicans), Romney (from Mass. and therefore verbotten to the Republicans AND not charismatic). So you've got possibilities like Owens, Johanns, Kempthorne, and a less than handful of others. Most of the current Republican governors have only been elected to office in 2002 or 2003. None of these guys' runs are going to be unchallenged.

So, unless Jeb steps in, their primary is going to be a repeat of the 1995-1996 process, which led to Bob Dole. A lot messier, and a lot less unified party than would be in the case of a Bush.

It's fine to look at the presidential level and see how unimpressive Bush's reelection was. But let's not overlook congress.

Most reelections are personal vindications, but not partisan. Bush's appears to be the opposite, and this is alarming.

Consider the following, for reelected incumbents since 1956. Below is the year, followed by the percentage change in the popular vote, then the change in electoral votes, and finally the change in House seats (measured by the result in the reelection year minus the result from the election four years earlier):

1956, +2.3, +15, -20
1972, +17.3, +219, 0
1984, +8.1, +36, -10
1996, +6.2, +9, -51
2004, +3, +15, +12 (and +4 in senate)

Only one positive in there in the House column.

And the senate could get worse in 2006. The two Nelsons (FL and NE) could be vulnerable. Plus, look at these victory margins in 2000. These three ran behind Gore's support in their state and are in states where Bush gained on the Democrat in 2004 relative to 2000:

Stabenow (won 49.5-47.9)
Corzine (50.1-47.1) Already announced run for gov '05
Cantwell (48.7-48.6)

Maybe Hillary should be in that list, too. And Dayton (5 point margin, but less than 49%).

Can we realistically hope for anything better than just having the presidential veto on our side after 2008?