Actually, Dukakis Still Would Have Won

by: Chris Bowers

Tue Jun 02, 2009 at 19:00

Regarding my claim that Dukakis would have won in 2008, Andrew Gelman writes:

From our analysis of the Current Population Survey post-election supplement, here are our estimates for voter turnout in 2008: 76.4% white, 11.9% black, 7.4% hispanic, 4.3% other, with the categories defined as mutually exclusive (for example, if you're white and hispanic, you count as "hispanic"). The exit polls say 74% white, 13% black, 9% hispanic, and 5% other (not adding to 100% because of rounding error), but I think CPS is more trustworthy.

Now we can take the Dukakis numbers and plug them into the 2008 turnout numbers, as long as we make some estimate for the votes of "other." I'll assume 55%, halfway between his performance among whites and among hispanics. (By comparison, we estimate from the Pew pre-election polls that Obama got 45% of the two-party vote among whites, 96% among blacks, 68% among hispanics, and 59% among others.)

Plugging in Dukakis's percentages by ethnic group and using the turnout numbers of 2008, we get a national adjusted Dukakis vote of .40*76.4% + .89*11.9% + .70*7.4% + .55*4.3% = 48.7%, which is better than the 46.1% he actually received but not quite enough to win.

My counter argument is that even if Andrew's numbers are correct (they do contradict the exit polls, but those polls have a margin of error), then while my 50%+1 claim was slightly off, Dukakis still would have won an electoral college victory and become President.

According to Dave Leip's Atlas of Presidential Elections, Barack Obama received 52.87% of the national popular vote, and 365 electoral votes. Had there been an even, nationwide shift of 4.17% away from that total in all 50 states, and even if all 4.17% of those voters had shifted to McCain, it would not have flipped enough electoral votes to put McCain over the top. Only Florida (27), Indiana (11), Nebraska-02 (1), North Carolina (15), Ohio (20), and Virginia (13) were decided by 8.34% or less, flipping 87 electoral votes to John McCain. However, that would have still left our temporally-relocated Dukakis with 278 electoral votes, due to narrow victories in Colorado (8.95% minus 8.34% = 0.61%), Iowa (9.35% minus 8.34% = 1.01%), and New Hampshire (9.61% minus 8.34% = 1.27%). Further, even with 2012 reapportionment, in 2012 the "Dukakis" states would still have, in all likelihood, either 270 or 271 electoral votes, and still enoguh to win.

The one place were I would quibble with Andrew's numbers (which are extremely useful, especially the more specific turnout estimates), would be with his estimate on voting trends for "Asians" and "others." He puts this vote halfway between whites and Latinos, but that is a low-end estimate. In 2006, the Asian and "other" vote was halfway between whites and Latinos, but in in 2004, Asians and others were within 2% of Latinos, and 14% of whites. In 2008, according even to the numbers that Andrew sites, Asians and "others" are within 9% of Latinos, but 14% of whites. So, while it is possible that the Asian and "other" vote could be halfway between Latinos and whites, it is more likely that the Asian and "other" vote would be closer to Latinos. This would up an estimate of Dukakis's performance among Asians and "others" to something more like 60% (twice as close to Latinos as whites), thus adding another 0.25% to his total. Given current trends, and given that at least 1% of the country voting for third-parties in ten of the last eleven elections, that would lead to a Dukakis popular vote vitory in 2012.

(Then again, given that Bill Clinton lost the Asian-American vote twice, it is likely that Dukakis lost as well. However, to push this counterfactual to a truly absurd level, 2008 Dukakis would have benefited from the Asian-American backlash against neoconservative foreign policy, just as current Democrats have done. But, if I start making arguments like that, then I am straying from my original point.)

No matter what the specific amounts are (somewhere between a net of 6% and 8%), roughly half of the Democratic electoral improvement since the dark days of 1988 has come from demographic change, rather than from either infrastructure / strategy / activist improvements, or from poor Republican governing performance. Further, this demographic change is actually more problematic for Republicans than the other two areas, because it requires changing the coalitions rather then developing better infrastructure or simply hoping that Democrats can't get the economy going again. It is an underlying problem Republicans face, and which requires them to break out of the ongoing Nixon-McGovern framework of American political coalitions. From now on, the McGoverns are just going to keep winning, even if they were to nominate another Michael Dukakis.

Update: Ugh--how embarrassing. I orginally credited Nate Silver for this response, not Andrew Gelman. As someone who was mistaken for Jerome Armstrong for about two years, I know how annoying that can be. My apologies.

Chris Bowers :: Actually, Dukakis Still Would Have Won

Tags: , , , , (All Tags)
Print Friendly View Send As Email

it wasn't nate. it was some guy named andrew. (4.00 / 5)

i'll take the position... (0.00 / 0)
that it doesn't matter nor will we ever know.

You've gotten me to agree with President Obama (0.00 / 0)
Time to put the past away and move on to relevant issues.

"It sounds wrong...
     ...but its right."

If it's all demographic change (0.00 / 0)
Then ideology doesn't matter that much and there is little electoral incentive for Democrats to become either more progressive or more centrist.

Things You Don't Talk About in Polite Company: Religion, Politics, the Occasional Intersection of Both

There are significant electoral incentives (0.00 / 0)
on immigration issues, a rather ideologically charged debate.  

[ Parent ]
Well (4.00 / 1)
I don't think that identity and ideology are really different things. Identity is ideology, just not in a left-right sort of way.

[ Parent ]
Either way, your main point stands (4.00 / 1)
The pool of voters in US elections has undergone significant demographic shifts that favor the Democratic party.

Statheads can be quite the literalists.  

This diary's topic and conclusion should be the catalyst for (0.00 / 0)
many more diaies regarding the complete implications contained within it.

It certainly does seem to confirm the book title:'The Emerging Democratic Majority' by Ruy Teixeira and John Judis

Whenever I read detailed analysis (4.00 / 3)
like this, I have two reactions:

1.  Admiration for the work that went into the analysis.
2.  The desire to point out that in any given moment we tend to underestimate, and not overestimate the amount of volatility in the electorate.  It is true that the underlying dynamics at work favor Democrats.  But it is also true that such advantages are often fleeting.  I think this is particular true given the economic crisis we are going through, and how that crisis will effect people's perception of free markets and the government.

Regardless if Dukakis would win or lose today... (0.00 / 0)
...the point is, it would be a very close call and not anywhere near the clear defeat he suffered in 1988. Remember, way back then he lost with a margin of 7.8% in the popular vote, and an abyss in the electoral college count, 111 vs. 426. Under the changed demographics today, which favor the Dems, even for somewhat weak candidate Dukakis, it would be a real head to head race. And that's the important point.

However, I don't think this should become a competition about having the last word. No doubt Nate, or Andrew at 538, will come up with new arguments why Dukakis would lose by 1% or 10 electors or something like that. But this is really missing the forest for the trees. You made your point, Chris, and it's a convincing one. A single percentage missing in the details doesn't reduce it's value at all.

What's important is that elections should become easier to win for the dems - if they don't totally screw up (which can't be ruled out, regarding historical precedence).

4.17% shift would have worked on Dukakis' map too. (0.00 / 0)
Based on Dave Leip's results, giving 4.17% to Dukakis and taking the same off Bush's total would have led to a 287-251 electoral college victory. Although reapportioning today would have made it a 269-269 map.

I'm not going to say that would have worked exactly - Dukakis wins Montana and South Dakota under this map, which are fairly lily-white - but the same general trends apply. Certainly 2008's demographics would have given Dukakis California and New Mexico, and probably Colorado, Illinois and Pennsylvania.

Interestingly, it's rather like the Obama map, but with less strength in the north-east, more in the north Plains and a wipe-out in the south.

If anybody's interested, here are the states Dukakis would have won, their electoral votes and his percentages:

Rhode Island (4) 59.81
Iowa (8) 58.88
Hawaii (4) 58.44
Massachusetts (13) 58.40
Minnesota (1) 57.08
West Virginia (6) 56.37
New York (36) 55.79
Wisconsin (11) 55.58
Oregon (7) 55.45
Washington (10) 54.22
Illinois (24) 52.77
Pennsylvania (25) 52.56
Maryland (10) 52.37
Missouri (11) 52.02
Vermont (3) 51.75
California (47) 51.73
New Mexico (5) 51.07
Connecticut (8) 51.04
South Dakota (3) 50.68
Montana (4) 50.37
Michigan (20) 49.84
Colorado (8) 49.45

Forgotten Countries - a foreign policy-focused blog

Dukakis would have won because of people who were 10 or younger in 1988 (0.00 / 0)
In addition to the racial/ethnic changes in the last 20 years, there's been a real turn around in D/R split among White voters born since 1978. If you look closely, Obama and Dukakis did almost exactly the same among Whites who were old enough to vote in both elections (38 y/o and up in '08). For Whites in 18-29 y/o demographic, Obama beat McCain 54-45. If you compare exit polls from 2006 and the numbers for 18-24 y/o in 2004, you'll see that those folks vote for all Dems at that rate.  

But... (0.00 / 0)
Would Dukakis have ever gotten the 2008 turnout? I'm not sure Edwards or Clinton or any other Democratic candidate would have. Would Dukakis have resonated as much with the under-30 crowd? Would he have offered a place for dissatisfied republicans to bail?

Moreover, would his vote have been distributed as well? I suspect any other candidate's relative strengths and weaknesses wouldn't hold steady.

Importantly, I suspect that Dukakis would have made an easier target for a lot of the divisive culture issues that never stuck to Obama.

In the end, campaigns are dynamic events. Yes, they are influenced by demographic trends, but I'm not sure that you can just plug Dukakis' percentages into the 2008 electorate without adjusting for how the 2008 turnout and composition would have looked after both a year and a half of Republican attacks on Dukakis instead of Obama, and a year and a half of Dukakis' pathetic responses.


Open Left Campaigns



Advanced Search

Powered by: SoapBlox