terça-feira, 12 de maio de 2009

As primeiras contas para as eleições intercalares de 2010


... feitas por Sean Trende, no Real Clear Politics:

«I was a little bit surprised about two weeks ago when Stu Rothenberg declared that Republicans' chances of taking back the House were not small, not tiny, but zero. The usually cautious Rothenberg states that the idea of a GOP takeover is "lunacy [that] ought to be put to rest immediately." Charlie Cook isn't as categorical, but he writes that "the 1934 model [where the President's party picked up nine seats] probably represents Democrats' best-case scenario and 1982 [where the President's party lost 26 seats] their worst-case scenario. As of now, Obama's Democrats are heading down a track much closer to 1934's."

Let me state at the outset that I think a full GOP comeback in the House is very unlikely, as I explained here. The 25% chance that Intrade currently gives Republicans of winning in 2010 is Pollyannaish. But declaring that the Republicans' chances of taking the House back are absolutely DOA is grossly premature.

The claim that parties cannot bounce back from rough elections and claim the mantle of "change" in the subsequent election is unsupported by history. In 1946, Republicans gained 55 seats and took control of the House of Representatives after sixteen years of unified Democratic control of government. Two years later Democrats were back, gaining 75 seats and re-taking both Houses. After double-digit losses in back-to-back years in 1978 and 1980, including a drubbing in the 1980 elections that gave Republicans de facto unified control of the government, the nation eventually became dissatisfied with the Reagan Administration, and Democrats gained back 27 seats.

And in 1996, Democrats actually picked up enough Republican seats to take back the House. They were only foiled by the high number of Southern Democratic retirements, which allowed Republicans to take enough heavily Republican open seats from the Democrats to keep their losses to a minimum. The Republicans' task in 2010 is not so much to take on the mantle of change, as it is to convince the voters that the change they are receiving is not the change they voted for in 2008.

Similarly, the fact that this is not a "six year itch" election is of little consequence. The number of years a party has held the Presidency has no statistically significant correlation with seats lost. To put it in statistical terms, the r-square is about .02. This is to say nothing of the problems with the entire concept of the "six year itch," but that is a story for another article.

So how can we know what kind of year 2010 will be? For those who are interested in playing the scrying game, I think there are two major factors to watch over the next year: the President's approval rating and disposable income growth.

President Obama's popularity ratings are currently robust, ranging from the mid-50s to high-60s. However, they are also not that unusual for this point in a Presidency. As David Paul Kuhn has illustrated nicely, Obama's approval ratings at the end of his first 100 days were pretty much in the middle of the pack for modern Presidents. Some Presidents with approval ratings lower than Obama's have had successful first midterms (Bush II, Nixon), while some who had approval ratings higher than Obama's have had rough first midterms (Reagan, LBJ).

To be sure, everything Obama touches seems to turn to gold, and the GOP's assaults on him have fluctuated between ineffective and laughable. This is reminiscent of how Democrats banged their heads against a wall with the Bush Administration for two years after 9-11, amazed that he retained approval ratings of over 60%. A slow economic recovery, blossoming budget deficits, corporate scandals involving Bush associates, a flailing stock market, none of these seemed to have much of an impact on Bush's approval rating. Of course, eventually these things did, but for a very long time, President Bush - like President Obama today - seemed to be coated in Reaganesque teflon.

Democrats didn't really land a knockout blow on Bush until Katrina, but it is clear the Democrats' assaults on Bush's intelligence and competence took their toll over time, and put them in a position to maximize their gains when New Orleans flooded. The first time Democrats said that Bush lied about Iraq and mismanaged the war, no one paid attention. The hundredth time, people called them fools (who were consigning themselves to permanent minority status with their extremist views). The thousandth time, people stopped and thought about it. The ten thousandth time, they ended twelve years of GOP control of Congress. And it wasn't just Iraq and lies; Democrats threw the kitchen sink at President Bush over the course of six years. Some things worked, others didn't, but taken together it was enough.

The lesson for Republicans is that you eat a donkey the same way you eat an elephant: one bite at a time. While GOP attacks on Obama aren't gathering much traction now, it is never too early to start a meme. And the attacks will likely take their toll eventually. Almost every President's approval ratings dip below 50% at some point in his first term (the one exception is Eisenhower, and Obama, for all his charms, didn't plan D-Day). To put this in further perspective, if Obama is able to maintain his 100-days approval rating of 63% in Gallup, he would enter the midterm election with the fourth-highest approval rating of any President in his midterm election since Roosevelt (the highest was Bush in 2002, at about 68%).

So what could possibly threaten Obama's soaring approval ratings? As British Prime Minister Harold Macmillan is said to have responded when he was asked a similar question by a young reporter: "Events, dear boy. Events." What if a 9-11 style attack occurs in 2010? Presumably the "rally around the flag effect" would inure to the Democrats' electoral benefit. But what if ten 9-11 style attacks occur? At that point, the public may decide that Democrats are ineffective and unable to keep them safe. What if Iran gets a nuke? What if Pakistan collapses? What if Obama gets bin Laden? What if the economy roars ahead? What if Paul Krugman is correct, and we are not near the bottom of the recession? We just don't know enough at this point to rule out the possibility of a collapse in Obama's approval ratings, though again, we shouldn't interpret a mere possibility as a statement of fact.

Presidents have little control over events, and they have little control over an equally important factor: their timing. Harry Truman suffered from horrible midterm timing. His sub-40% approval ratings around the time of the 1946 midterms wouldn't be matched until 1950, just in time for the second midterms. Carter, on the other hand, had great midterm timing; his approval ratings in October/November 1978 were the highest recorded for the year prior to the election and for the year after; had the midterm elections been held a few months before or a few months after, the Republicans likely would have fared much better than they did. The first President Bush enjoyed high approval ratings for the 1990 midterms, but they were relatively poorly timed - from May of 1989 through October of 1991, this was the only time his approval rating dipped below 55%. Bush 41 also demonstrates how volatile approval ratings can be; from late October 1991 to early February 1992, Bush's approvals dropped 26 points.

If a sharp downturn in Obama's approval does come in time for the 2010 midterms, it is not difficult to imagine the GOP taking back the House. The average midterm loss for a President with sub-fifty percent approval ratings is 39 seats - just shy of what the GOP needs to take back the House. The median loss is 48 seats, which is enough to put the GOP back where it was in the Bush years.

The growth of disposable income further affects midterm election outcomes. When people have much more disposable income than they have in the past, as in 1962, 1970, 1978, or 1998, the President's party fares well. When they feel poorer, the President's party is punished; see 1946, 1958, 1974, 1982, and 1994. To draw something of a line in the sand: When disposable income has grown by less than a point, the President's party loses an average of 43 seats; the median loss is again 48.

What will 2010 look like here? Again, we get a giant "it depends." Disposable income growth is typically brisk in the year after a recession ends, so if analysts are correct in calling the end of the recession in the third quarter of this year, Team Obama has reason to be hopeful. Then again, the damage wrought by this recession is particularly wide-ranging, and disposable income growth after recent recessions has been atypically slow. And the recession may not, in fact, have ended by the third quarter.

Of course, there are other factors to consider, such as recruiting, fundraising, and retirements. We won't have a good sense of that until this time next year. One imagines that there is a pent-up supply of quality Republican challengers who have taken a pass on running in the past two cycles who can be enticed into making a go of it this year, but if things are looking good again for the Democrats this time around, Republican recruiting could be depressed. As Cook observes, fundraising is likely to be depressed for both parties, though it is still an open question whether one party or the other will suffer the brunt of the expected decline. And as Rothenberg has noted, the House GOP retirement disadvantage is unlikely to continue forever.

Obama's coattails - something we haven't seen at the Presidential level for quite some time - also present a problem for Democrats. Obama's candidacy drew numerous African American and college-age voters to the polls in 2008, and this likely proved critical for successful Democrats in close races such as AL-02 (Bobby Bright), VA-02 (Glen Nye) and VA-05 (Tom Perriello). These voters may still turn out in 2010 without Obama at the top of the ballot, and Obama may decide to expend enough political capital on 2010 to place himself virtually at the top of the ballot (assuming he has capital to spend at that point). But it may also be time to revive "surge and decline" theory.

It is also worth noting Cook's observation that "10 years ago, 68 House districts voted in the presidential race within 2 percentage points of the winning presidential candidate's national average. Now, just 38 districts fit that description." This is interesting, and it is relevant to a discussion on the abuses of gerrymandering. But I don't think it is relevant here. What is relevant is that the Democrats have about as many members in Republican-leaning districts today as they did in 1994. It is also worth observing that 1982, Republican losses were sharp, but still limited, because they only had 192 members of Congress to begin with; they still began 1983 with only 22 more members than they had in their rock-bottom year of 1974.

Finally a note on potential Democratic gains. Midterm gains in Congress are incredibly rare for a party - they have occurred only three times since the Civil War: 1934, 1998, and 2002 (in absolute terms it also happened in 1902, but relative to the growth of the House, Republicans lost seats). All three years involved Presidents with approval ratings above 65% (one imagines this was the case for FDR). And all three involved fairly robust income growth: 2.2% in 1998, 1.8% in 2002, and an astonishing 8.9% in 1934. If the economy is on track to produce income gains of 8.9% by fall of 2010, I will concede that the 1934 analogy is an apt one. And perhaps, given the absolute mess Obama inherited, voters will grade his economy on a curve. But if this is, in fact, a "U" or "L"-shaped recession, I have an even harder time envisioning Democratic gains than I do a Republican Congress.»

Sem comentários: