Public Opinion Watch


Ruy Teixeira
Ruy Teixeira

(covering polls and related articles from the week of May 16–22, 2005)

In this edition of Public Opinion Watch:

Like Ouch, Man
• Hunting for EVs
• It’s a White Protestant Thing, You Wouldn’t Understand

Like Ouch, Man

If Maynard G. Krebs, beatnik extraordinaire, worked down at RNC headquarters, that’s what he’d likely be saying about the latest round of public polls.

Newly released data from the latest Pew Research Center poll include the following dreadful approval ratings for Bush: 43 percent approval/50 percent disapproval overall; 42 percent/43 percent on the environment; 38 percent/46 percent on foreign policy; 37 percent/56 percent on the Iraq situation; 35 percent/57 percent on the economy; 31 percent/49 percent on energy policy; and 29 percent/56 percent on Social Security.

The Pew analysis of the poll notes that the biggest factors (based on a regression model) driving Bush’s poor overall approval rating are the public’s negative views of his handling of the economy and of the Iraq situation.

The Pew poll also includes a series of questions asking respondents whether the country is making progress, losing ground, or staying about the same on a series of important issues. The worst result was on the budget deficit, where 65 percent say we’re losing ground and just 6 percent think we’re making progress. That’s followed by Social Security finances (63 percent/6 percent), how the health care system is working (62 percent/9 percent), Medicare finances (56 percent/5 percent), availability of good-paying jobs (55 percent/15 percent), illegal immigration (52 percent/11 percent) and the quality of public education (50 percent/20 percent). On the health care system, going back to 1994, and the budget deficit, going back to 1989, these are the most negative assessments ever. And on job availability, only an early 1994 reading is more negative than the public’s assessment today.

Speaking of job availability and the economy, the latest ARG poll indicates extraordinarily high levels of economic pessimism. Bush’s economic approval rating in the poll, 35 percent/57 percent, closely matches Pew’s rating (as does Bush’s overall approval rating at 43 percent/51 percent). And just 19 percent in the poll say the national economy is getting better, compared to 59 percent who say it is getting worse. Moreover, only 21 percent expect the economy to be better in a year, compared to 51 percent who say it will be worse.

In terms of their household financial situation, a mere 9 percent report that their financial situation is getting better, while 61 percent say it is getting worse. And expectations for a year from now are only slightly more positive: 23 percent say their finances should be better, while 50 percent expect them to be worse.

The latest NBC/Wall Street Journal poll has right direction/wrong track at 35 percent/52 percent and indicates a number of ways in which Bush and his administration are seriously out of step with the American public= Just 35 percent say Bush has the same priorities for the country as they do, compared to 57 percent who say his priorities are different. By 49 percent to 12 percent, the public says that Bush and his administration are placing too much, rather than too little, emphasis on Iraq and, by 30 percent to 27 percent, they feel the same way about “issues related to moral values.” On the other hand, they feel very strongly that too little (65 percent) rather than too much (1 percent) emphasis is being placed on jobs and the economy and express similar sentiments about health care (75 percent/3 percent), education (57 percent/8 percent) and gas prices (64 percent/9 percent).

Consistent with this overwhelming sense that the Bush administration is putting too little emphasis on jobs and the economy, the public finds Bush administration economic policy falling short in almost every area of the economy (the one exception is on keeping interest rates low). Bush administration policies receive their worst ratings on keeping manufacturing jobs in the country (69 percent not working well versus 10 percent working well) followed by dealing with the price of gas (67 percent/11 percent), managing the federal budget (65 percent/15 percent), keeping white collar jobs in the country (48 percent/23 percent), expanding the number of new jobs (48 percent/24 percent), controlling inflation (43 percent/28 percent), improving the overall economy (39 percent/30 percent), encouraging retirement savings (38 percent/32 percent) and keeping taxes low (43 percent/34 percent).

The public also expresses lop-sided support for Congress holding hearings on gas prices (66 percent support/13 percent oppose) and for Congress investigating Tom DeLay’s relationships with lobbyists (52 percent/12 percent). And the public continues to think, by a wide margin, that is a bad idea (56 percent), rather than good idea (36 percent), to change the Social Security system to allow workers to invest their Social Security contributions in the stock market. Moreover, those who believe private accounts are a bad idea are quite unlikely to change their minds (62 percent say their position is firm), while those who believe these accounts are a good idea are quite open to shifting their position (62 percent say they’re open to changing their minds).

Things just seem to be going from bad to worse for the Bush administration. Or, as Mr. Krebs might put it: Like ouch, man—like double ouch.

Hunting for EVs

Demographer William Frey recently released an interesting study for the Brookings Institution, “The Electoral College Moves to the Sunbelt,” analyzing likely changes in the distribution of electoral votes (EVs) between now and 2030. Based on Census Bureau population projections, he expects the following to happen:

“All else being equal, in 2030 the red-blue Electoral College vote would come to 303 to 235 (compared with 286 to 252 last November). . . . [T]his change is largely due to Snow Belt to Sun Belt demographic shifts.”

The states that gain the most EVs in Frey’s analysis are Florida (+9), Texas (+8), and Arizona (+5), all red states in the 2004 election.

Ron Brownstein’s article based on the Frey study, “Democrats Covet the West, but Can’t Keep Losing the South,” looks at it from a regional angle and correctly observes:

“In 2000 and 2004, Bush won all 11 states of the old Confederacy, plus Oklahoma and Kentucky. In those two elections it netted him 168 electoral college votes. That meant Democrats had to win about 73% of the remaining votes to secure a majority—a hurdle they found a little too high each time.

Frey projects that those 13 Southern states would cast 173 electoral college votes after 2010, and account for 186 after the 2030 census. If Republicans can still sweep the South at that point, Democrats would need to win a daunting 77% of the remaining votes to construct a majority.

Victories in the West might temporarily help Democrats offset the South’s rising influence. But it doesn’t seem a long-term solution.”

I don’t dispute this. Democrats can’t afford to cede the entire south to the opposing team, especially given that its share of the nation’s EVs will be increasing. That just gives them too little margin for error in the rest of the country.

But I do think it’s instructive to look a little bit more closely at the Frey data and understand that, while cause for concern, the changes Frey analyzes are not quite as daunting as they might appear at first glance.

Essentially the Frey data say that the red–blue EV margin, under 2004 results, would expand from 34 to 68 over the twenty-six-year period from 2004 to 2030 (a substantial period of time, let’s not forget). But the swing of EVs needed to reverse the 2004 outcome is only about half of that in each year: from 18 today to 35 in 2030.

How daunting is a swing of 35 EVs, using the 2030 EV projections? Well, Ohio will still have 16 EVs in 2030, so combining that with New Mexico (5), Colorado (9) and Nevada (7), all within five points in 2004, gets you to a 2030 blue majority.

That doesn’t seem so daunting.

And Florida, which the Democrats won in 1996 and, arguably, 2000, and only lost by five points in 2004, will have 36 EVs in 2030, so turning that one single state is enough to produce a 2030 blue majority.

And then there’s Arizona, which, despite the Democrats’ ten-point loss there in 2004, still seems likely to become more contestable over time due to demographic and other trends. That’s 15 EVs in 2030. Turn those EVs plus Ohio’s and New Mexico’s and once again you’ve got a 2030 blue electoral majority.

So: Democrats should study the Frey data carefully. There is much to learn there. But they shouldn’t let it spook them. The increase in Sunbelt EVs is a trend a smart and energized Democratic party can easily overcome.

It’s the “smart and energized” that’s the hard part.

It’s a White Protestant Thing, You Wouldn’t Understand

What is the relationship between church attendance and party identification? The conventional wisdom is that those who attend church most frequently lean heavily Republican, while those who attend least frequently lean heavily Democratic. A new Gallup report, based on 30,000 interviews conducted during 2004, confirms this perception.

According to the report, a “macropartisanship” measure tapping the Democratic leanings of a group (defined here as the percentage of Democrats in a group divided by the percent of Democrats plus the percent of Republicans in that group) has a value of 40 among those who attend church once a week, 45 among those who attend almost every week, 54 among those who go once a month, 56 among those who seldom attend, and 61 among those who never attend. That indicates a pretty strong and uncomplicated relationship between church attendance and Democratic leanings.

But among important subgroups of the population this relationship is considerably more muddled. Among blacks, for example, the relationship is considerably weaker and more erratic, going from 88 to 92 to 94 to 95 and back to 94, as you go from highest to lowest attendance. And among white Catholics the relationship is also quite weak and even more erratic, going from 49 to 47 to 46 to 57 to 54, as you move from highest to lowest levels of church attendance.

Given this, what’s driving the strong relationship we see in the overall data on church attendance and partisanship? It’s all about white Protestants: at the highest level of church attendance, macropartisanship is 25, rising to 32, 41, 47, and finally 52 at the lowest level of church attendance.

So when Democrats worry about the relationship between religious observance and supporting the Republican Party, it appears they should focus most of that worry on white Protestants. Among other groups, it doesn’t seem to be that big a deal (though see Alan Abramowitz’s analysis of 2004 National Election Pool exit poll data for some alternative findings and analysis).

Ruy Teixeira is a joint fellow at the Center for American Progress and The Century Foundation.