Open Left - partisan ID

Michael Lind Secedes From Reality: Mixing Up Race, Class, And Party ID In The South

Paul Rosenberg — Sat, 15 Aug 2009 16:30:00 GMT

In much of his writing Michael Lind generally makes a very good point: focusing on economic populism is both the smart thing, and the right thing for Democrats to do. Sometimes this leads him to surprisingly insightful critiques--as it did in parts of his article critiquing Obama's neoliberalism that I wrote about last weekend. But other times this leads him to neglect--or worse yet, severely misrepresent--the opposing forces that stand in the way. Last weekend, I criticized him for misrepresenting the role of racism during the New Deal, but it was only a minor point with respect to the main thrust of that argument. This week, Lind's mis-apprehension of racism emerged full-blown as the center-piece of a truly delusional attack on "liberals".

Before turning to the issue of how Lind mis-represents liberals in my next diary (a task I already introduced in my earlier diary, "Going After Michael Lind With Occam's Razor"), I want to clearly show how Lind mis-represents those he claims to champion, and how confused he is more generally about issues of race, class and regional identity. As I've noted in the past, the South really is different, and I'm down with Tom Schaller on his "Whistling Part Dixie" thesis--not that the South can go to hell, but that Dems should stop trying win back the South on terms set by White Southerners.

Here is how Lind's article concluded:

o be expected that people, black and white, who have been deprived of adequate education will be more likely than educated people to believe in nonsense like Birther conspiracy theories and AIDS conspiracy theories. And it is only to be expected that people, black and white, who have been frozen out of politics by oligarchic elites will turn to flamboyant populist tribunes as their leaders, including theatrical preachers like Pat Robertson and Jeremiah Wright, Al Sharpton and Jerry Falwell.
The traditional liberal solution to such alienation is economic reform, education and political empowerment. But reform is difficult and expensive. And it is much less fun than caricaturing entire ethnic or regional groups, particularly those whose members tend to have less money, less education and less power than those who lampoon them.

Lind's attempt to equate Sharpton and Wright with Robertson and Falwell is an insult I'll return to in a later diary, along with his attempt to equate the Birther conspiracy theory with AIDs conspiracy theories. What I want to focus on here is Lind's wholly unfounded assumption that it is poor, uneducated White Southerners who are the core demographic who buy into this nonsense. Of course, Lind isn't the only one who makes this mistake. But he is, after all, the one who pretends to be both holier-than-thou and more knowledgeable-than-thou. And so it must be forcefully recalled that (a) the Birther belief is concentrated among White Southern Republicans, and (b) White Southern Republicans skew wealthy, not poor. On the first point,
PPP's North Carolina poll (pdf) had Republicans saying Obama wasn't born in America by almost 2-1 (47-24, with 29% unsure) while Democrats saying he was born in America by over 6-1 (75-12 with 13% unsure). On the second point, the following table, based on ANES data, is particularly clear:

(It should be noted that the sub-sample for the top income group is particularly small, especially for this decade, and thus the data is noisier and less reliable.)

Ever since the 1950s, the wealthiest Southern Whites have been more Republican than the poorest Southern Whites--indeed, the entire top third was more Republican than the bottom third. The last two decades, the relationship has been perfectly monotonic--every income group is more Republican than the group just below it--and the 2000s were more markedly so than the 1990s. Whatever may have been true in the past, the picture that Lind has in his head is 100% the opposite of what the NES data tells us.

On the flip, more tables to put the trend among White Southerners into context.
Have the Democrats lost strength among poor whites in the South? Undeniably. Their 38.7% edge among the bottom 1/6th has been cut virtually in half to 19.7%. But that 19% drop is nothing compared to the 45% drop amongst the next-highest 1/6th in the income ladder, much less the 85%+ drop amongst all three categories above that: a 95%+ drop among the middle third, an 89%+ drop among those from the 68th to 95th percentile, and an 86% drop for those at the very top.

Indeed, the poorest 1/6 of Southern Whites was the least Democratic demographic in the 1950s, and it's the most Democratic demographic today. They were 19.5% more Republican than the White South as a whole, now they're 29.35 more Democratic.

Say what you will about their poverty and lack of education, but their grasp on reality is much stronger than their so-called "betters."

Thus, Lind's entire sub-premise--that mocking Birthers means dissing poor White Southerners--is simply mistaken. Lind loves to lambaste liberals for their supposed stereotypes of White Southerners, and how mistaken they are. But what does this say about Lind's stereotypes?

Comparative Trends: Whites Outside The South

Having just looked at how partisan ID has shifted by income in the White South sine 1950, it's instructive to compare this with shifts in the White non-South:

Note that the poorest Whites outside the South were not the most Democratic until the 1980s. From then on, the margin in party ID has been completely monotonic--every income level is more Republican than the one before in every decade. (The sharp drop in GOP edge among the wealthiest Whites this decade is almost certainly at least partially a result of a small sample size amplifying outlier data, but given the similar, but smaller drop in the next-wealthiest group, the drop itself is almost certainly real, just a bit exaggerated.) Despite the large drop in GOP margin amongst the wealthiest group, non-South Whites still tilt more toward the working class this decade than they did in the 50s and 60s. And their overall margin--5.7%--is virtually the same as in the 1980s--5.9%.

If we take the difference between the figures for non-Southern and Southern Whites, we get the following table:

Southern Whites were almost 50% more Democratic than Northern Whites in the 1950s, and remarkably they were still more than 7% more Democratic in the 1990s. But this decade, non-Southern Whites have become 15% more Democratic. Looking at the class breakdown, there was almost no class skew in the 1970s--Southern Whites were about 15% more Democratic than non-Southern Whites across the board, except for those in the 96th income percentile or higher. There, Southern Whites were about 35% more Democratic. This decade, they are about 35% more Republican, while the next two income groups are about 22% more Republican in the South, and the bottom two income groups are a mere 2% or so more Republican in the South.

Comparative Trends: White America As A Whole

Looking at White America as a whole, we see that it has gone from a 30% Democratic margin in the 1950s to a 10% margin in the 1980s, to almost dead even in the 1990s and 2000s. Having lost just 3.1% outside the South since the 1950s, and a mere 0.2% since the 1980s, it's clear that White Southern losses have dominated.

But it's equally clear that losses have been smallest among the lower incomes (aside from the anomaly of the top income group this decade--but not last.) From the 1950s to 2000s, Democrats have lost 6.6% in the lowest income group and 11.5% in the second-lowest, compared to 24% in the middle third, and 28.7% in the upper third below the top level.

Comparative Trends: America As A Whole

Looking at America as a whole, Democratic losses since the 1950s are far less severe:

Indeed, there's been almost no change at all in the lowest two income groups, while the Democratic margin in the middle third has been cut in half, and Democratic margin in the 68-95th percentile groups has vanished entirely. In short, the Democratic Party has become a demonstrably more working-class party in the 2000s than it was during the 1950s, just as it has become demonstrably more non-Southern.

Finally, the last chart, All Americans minus White Americansm, shows the growth in Democratic margins among minorities, from about 2% in the 1950s to 12% in the 1990s and 2000s:

Equally striking is the fact that the Democratic margin is virtually flat for the bottom three income levels in both the 1990s and 2000s. Indeed, the minority margin for Democrats above that level is actually less than the White margin for all Democrats as recently as the 1980s.

All the above goes to show that the Democratic Party today is very much a working class party, both for Whites and minorities, both North and South. Lind's assumption that Birthers = working class Southern Whites that the Democrats cannot afford to alienate is simply one more example of how deeply distorted, and frequently mistaken America's views of race, class and national identity are.

One More Thing...

As a sort of cross-check on all the above, here's some data from the General Social Survey, which breaks down ideology and party ID in the White South, the rest of country, and the country as a whole. There's no economic breakdown here, but there is a big-picture take-away that strongly reinforces the message of Southern White Conservatives as politically anomalous, rather than representative of America as a whole:

In the White South, the percentage of conservative Republicans doubles from the first period to the third, from 11.8% to 23.6%, and the growth is continuous. They go from just slightly more than the number of liberal Democrats to 2 1/2 times the number of them.

Outside the White South, things look very different, particularly moving from the second to the third time period:

Conservative Republicans actually decreased in numbers, as did liberal and moderate Republicans, too. By contrast, moderate Democrats increased roughly the same small amount that liberal Democrats decreased during this same period of time.

In short, virtually all of the continuing conservative shift seen nationwide from the second period to the third:

was a product of the White South alone.

They are a poltical anomaly in American politics. A sizable one, no doubt. But they are in no wise typical of all Americans. Only in their own minds. And Lind, unfortunately, appears to share their narcissistic myopia.

Gallup Shows Broad GOP Losses In Almost All Demographics

Paul Rosenberg — Tue, 19 May 2009 16:00:00 GMT

Sub-Group Shifts Run Counter To Obama Strategy

Frequent church-goers were the only demographic subgroup to show no decline in GOP allegiance from 2001 to 2009, according to a new survey brief from Gallup. Declines among conservatives and those 65 and older were also minimal--the only bright spots reported for the GOP:

Democrats gained most from further consolidating support in their strongest demographic groups, rather than winning over Republican core groups, a shift that goes contrary to President Obama's repeated overtures to the GOP base:

Aside from education, for which the parties were basically at even strength in 2001, the Republicans' losses tend to be greater among groups that were not strong GOP supporters to begin with. These include self-identified liberals and moderates, church non-attenders, and lower-income and young adults. Thus, a big factor in the GOP's overall decline is the Democratic Party's consolidating its support among normally Democratically leaning groups.

Losses among Blacks and minorities in general were also small, but these groups already have low levels of GOP support. In fact, the 2% loss in total black support translated into a 17% loss of black Republicans.

The chart below (which adds a fourth column to the one released by Gallup) shows a loss of at least one in five GOP supporters among eight of the nine groups where the GOP lost eight or more percent support among the population at large. For example, the nine percent loss among moderates in general, from 37% to 28%, translated into a loss of 24% of GOP moderates--just shy of one in five. Combined with a 47% loss of GOP liberals and a 0% loss of GOP conservatives, this is yet another indication that the GOP is becoming more extreme as it shrinks:

Thus, although Gallup did not focus on this, one of the strongest take-aways from these results is that Obama's strategy of outreach to GOP stronghold groups runs directly counter to shifts in partisan trend: he is striving most visibly to reach out to those least inclined to listen, while offering a wide range of compromise, or at least soft-peddling with respect to issues that have have traditionally appealed to the groups where Democratic support has been growing.

The aggregate shift is represented in two charts, the first showing the shift in party support--members plus leaners:

The second shows the shift in support between party members and independents:

This simply reiterates previous reporting from Gallup as well as others, such as Pew, showing that GOP losses exceed Democratic gains. The combined picture is that roughly half of Democratic gains in support are from people who lean Democratic, but don't identify as party members--a further indication that Obama's relative de-emphasis on outreach to the Democratic base looks to be at odds with what would build the party at this time.

Gallup: Vanishing Red-State America--A Center-Right Nation No More

Paul Rosenberg — Wed, 28 Jan 2009 22:15:00 GMT

Gallup has just released "State of the States: Political Party Affiliation", the first in a four-part series to be released this week on Gallup.com, based on Gallup Poll Daily tracking data collected throughout 2008. And the results could not be clearer: GOP plurality states (including leaners) have been reduced to a mere handful: the Mormon mountain heartland, plus Alaska and Nebraska. That's it:

This contrasts sharply with what Gallup found as recently as 2002:

I'll work up a more extensive analysis for this weekend, but this shift is so dramatic, and so sharply at odds with the ongoing Beltway narrative, that I didn't think I should wait till then to toss this out for discussion. As a further indication of what this means, consider Gallup's list of the top 10 states for both parties:

As for the amount of data, Gallup reports:

In 2008, Gallup interviewed more than 350,000 U.S. adults as part of Gallup Poll Daily tracking. That includes interviews with 1,000 or more residents of every U.S. state except Wyoming (885) and North Dakota (953), as well as the District of Columbia (689). There were more than 15,000 interviews conducted with residents of California, New York, Pennsylvania, Texas, and Florida.

Of course, one can easily come up with plenty of caveats, as Nate Silver does at 538:

Now then for a couple of caveats. Firstly, Gallup's numbers consist of interviews with all adults -- not registered voters, and certainly not likely voters. Depending on the particular application that we're using this data for, that may be helpful or unhelpful. What this perhaps indicates, however, is that even after all the millions of new voters that the Democrats registered and brought to the polls in 2008, there are still probably some marginal gains to be had, particularly in areas like the deep South that the Obama campaign did not really concentrate in.
Secondly, these totals include "leaners" -- independents who lean toward one party or another, but don't identify themselves as such. This tends to increase the Democratic margin by a couple of points.
Thirdly and perhaps most importantly is a point that both Michael Barone and I have raised at various times: one consequence of the Democratic coalition being larger, particularly as it tends to include a miscellany of groups that don't always see eye-to-eye with one another (African-Americans, Hispanics, coastal liberals, union workers, young voters, etc.), is that it is more difficult to harness the entirety of that coalition in national elections. A Democratic presidential candidate from the North might have trouble appealing to voters in the South. A candidate from the South might have trouble appealing to voters in the North and West. A theoretic "generic Democrat" might have a chance at a rather large majority -- but a "generic Democrat" is an abstraction, and most real Democrats will offend the sensibilities of some or another region. In Barack Obama's case, these were voters in Appalachian and "Highlands" states like West Virginia and Kentucky, states that remain highly Democratic at the state level but which have not recently voted for Northern presidential candidates.

But, rather than get down to the micro-level and pick these apart (I promise a bit of that this weekend), just look at those maps like a board game, and ask yourself: which would you rather be: red or blue?

And then ask, given the map at the top of this post, which should Dems be playing right now, every single sound-bite of every single news cycle? Offense, or defense?

Center-Left Nation Update--Party ID Since 1972

Paul Rosenberg — Sun, 23 Nov 2008 22:18:29 GMT

The main objections to my diary "Center-Left Nation: Congress Since WWII" seemed to be two-fold: first that I was only using a single data source. Second, more significantly, that I was making an argument that ignored the racism of the Southern Dems. My answers were that (1) You only need one data source to disprove a thesis ("center-right nation"), and advance (not prove) another ("center-left nation"); (2) It was still true in the aggregate, as shown, for example, by the Congressional DW-Nominate scores.

I stand by those answers, but of course, they don't need to be the last word. And, indeed, they shouldn't be the last word. Hence, a look at party ID since 1972, from the General Social Survey (GSS). It should be noted that the figures here--which include partisan leaners--are not as strikingly Democratic as those from Pew, which is also a very reputable pollster, and I have no explanation for the discrepancy. But GSS figures are available going much further back.

(GSS polls not taken every year, see tables below for exact dates.)

So, the Democrats have lost their crushing 2-1 advantage enjoyed in the early 70s, but still have maintained an edge in every GSS poll since that time. And as for arguments about Southern racist Dems, well, that's what God made regional crosstabs for...
As it turns out, though there are clear regional differences, they are generally less striking than the similarities. And though we don't have figures from the '40s, '50s and '60s, the figures from 1972 on show that Democratic majorities were never solely dependent on Southern voters during this time, even when those voters were still overwhelmingly registered as Democrats.

First up, in the Northeast--New England plus the Mid-Atlantic states--Republicans briefly outnumbered Democrats in 1992, but the Democratic advantage has been up around 1970s levels for a while now.

Second, in the South, the GOP's new stronghold--from the Southern Atlantic seaboard to Texas and Oaklahoma--they came close to surpassing the Democrats in 1994, but lost ground after that, with a post-9/11 resurgence in 2002.

Third, in the Midwest--from Rustbelt Ohio to the Great Plains--the Republicans pulled ahead of the Democrats in the early 1990s, but have since fallen behind again.

Fourth, in the West--Mountain and Coastal--Republicans pulled ahead during Reagan's second term, but fell behind again under Clinton--an early profile of GOP gains, followed by losses, than in any other region.

Summary

Again, this is not presented as definitive proof that America is center-left, but it is proof that we are not center-right on the level of partisan politics, which is the level this debate is all about. The Versailles rhetoric is out of touch with reality, once again.

Tables

Party ID--Nationwide
Year	Dem	Ind	Rep
1972	60.9	10.6	28.5
1973	55.8	10.1	34.1
1974	59.0	10.5	30.4
1975	55.5	13.7	30.8
1976	55.9	16.3	27.9
1977	58.0	11.2	30.9
1978	52.9	14.4	32.7
1980	51.7	16.9	31.3
1982	60.2	11.7	28.1
1983	52.4	13.0	34.6
1984	51.4	11.5	37.0
1985	49.8	9.3	40.9
1986	49.7	13.2	37.0
1987	56.4	10.8	32.8
1988	48.8	12.6	38.6
1989	46.1	13.0	41.0
1990	45.4	11.0	43.4
1991	44.2	13.1	42.7
1993	46.5	12.5	41.2
1994	47.5	13.0	39.5
1996	45.2	16.4	38.4
1998	46.5	17.8	35.8
2000	44.0	20.7	35.3
2002	43.4	20.0	36.6
2004	44.4	16.2	39.4
2006	42.3	23.5	34.2

Party ID--Northeast
Year	Dem	Ind	Rep
1972	53.2	12.6	33.9
1973	58.8	9.6	31.6
1974	62.3	8.4	29.7
1975	55.4	13.0	31.6
1976	54.4	15.7	29.7
1977	53.9	11.8	34.5
1978	49.4	13.6	37.0
1980	45.9	19.2	34.9
1982	55.6	14.1	30.2
1983	51.0	15.5	33.5
1984	52.5	12.5	35.0
1985	47.9	11.4	40.7
1986	53.0	10.9	36.1
1987	49.7	12.5	37.8
1988	51.5	14.2	34.3
1989	43.6	14.7	41.7
1990	46.4	10.4	43.2
1991	37.6	15.2	47.2
1993	49.3	12.1	38.6
1994	50.6	11.8	37.7
1996	48.1	18.9	32.8
1998	50.3	20.8	28.9
2000	47.4	25.2	27.4
2002	45.6	24.7	29.6
2004	49.2	14.4	36.4
2006	48.3	23.0	28.7

Party ID--South
Year	Dem	Ind	Rep
1972	71.3	8.6	20.1
1973	54.8	9.4	35.7
1974	59.5	8.6	32.0
1975	60.6	12.4	27.1
1976	55.6	16.9	27.4
1977	59.4	11.7	28.9
1978	56.6	14.5	28.9
1980	57.1	14.8	28.3
1982	66.8	9.8	23.4
1983	56.8	13.2	30.0
1984	55.7	11.0	33.3
1985	56.3	7.2	36.7
1986	56.0	12.6	31.5
1987	60.8	9.9	29.2
1988	51.0	12.4	36.6
1989	47.7	12.5	39.8
1990	51.6	9.2	39.2
1991	50.6	10.5	38.8
1993	44.9	11.3	43.8
1994	44.5	13.8	41.7
1996	43.0	14.7	42.4
1998	45.0	16.2	38.6
2000	44.0	19.5	36.4
2002	41.4	18.3	40.2
2004	44.8	16.1	39.2
2006	41.7	24.8	33.5

Party ID--Midwest
Year	Dem	Ind	Rep
1972	54.6	12.5	32.9
1973	54.6	10.7	34.5
1974	58.1	13.2	28.7
1975	46.9	16.9	36.2
1976	53.9	16.4	29.7
1977	59.5	9.4	30.9
1978	47.3	17.0	35.4
1980	50.5	17.8	31.4
1982	59.2	11.3	29.3
1983	50.4	12.7	36.9
1984	52.4	12.9	34.7
1985	47.7	10.6	41.4
1986	48.3	14.7	37.3
1987	61.4	10.0	28.6
1988	48.5	16.0	35.5
1989	44.7	13.2	42.1
1990	38.4	14.9	47.0
1991	42.6	14.8	42.6
1993	41.1	15.0	44.0
1994	47.3	13.0	39.8
1996	45.9	16.7	37.4
1998	44.6	17.0	38.4
2000	38.8	22.3	38.8
2002	43.1	20.6	36.4
2004	41.8	17.6	40.8
2006	39.8	22.2	38.0

Party ID--West
Year	Dem	Ind	Rep
1972	63.7	7.9	28.5
1973	54.9	11.6	33.5
1974	56.2	12.4	31.4
1975	61.1	11.1	27.9
1976	61.6	16.1	22.7
1977	57.0	12.0	30.5
1978	60.2	10.6	29.2
1980	50.0	16.4	33.2
1982	51.5	13.7	34.4
1983	49.8	10.3	39.9
1984	41.4	8.8	49.8
1985	42.3	9.3	48.5
1986	37.9	15.1	47.0
1987	45.3	12.0	42.8
1988	42.6	6.6	50.9
1989	47.8	11.6	40.9
1990	43.7	9.5	46.4
1991	41.0	13.3	45.4
1993	53.1	11.3	35.6
1994	50.0	12.9	37.1
1996	45.4	16.2	38.2
1998	47.3	18.1	34.4
2000	46.3	16.4	37.5
2002	44.7	17.3	37.9
2004	43.1	16.4	40.5
2006	41.2	22.9	35.7

The Democratic Wave Since 2004--Rasmussen Tracking

Paul Rosenberg — Sun, 11 May 2008 19:46:57 GMT

First the Republicans began losing voters, then, only quite recently the Democrats began gaining them. Although the magnitude is less than that shown in a recent Pew Poll, the broad storyline is the same. Indeed, within the last month, the Rasmussen figures have suddenly jumped into the same range that Pew gave:

Table on the flip.
The movement since the primaries started is particularly striking--though the potential for it has been building for some time:

Partisan ID Tracking, 2004-Date: Rasmussen
	Republican	Democrat	Other	R - D	Quarterly	Dem/Rep
2004
January	34.60%	36.90%	28.50%	-2.30%		1.07
February	34.50%	38.40%	27.10%	-3.90%		1.11
March	34.30%	38.30%	27.40%	-4.00%	-3.40%	1.12
April	34.20%	37.00%	28.80%	-2.80%		1.08
May	34.70%	37.40%	27.90%	-2.60%		1.08
June	35.30%	37.30%	27.50%	-2.00%	-2.50%	1.06
July	35.50%	38.40%	26.10%	-3.00%		1.08
August	35.10%	37.70%	27.30%	-2.60%		1.07
September	37.30%	37.90%	24.80%	-0.60%	-2.10%	1.02
October	37.20%	38.70%	24.10%	-1.50%		1.04
November	37.10%	38.60%	24.30%	-1.60%		1.04
December	37.10%	38.80%	24.10%	-1.70%	-1.60%	1.05

2005
January	36.00%	37.00%	27.00%	-1.10%		1.03
February	36.20%	36.90%	26.80%	-0.70%		1.02
March	35.30%	38.10%	26.60%	-2.80%	-1.50%	1.08
April	34.40%	38.10%	27.50%	-3.70%		1.11
May	35.10%	37.00%	27.90%	-2.00%		1.05
June	35.60%	36.40%	27.90%	-0.80%	-2.20%	1.02
July	35.30%	35.90%	28.80%	-0.70%		1.02
August	34.90%	36.30%	28.70%	-1.40%		1.04
September	34.80%	36.80%	28.40%	-1.90%	-1.30%	1.06
October	34.10%	36.20%	29.60%	-2.10%		1.06
November	34.10%	37.90%	28.00%	-3.90%		1.11
December	33.90%	35.90%	30.20%	-1.90%	-2.60%	1.06

2006
January	34.50%	36.10%	29.50%	-1.60%		1.05
February	33.80%	36.50%	29.60%	-2.70%		1.08
March	34.00%	36.70%	29.30%	-2.70%	-2.40%	1.08
April	32.70%	36.40%	30.90%	-3.70%		1.11
May	33.60%	36.40%	30.00%	-2.80%		1.08
June	33.50%	37.00%	29.50%	-3.40%	-3.30%	1.1
July	32.80%	36.80%	30.40%	-3.90%		1.12
August	31.90%	37.30%	30.80%	-5.30%		1.17
September	32.30%	37.00%	30.70%	-4.80%	-4.70%	1.15
October	31.50%	37.70%	30.70%	-6.20%		1.2
November	31.40%	37.50%	31.20%	-6.10%		1.19
December	31.10%	38.00%	30.80%	-6.90%	-6.40%	1.22

2007
January	32.10%	37.50%	30.30%	-5.40%		1.17
February	31.70%	37.80%	30.50%	-6.10%		1.19
March	31.50%	37.20%	31.30%	-5.70%	-5.70%	1.18
April	31.00%	36.50%	32.40%	-5.50%		1.18
May	30.80%	36.30%	32.90%	-5.50%		1.18
June	32.00%	36.10%	31.90%	-4.10%	-5.00%	1.13
July	31.30%	35.90%	32.90%	-4.60%		1.15
August	32.50%	37.40%	30.10%	-4.90%		1.15
September	32.60%	37.20%	30.20%	-4.50%	-4.70%	1.14
October	32.70%	37.30%	30.00%	-4.50%		1.14
November	32.50%	37.40%	30.20%	-4.90%		1.15
December	34.20%	36.30%	29.50%	-2.10%	-3.80%	1.06

2008
January	33.10%	38.70%	28.20%	-5.60%		1.17
February	31.80%	41.50%	26.70%	-9.70%		1.31
March	32.10%	41.10%	26.80%	-9.10%	-8.10%	1.28
April	31.40%	41.40%	27.20%	-10.00%		1.32