The Republican party's strategy seems to be to Just Say No as loud as they can.
It doesn't seem to be working, of course, as we've seen in poll after poll. But let's look at the last 20 years of favorability ratings for the parties for some perspective:
Click to enlarge. Dotted lines show elections.
These are net favorability ratings - percent favorable minus percent unfavorable. While Republicans have been bouncing off of record lows for the past three years, in 2009 they've managed to sink well below their prior record. More on the flip.
Here's another way to look at it: the net favorability difference, defined here as Democratic net favorability minus Republican net favorability.
Click to enlarge. Dotted lines show elections.
Between 1985 and 2005, negative net approval numbers were rare. The long term trends show Democrats bouncing around, typically between 0 and +20, while Republicans are in a long-term decline, from +30 in the 1980s to –30 now. Unlike the Republicans' previous brief dalliance with –20 during the Clinton impeachment, these recent negative values have been sustained since the end of 2005.
Here's a guided tour:
- Public is not enamored with either party, leaving an opening for Ross Perot. Democratic favorability increases over the course of the campaign. Republicans stuck in the mud. Clinton wins.
- Democratic scandals in the House and the failure of health care reform. Republicans have their biggest advantages over Democrats in favorability in November, coincident with massive Republican gains in the House.
- After Newt Gingrinch leads Republicans in a face-off with Clinton resulting in the shutdown of the federal government, polls starting in April show Republican net favorability has dropped to zero. This rebounds quickly, but Democratic favorability increases as well.
- Impeachment throws Republican favorability into a nosedive, albeit only briefly. To complete the backfire, Democratic favorability rises during impeachment proceedings.
- Republican net favorability rises 30 points following 9/11. They milk it for all they can for the next two election cycles.
- Everybody loves a winner. Democratic net favorability jumps following the election of 2006. It falls back down shortly.
- Everybody loves a good primary. Favorability increases for both parties during their primary seasons. The Republican Convention produces a large but short-lived spike that almost brings Republicans up to Democratic favorability levels. And again, everybody loves a winner.
- The Democratic bounce is returning to earth, but the Party of No is sinking even faster. Especially among non-whites.
Clearly, scandals, campaigns, and other events can have large effects on favorability, but they don't last that long - weeks to months. This is certainly enough to affect elections - I'm looking at you, Mark Foley - but the event in question seems to be forgotten a year later. The question remains, are the recent low favorability numbers of Republicans simply the result of one scandal after another following so quickly that the numbers can never fully recover? Or do we have a new 'normal' where the Republican brand is so tarnished, even in good times favorability would still languish in the negative numbers? The continued decrease to even more negative numbers, representing only the die-hard 20-percenters who still approve of Bush, suggests the latter. And the pandering to the racist birther base suggests Republicans aren't seriously looking to change things any time soon.
To show how favorability numbers are relevant, here's the change in House seats compared to net favorability differences:
Click to enlarge.
There is a clear relationship. Of course, House elections aren't that simple, and the number of seats gained or lost has to do with the number and partisan distribution of seats that are held by a party as well, but as an indicator, net favorability difference is pretty good. It's a good thing to keep an eye on.
Fine Print: I normalized all data by subtracting a House Effect for each party for each polling institute. The House Effect was calculated by comparing to the average of CBS/NYT, NCB/WSJ, and Gallup. These are the 'norm' because they are the three polls with the most data. It doesn't mean it's the 'true' number, but it allows for us to look at trends. Normalizing the data is necessary because of the difference in timing and beginning of polling on this question, and large differences in House Effects. For instance, CBS/NYT has asked the question 76 times over 23 years, while Daily Kos has asked the question 28 times in just the past seven months. Also note the lines in the graphs above show a three-point running average, except where there are large gaps in time between data points prior to 1998.
Cross posted at Open Left.