Top Videos
Related Topics
election 2006
2008 Polls NationalIowaNew HampshireGeneral Election
GOP | DemGOP | DemGOP | DemHead-to-Head

Send to a Friend | Print Article

What Does the Money Tell Us About Tuesday?

By Jay Cost

Just last week, all candidates for federal office were required to submit their final pre-election financial disclosures to the Federal Elections Commission. The results were illuminating - a nice break from what I have taken to be the real darkness that surrounds the House elections.

I have felt that the conventional wisdom on the House has looked an awful lot like a circular track of footnotes. People who study for a living know what I am talking about. One author makes a claim about some such thing, and, in a footnote, cites another author. Because you are interested in the claim, you go look up that author, who then cites a third author. The third cites a fourth, and the fourth cites...a prior edition of the first! That is how the post-Foley environment has felt to me. The pundits are all just quoting each other - nobody is working from any hard data and/or well-tested theories.

Well - the FEC is nothing if not a provider of hard data! But we have to read it in light of good theory. Barron's tried to interpret the FEC numbers last week and failed because they did a straight head-to-head comparison of candidate dollars per race. The problem with that, as I argued, is that victorious challengers almost never spend as much as incumbents. I view money as a necessary, but insufficient, criterion of electoral success. So - money can give us a sense, not of which way the close races are leaning, but rather of the set of close races. I made this point back in August when I crunched the numbers on the Quarter II financial reports. I am going to do the same now.

With a bit of tweaking. In the summer, I compared candidates this year to candidates in 2004 in light of both cash on hand and total receipts. My interest was in whether or not the 2006 candidates were on track to raise what the 2004 candidates eventually raised. I also included a factor to account for debt. This time around, I am just going to look at total receipts - to see who is on track to raise what was raised in 2004. Cash on hand becomes very tricky this late in the season because candidates have paid for advertisements that have not yet run, or paid for other services that have not yet been delivered. Just as just made me pay for Music from Big Pink before they have actually delivered it, so could candidates be paying in advance for products yet to be delivered. FL 22's Ron Klein, a qualified, professional candidate if there ever was one, has less than $30,000 in the bank right now. Is that because he wasted all of his money? Unlikely. It is probably because he has paid for advertisements, mailings, and other products/services that have net yet been delivered. So, we cannot "fault" him for that.

I have also eliminated the debt factor. Whereas cash on hand was certainly an appropriate marker for the summer - as it was a sign that candidates, in the pre-campaign season, were not wasting money on that extra box of Krispy Kremes for their volunteers - I have now decided that debt is probably not a good indicator to include at all. The debt factor's inclusion was meant to account for candidates who were genuinely hopeless and were simply self-funding. They still present a problem, but - my intention here is to look at dollars as capacity to purchase campaign products, not dollars as a sign of viability. This means that I should not have factored debt into the summer analysis. So - take this exclusion as a sign that, while the money race illuminates some, it does not illuminate all. It is a necessary but insufficient condition.

Another change - I have decided to expand the number of ways I count the minimum floor for money. In the summer, I looked at the 2004 results - and counted the number of candidates on track to raise the minimum amount required to take a House seat, and the number on track this year to raise the average amount raised by those who would eventually raise enough to take a House seat in 2004. Now - to "flesh" things out a big - I am going to add two additional models.

The first additional model is more stringent upon challengers. It requires this year that the minimum amount for a viable challenge is 2004's average amount adjusted for CPI inflation. That means that, to be a viable challenger of an incumbent, you'd have to be on track to raise about $2 million. To be a viable challenger in an open seat, you'd have to be on track to raise about $2.1 million.

The second model is much less stringent. It looks back to the last election cycle where many seats switched - 1996 - and simply adjusts for inflation the median amount to take down a GOP challenger. This was a year when the Democrats took down 17 vulnerable Republican incumbents and offered a strong "correction" to the 1994 Republican surge. Under this model, the inflation-adjusted minimum amount to be a viable Democrat is $1.2 million dollars. The intuition driving this model is that Democrats have such salient lines of attack this year that less this year is just as good as more in 2004 - and, effectively speaking, there has been little-to-no inflation in campaign costs (for Democrats) beyond the CPI because of the salience of their line of attack. Meanwhile, because of the negative political environment, the minimum amount to be a viable Republican is what it is in the most stringent model, $2.2 million.

I do not think either of these models is very accurate. I view the first mostly as a floor and the latter mostly as a ceiling.

That gives us the following number of party seats where the other side is offering a viable challenge.

Model 1 - Average 2004, Corrected for CPI Inflation: 23 Republican seats, 6 Democratic seats.
Model 2 - Average 2004, Not Corrected for CPI Inflation: 27 Republican seats, 6 Democratic seats.
Model 3 - Minimum 2004, Not Corrected for CPI Inflation: 39 Republican seats, 7 Democratic seats.
Model 4 - Median 1996, Corrected for CPI Inflation: 48 Republican seats, 6 Democratic seats.

This is quite a wide bit of variation, but it does give us some additional purchase on the current battlefield. The (on-the-record) talk from Republican strategists points toward the figures from Model 1. If this was the true model, it would indeed be unlikely for the House to tip, and the Republicans should be confident. Model 4 is the talk both of Democrats and many in the media. Rahm Emanuel boasts that 50 seats on the table, which is 2 seats more than the ceiling. Charlie Cook sees 54 seats as competitive, 6 more than this ceiling. So - even with a range of 25 Republican seats, these 4 models still offer us some kind of guidance through the noise.

I think we can push the analysis a bit further than this, though - and develop a more specific estimate than "more than 23, less than 48." The House party committees - the NRCC and the DCCC - are allowed to spend unlimited quantities of cash on races, provided that the expenditures are "independent" of candidate allocations. They are also required to report those numbers to the FEC. National Journal's House Race Hotline has been keeping track of the races in which the parties are spending. I have gathered their most recent list, from Tuesday, and updated it with Wednesday's figures.

Parties can, I would say, spend for 3 distinct reasons. First, they believe that they will win the race, but want to put down some cash to guarantee it, or shore it up. Think of the money the NRCC has put into WY-AL. They spend money in this instance because they are risk averse. Second, they believe that they may stand a chance in a race, and want to put down some cash to test their position. Think of the money the RNC is pumping into New Jersey. They spend money in this instance because they have imperfect information and they wish to procure some data. They might also spend money to throw the other party off the scent, to project a sense of strength, etc. Third, they believe that the race could go in either direction, and they put the cash down to swing things in their direction. Think of the money both sides are pouring into CT 04. They spend money in this instance to increase the chance of winning.

It seems reasonable, further, to conclude that seats where only the party in power is spending its money most likely fall into the first category - i.e. they are just shoring things up. It also seems reasonable to conclude that seats where only the party out of power is spending its money most likely fall into the second category - i.e. they are just testing the waters. It also seems reasonable to conclude that seats where both parties are spending money most likely fall into the third category. However, this is not necessarily the case. Both parties might be spending money for the first and the second reasons. The in-party spending to shore up. The out party spending to test the waters. The tip-off would be the position of the out-party candidate. If they are well-financed, then the parties are probably in the midst of genuine competition. If they are not well-financed, then it probably is a combination of the shoring up and testing the waters/spinning the media.

Accordingly, a cross-tabulation of Model 4, the broadest model I offer, with DCCC and NRCC spending would be illuminating. It seems to me that the districts where the parties are both spending their money, and the out-party candidate has raised at least $1.2 million, are likely to be part of the true battlefield - i.e. the races where neither party has a decisive advantage. Conversely, the places where either the parties are not engaged, and/or the out-party candidates have not raised enough, are not likely to be part of the battlefield. When we cross-tabulate the three sets of data (candidate receipts, NRCC spending, DCCC spending), we can thus get a good estimate of the size of the true battlefield.

When we do that, we get an interesting result. The candidates have raised the minimum amount necessary and both parties are spending in 32 Republican-held districts and 3 Democrat-held districts.

By and large, this method produces a set of seats that does a superlative job of covering all of the generally-accepted vulnerable seats, illuminating some seats as being vulnerable that not everybody thinks are vulnerable, and cutting through the party-generated spin about which seats are vulnerable and which are not.

The seats that this method includes are: AZ 05, CA 11, CO 07, CT 02, CT 04, CT 05, GA 08, GA 12, FL 16, FL 22, IL 06, IN 08, IN 02, IN 08, IN 09, IA 01, IA 03, KY 03, KY 04, MN 06, NC 11, NV 02, NV 03, NH 02, NM 01, NY 20, NY 24, OH 01, OH 15, OH 18, PA 06, PA 07, PA 08, PA 10, VA 02, WA 08, WI 08.

Oft-mentioned seats that are not included (and remember a lack of inclusion means that either (a) the NRCC is not spending there, or (b) the DCCC is not spending there, or (c) the out-party candidate is not on track to raise $1.2 million) are: AZ 01, AZ 08, CA 04, CA 50, CO 04, CO 05, FL 13, ID 01, KS 02, MN 01, NJ 07, NY 19, NY 25, NY 26, NY 29, PA 04, VT AL, WA 05, WY AL.

The only three races that are excluded from this list that certainly should be included, i.e. the known Type 2 error, are NY 26, FL 13 and TX 22. However, these seats do not make the list for a theoretically reconcilable reason: well-funded candidates have induced the parties to stay away. All three races do not get into the first list because one party or both are not involved; however, each candidate who corresponds with a demurring party has access to hundreds of thousands of dollars (Nick Lampson, Jack Davis, Tom Reynolds, Christine Jennings, and Vern Buchanan). Another Republican-held seat that is excluded from the first list is AZ 08, but that would be because the GOP has ostensibly ceded it to Gabrielle Giffords and the Democrats.

All in all, then, this list seems to me to be very sound. It (a) is theoretically sensible, (b) produces results that cohere with what we know to be true, (c) produces results that do not cohere with what we suspect is over-estimation of the battlefield, and (d) produces no inexplicable error. It thus has some real virtues to it. Absent reliable polling in each district, I would say that these 35 seats, plus the 4 seats that it fails to capture, are the real battlefield. That would mean that 37 Republican seats and 4 Democratic seats are, in one way or another, up for grabs.

This might seem like a lot for the Republicans to defend, and from a certain perspective it is. However, whether or not the Democrats pick up control of the House by plucking a net of 15 of these districts really depends upon the probability of flipping we assign to each race. If, for instance, both parties have an equal shot in every seat, we should expect the Democrats to net 16 to 17 seats - and the Democrats have a 73% chance of taking the House. If the Democrats have a 40% chance in every seat, we should expect them to net 12 to 13 seats - and they have a 25% chance of taking the House. If the Democrats have a 60% chance in every seat, we should expect them to net 20 to 21 seats - and they have a 97% chance of taking the House.

This is the main reason I am skeptical of the "wave," i.e. a net of 25 or more for the Democrats. Even if we give the Democrats 2/1 odds in each contest, with this battlefield there is still only a 35% chance that they net 25 or more seats.

What are the true probabilities for these races? I honestly have not the foggiest idea beyond some basic intuitions (e.g. the GOP has a less than 50% chance in at least 6 to 7 seats). Of course, lots and lots (and lots and lots) of people are ready and willing to assign very specific probabilities to these races. But are they able? What are the data points we should use in such an endeavor?

Should we use House polls from companies we have never heard of, who are obviously pushing polls to drum up business for themselves after the election, who use samples that have strange origins, who use methods that are unpublished and probably underdetermined, who publish results that contradict other polls?

Should we use the rumors and innuendos we happen to stumble upon, the inside gossip to which almost all of us are not privy, and to which - if we are privy - we hear third, fourth, or fifth hand?

Should we use a favored set of anecdotes, interesting stories told by local news outlets on a given horse race that capture our attention, even if its actual effect on the race is undeterminable?

Assuming we can use any of this data - a condition that I think remains unfulfilled - another question presents itself: how do we use this data to assign odds? What weight should we give each data point? I honestly have no clue. It seems to me to be quite easy, in almost all districts, to write a storyline, a believable storyline, that favors one side over the other - and from that assign a probability. But, at the end of the day, what is the data that is inducing this assignment? It seems to me that it is is just a set of questionable polls, unfounded rumors and potentially irrelevant anecdotes to which we have, without any real justification, assigned determinative weight.

Like I said - there is just so much cloudiness out there. Lots of cloudiness over what the actual battlefield is. Lots of cloudiness over the chances that each party has in each contest. I cannot really use the money clear up the cloudiness over the probabilities. It will only take us so far. Fortunately, I think it takes us some distance. I think it clears up the actual battlefield.

© 2000-2006 All Rights Reserved

Email Friend | Print | RSS | Add to | Add to Digg
Sponsored Links

Jay Cost
Author Archive