Friday’s disappointing jobs report has already generated a lot of speculation about how it might impact the election in November. See, for example, this piece by Helene Cooper and Annie Lowrey, which quotes both Andy and me. My basic take is that the jobs report won’t shift the dynamics of this election very much. Here is why:
1) Late changes in objective economic indicators may not matter. Consider this interesting finding from Robert Erikson and Christopher Wlezien’s book, The Timeline of Presidential Elections. They find that from 1952-2008, one objective measure of the economy — per capita income growth — correlated as strongly with the presidential election outcome in April as in November. To quote them:
bq. …one can predict that actual vote from the current income growth about as well in April as in November. By April, the economic cake is largely baked.
2) The jobs report doesn’t represent a big change. The rejoinder to #1 is, “Well, what if the late change was drastic?” What if the economy collapsed, as in the fourth quarter of 2008? That probably would have a larger impact on the race. (It certainly appeared to in 2008, although it is difficult to peg discrete changes in the tracking polls to specific events in the financial crisis. See the graph here.) But this jobs report is mostly status quo.
3) News reporting didn’t change. Because the jobs report was mostly status quo, so was the news reporting — or at least such was my impression. The jobs report got covered, but not in a way that signals to the average voter that something dramatically bad has happened. It’s important to focus on news coverage because most voters aren’t refreshing the BEA webpage and instead rely on news coverage for their impressions of the national economy.
This gets us to the possibility raised by Andy’s comment in the NY Times piece:
bq. I’m reminded of 1992, when, after the election, a story went around that G. H. W. Bush lost because there was a perception of the economy in recession even though the economy was already improving.
One reason for that perception might have been news coverage. As Marc Hetherington has argued, the more news people reported consuming in 1992, the more they were dissatisfied with the economy. The same thing was not true in 1984 and 1988, suggesting that 1992 was different.
What about this year? It’s the opposite: the more people pay attention to news coverage, the better they think the economy is doing. Here’s a graph the past 7 months of YouGov data:
People who watched a lot of news were much more likely to say that the economy was doing better. To be sure, this relationship is strongest for Democrats. The difference between the least and most attentive Democratic news consumers is 34 points. Among independents, it’s 24 points. Among Republicans, it’s only 7 points. But still, among the group most likely to vote based on its perceptions the economy — independents — there’s no evidence that watching the news makes one more pessimistic about the economy. The same is also true in more recent polls — I looked at the June-August polls specifically — which were conducted when consumer confidence has been dropping.
In short, 2012 doesn’t look like 1992 in this respect: it’s not an election where negative news coverage of the economy may hurt the incumbent even as the objective economy improves.
And unless the objective economy takes a much sharper upturn or nosedive than this most recent jobs report suggests, I don’t think that any late economic trends will matter much come November.