03/05/2009 02:45 pm ET Updated May 25, 2011

Are Stocks Like Tracking Polls?

Earlier this week, President Obama analogized recent short term stock market fluctuation to changes in a political "tracking poll." TalkingPointsMemo captured the video and Ben Smith provided a transcript:

What I'm looking at is not the day-to-day gyrations of the stock market, but the long-term ability for the United States and the entire world economy to regain its footing. And, you know, the stock market is sort of like a tracking poll in politics. You know, it bobs up and down day to day. And if you spend all your time worrying about that, then you're probably going to get the long-term strategy wrong.

I took Obama's point about the potential for meaningless "day-to-day gyrations" in both tracking polls and stock prices, yet still found the analogy a little jarring since tracking polls and the stock market differ so much. Wondering if other pollsters had a similar reaction, so I asked using the listserv of the American Association for Public Opinion Research (AAPOR) and the smaller universe of those following me on Twitter.

Most agreed that Obama's underlying point was basically valid: The stock market, like a tracking poll, is subject to a lot of short term and seemingly random variation, albeit for different reasons (see the "random walk" hypothesis). As Democratic researcher Jason Boxt put it, "one blip doesn't a trend make."

Yet several agreed with Republican pollster Alex Lundry that the analogy is "crass" or otherwise imperfect. As Doug Strand points out, "the stock market is not going up and down right now, it's largely going down." A sustained four-month drop in a tracking poll would certainly cause a pollster to reevaluate their long term strategy. But how much of that decline is a reaction to Obama's economic proposals? Opinions on that question will differ.

Whatever your judgement about the value of the stock market and opinion polls as tools to evaluate the state of the economy or Obama's long term economic strategy, it is important to remember that stock traders on any given day are not a representative sample of all Americans or even of those that own stocks or mutual funds. And while virtually all stock traders make some sort of judgement about the future value of the stocks they buy or sell, those judgements are influenced only partially by their view of the overall economy and their evaluation of the merits of Obama's economic policies.

I have reproduced the longer comments from the AAPOR listserv below, as they raise some interesting points I had not considered. Most of these comments come from survey researchers with an academic bent.

In his comments, Doug Strand asks to see more of the context of Obama's remarks. I've embedded below the full C-SPAN video of the question and answer session with British Prime Minister Gordon Brown. The question about stock market trends comes at 8:33. The version at includes an interactive transcript.

Also, our comments section is wide open, so please post your thoughts.

Mellissa Marcello, president of Pursuant Research:

I read the quote prior to your posting on AAPORnet, and I have to admit I was not too troubled by it. I believe that public opinion is fluid, and if we look at any single point in time, we run the risk of misreading how people (or the markets) really feel (perform) over the long-term. What I have to believe Obama was really saying was that he was not going to grandstand for the sake of keeping Wall Street happy. This was directed, in part, towards Bill Clinton who recently publicly urged Obama to speak more positively about the economy. Obama would like us to believe (and I have no reason to believe otherwise) that he is not going to pander to the markets, or public opinion, when it comes to working on long-term solutions to the country's economic woes.

Doug Strand, an election researcher at UC Berkeley's Survey Research Center (speaking for himself only):

It would have been helpful to see more of the context of his remarks, like a full transcript of that session with the press? (I can't find a transcript or video that encompasses more than the quote you have below).

But based just on that paragraph you quoted below, it appears to me that Obama's words were very poorly-chosen. As conservative talk show hosts have gleefully pointed out, the stock market is not going up and down right now, it's largely going down. I believe the Dow is down about 1/3 since election day. If Obama's percentage in a tracking poll had dropped 1/3 over a 4 month period after the primaries in '08, I bet Hillary would have drafted to replace him. What he may well have meant is that the stock market can go up and down from year to year -- rather than day to day -- and that alone shouldn't mislead one about the long-term growth of the stock market for investment. And that is what I believe most investment advisers would say.

Matthew Jans:

To the degree that both are quantitative time series, the analogy seems appropriate. I think his point about looking at overall trends, not dayto day (sometimes random) fluctuation is a good one (for polls and the stock market both, as well as other longitudinal quantitative measures).

Francine Cafarchia:

He is correct if you consider only those that trade and/or have some influence on Wall Street market fluctuations, but it is obviously not a representative sample. That said, the balance of the population whose reactions are not directly registered in the stock market's fluctuations will likely be influenced by the market's performance and as such, may lead to a correlated response among the rest of the population. Of course, the general population's response may be tempered by their attitudes towards "Big Business", but most people's sentiments about the economy and "how things are going" follow what is happening in the stock market. So in the end, you could have the similar results had you done a formal poll.

Phillip Downs:

I think the analogy between stock market fluctuations and fluctuations across poll results over time has some merit in that our basic values keep poll results over time from being absolutely rudderless just as the underlying value of our economy keeps the stock market from being absolutely rudderless over time.

While our attitudes on specific issues (as with valuation of specific stocks) can be fluid over time given different circumstances and information, we do not change our basic value system (or the worth of the market) from month to month or from year to year.

However, when there are extreme shocks to either system (our value system or the market), results tend to become more rudderless. For example, under life and death and other extreme circumstances, human behavior and attitudes become more variable. Under our present near-death scenario for the market, valuations become more variable.