« FRB Dallas: Economic Outlook and Quarterly Energy Update | Main | Where Were You When I Needed You? »

Saturday, May 12, 2007

Spinning in Circles

I want to return to the idea of spin, partly because I didn't like my first stab at this in response to a Stanly Fish's piece in a blog at the NY Times, or my second response after Brad DeLong's follow-up.

Stanley Fish took the position that Karl Rove is not spinning the data when he talks about the increase in average income, but does not note the increase in inequality that comes along with it:

...But... a 2006 statement by Karl Rove to the effect that “Real disposable income has risen almost 14 percent since President Bush took office.” Jackson and Jamieson regard this claim as “so divorced from reality as to seem unhinged.” Why? Because the real disposable income Rove cited “was a statistic that measures the total increase in income, not how that income is distributed.” That is to say, the 14-percent increase did not benefit everyone, but went largely “to those in the upper half of society”; the disposable income of the lower half had “fallen by 3.6 percent.”

Does this prove spin? I don’t think so. What it proves is that in Rove’s view, the health of the economy is to be gauged by looking at how big investors and property owners are doing, while in Jackson’s and Jamieson’s view, an economy is not healthy unless the fruits of its growth are widely shared. This is a real difference, but it is a difference in beliefs about what conditions must obtain if an economy is to be pronounced healthy. It is not a difference between a clear-eyed view of the matter and a view colored by a partisan agenda...

Brad DeLong objects:

...Rove says ... "Real disposable income has risen almost 14 percent since President Bush took office." And then he is silent. And his subsequent silence speaks. His subsequent silence says that that 14% number is what you need to know--that there is no major qualification that needs to be made. And so Rove's subsequent silence is a loudly spoken lie: Rove agrees with Jackson and Jamieson that a healthy economy is one with large and broadly-distributed income gains. And he agrees with Jackson and Jamieson that that is not the case. But Rove wants his auditors to think that that is in fact the case--hence his one sentence, followed by silence. ...

So Brad is arguing this is spin, Stanley Fish is arguing it isn't. I initially disagreed with Brad, but once spin is defined - what is meant by spin wasn't clear initially - I think we are closer together. But we shall see.

Suppose we have data on three different things characterizing the economy, A, B, and C. For example, A could be the inflation rate, B the unemployment rate, and C the Gini coefficient.

There are two groups of people, Rs and Ds.

Rs care about A and B, but their votes don't depend upon C.

Ds care about B and C, but their votes don't depend upon A.

A good economy then, broadly speaking, is when all three variables go up. A bad economy is when one of the three goes down. Everyone knows this. When B goes down, both groups are negatively impacted, and if A or C goes down, then one group or the other will be negatively impacted.

This is the first place where I am going to differ with Stanley Fish because it leads him to the wrong conclusion. In his view, Rove doesn't believe that the health of the economy depends upon C and the previous debate, unfortunately, got stuck on whether Rove would believe this or not. I think it's better to proceed along the lines of this model - there's a difference between what's good for the economy overall and what's good for the subset of the population Rove represents. Rove knows that C matters overall, but he also knows it only hurts his cause to bring it up.

Suppose that data comes out on A, B, and C and the news is good about A and B, but the news about C is bad.

I have no problem with Rs emphasizing data on A and B as it comes out, and downplaying data about C when it's negative. To me, that's spin and that's what Rove was doing - it's not an accurate reflection of the broader economy - it's just the good news his constituents care about along with the good news both groups care about. Maybe there's some larger obligation for, say, the president to acknowledge all three pieces of data when discussing the state of the economy, not just what's good for one of the two parties, but generally Rs would be expected to shout about A and B, and forget to mention C. The other side can counter with their arguments and data, and their are other sources of information.

I also don't have a problem - to some degree - with a group using the measures of A, B, and C that best represent their group's case. If there is more than one measure of inflation, and there is, and if we aren't completely sure which is best, and we aren't, then using one or the other to make your case in the political arena is not something I would get too excited about (though in an academic setting I would expect a footnote and a robustness check to see if the measurement choice matters). Thus, I don't have a problem with a group choosing particular measures of aggregate activity, inflation, wealth, etc. to highlight a political case. Again, this is spin. By always making choices in one direction and failing to qualify the results accordingly, the picture is distorted.

But there's a line. So far, this says take an honest cut at the data and make your best case, and since this is politics we understand the choices you make will be in your favor.

But it's not okay to misrepresent the data, to "spin" it into something that it isn't by deceptiveness. For example, intentionally selecting particular start and end dates for a data series to amplify a point in your favor or your opponents disfavor is a dishonest presentation of the facts. To me, that's not spin, that's intentionally misleading. Presenting a measure you know is flawed, but that supports your case does the same.

Suppose it were a product on the market and you were to ask the manufacturer about reliability. There may be some very legitimate choices to make about what to include in the measure of product failure and what to leave out, and left up to the manufacturer those choices will tend to fall a particular way, toward leaving things out. But it wouldn't be okay to, say, pick a period when failures were abnormally low and report that as your failure rate in product advertising. It intentionally misleads.

The use of data in this way in the political arena is common, unfortunately, and I don't consider that spin. It is hackery, it is designed to mislead rather than convince, it is lying to the public.

Was Stanley Fish correct to argue that Rove was not spinning the data? I don't think so. It's partly definitional - what is meant by spin. But by my definition, there's no doubt. Rove was spinning the data.

Bruce Wilder in comments. As I noted in response, Bruce has a nice statement of the first point I was trying to make:

Spin is making statements of fact, which encourage a particular judgment by suggesting a narrative analysis, favorable to one's own point of view and values.

The fact that the statements of fact are selective, and the selection favors a particular narrative, is, by itself, legitimate.

Political affairs necessarily involve contests in which conflicts are resolved, conflicts where the goals and desires of opposing groups are mutually exclusive. Rs want lower labor cost; Ds want higher wages. There might be resolutions, where both get some of what they want, but such happy solutions are not privileged or guaranteed.

Bruce goes on to add:

The classic and definitive example of BushSpeak spin is certainly the infamous 16 words. Here is an analysis by Tristero of Hullabaloo:

For many of us, it was quite clear from the instant they were uttered that the famous 16 words were a bald-faced lie of monumental proportions: The British government has learned that Saddam Hussein recently sought significant quantities of uranium from Africa. ...

[I]t was clear that no one in the Bush administration - including Bush himself, or he surely would have approved a stronger statement that was less carefully hedged - believed for a second that Saddam had recently sought significant quantities of uranium from Africa. They just wanted you to believe it, and they were prepared to lie about it.

Now, it is certainly possible to look at the 16 words, and confirm that they are perfectly truthful, and, therefore, Bush was not "lying". Here is the analysis by factcheck.org. But, of course, factcheck.org is missing the deliberate deception, which Tristero (correctly, imo) immediately detects.

Brad's claim is like Tristero's, and Fish's claim is like that by factcheck.org.

Like Tristero's claim, Brad's turns on the speaker's genuine judgment being substantially different from the one he is trying to induce in his audience.

That the speaker should choose not to acknowledge some validity to the judgment of an opponent, not in the room, seems, well, par for the course.

And I agree. What Bruce describes is not my definition of spinning and one test of that is, as Bruce says, a conflict between what the speaker knows or believes and what he or she attempts to convey, whether or not the words are technically true by some metric.

    Posted by on Saturday, May 12, 2007 at 12:15 AM in Economics, Politics | Permalink  TrackBack (0)  Comments (16)


    TrackBack URL for this entry:

    Listed below are links to weblogs that reference Spinning in Circles:


    Feed You can follow this conversation by subscribing to the comment feed for this post.