Quickie math (arithmetic)/journalism peeve: I don’t think “party lines” means what you think it means/NY Times edition.

Update: a couple of changes below to try to make the point I thought I was making.

From today’s NY Times article on the Senate vote to extend the Cash for Clunkers program:

In the end, the bill passed largely along party lines, with 51 Democrats, 2 independents and 7 Republicans voting for it.

What’s wrong with that?

This phrase:  “largely along party lines.”

Let’s do the math, shall we?

7 Republican senators voted with the majority to pass the bill.  4 Democrats joined the minority opposing the bill.  So the Democrats did come close to holding party lines on the bill; only 6.7 % of the Democratic caucus defected.

But 7 Republicans?   That’s 17.5% of the rump opposition party.  That’s a substantial defection, a significant breach in party discipline, especially given the record of GOP monolithic “no-ness” on most major votes this year.  Even more so when you consider that this is in essence a small supplement to the economic stimulus bill, and as such would naturally be a target for a Republican party that has pushed a lot of chips onto the table betting against the success of Obama’s economic policies.

Numeracy matters — and it does so not just at the level in which an understanding of statistics in any kind of deep way might be helpful, but even at this simplest level of fifth grade math.

I’ve written before on the importance of distinguishing between raw numbers and some representation of the data that permits comparisons to be made across different circumstances or events.  See this post for a much sadder example.  But the point keeps getting made by the way journalists handle numbers.  Simply the act of deriving a percentage from the tally sheets is way too often a bridge too far.

Why the failure to count on one’s fingers at such an elevated institution as the Times?

I don’t know for sure, certainly — I wasn’t in Matthew Wald’s head when he wrote the piece, nor the room in which his editors ran their eyes ove rhis copy.  But I do note that the running journalistic line for weeks now, significant especially in the context of the health care debate, has been that the Democrats have failed to meet promises of bipartisanship, and that hence, their legislative agenda and even their (our) legitimacy is in doubt.

If that’s the story line, then clearly each significant vote must be characterized as evidence of the ongoing division between the parties, and the inability of the Democrats to construct legislation attractive to a rumored center and the mostly mythical moderate wing of a party dominated by Limbaugh dead-enders.

And so, even in this case, when Democratic legislation on a high profile economic and environmental* program attracts a significant number of Republican votes, enough to release some Democratic senators to cast what may be either or both politically useful and or principled nay votes, the Times still feels compelled to assert, in essence, that this Democratic — and partly Republican — legislative victory was in fact yet another blow to bipartisanship.

The moral of this story:  this is how destructive tropes get made.  In the construction of myths epithets matter.  Little phrases, rote characterizations —  “the wily Oddysseus,” “loose-tongued Thersites” — they frame the reader’s interpretation of all the action surrounding the thus-labeled person or event.  They serve as received wisdom, a collectively agreed value judgment, to speed the plot.

And here, the phrase approximating an epithet, “party lines,” is the tell that the writer has substituted such prior qualitative interpretation for thought.  The cash for clunkers bill saw close to one fifth of the opposition party break ranks.

But to say so, to recognize that a Democratic policy initiative was sufficiently successful/popular that a significant subset of the GOP senate caucus chose to embrace it would have forced the writer (and editors) of the Times’ story  to break out of that sweet sleep in which there is no need to actually do the simplest bit of arithmetic.

Image:  Jan Breugel the Elder, “A Fantastic cave with Odysseus and Calypso,” c. 1616.

Explore posts in the same categories: Arithmetic, Journalism and its discontents, Mathematics, Politics

3 Comments on “Quickie math (arithmetic)/journalism peeve: I don’t think “party lines” means what you think it means/NY Times edition.”

  1. Jonathan Livengood Says:

    While I agree with the moral of the post, I disagree with your criticism of Wald’s article in particular. Given that better than 90% of Party A voted “yes” and better than 80% of Party B voted “no,” I find it perfectly reasonable to say that the vote was “largely” along party lines. It seems like a judgment call with numbers like this, not a grossly misleading, politically inspired mistake. How much agreement would you have needed before “largely” was appropriate? After all, Wald didn’t say the vote was “completely,” “totally,” or “almost entirely” along party lines. Maybe you think he should have included how many Democrats and Republicans the totals were out of (51/55 and 34/41 holding the lines)?

    • Tom Says:

      Fair enough, and it may be that I am excessively sensitive to what I see as the danger of stealth commentary within what purport to be straight news articles.

      But before I concede entirely, think about this way of framing the question. The point I was trying to make was that the NYT was fooled by the relatively small number — 7 — in the numerator of the fraction of the Senate Republicans who voted in favor of the bill, failing to take into account that the denominator, 40, was also small.

      Now imagine the counterfactual inverse of the situation that actually occured: the fractions of aisle crossers from each party switched magnitudes. 10.5 — say 11, rounding up — out of 60 democrats crossed party lines to vote against the bill, while on the GOP side, 3 senators joined the majority.

      In that case you would have seen the measure pass 52 to 48, and the headlines would have read that the measure barely passed in the face of GOP opposition joined by a significant number of Democratic party dissidents, i.e. bipartisan opposition. I don’ think — though I certainly can’t prove — that there would have been any talk of a party line vote.

      Why do I think that? 2 reasons. First, I think it is harder for MSM journalists to credit Democratic party successes in dividing the GOP on key votes — but more significantly, I think it is much easier to imagine that when senators votes against party in double digit numbers, that feels like a big deal in a way that a mere seven aisle crossers do not. And that’s the point I’m really trying to make, which is that most people and way to many MSM folks have a very poor grasp of what numbers mean.

      • Jonathan Livengood Says:

        The hypothetical is helpful, and as I said, I agree with the moral you drew in the original post and re-draw in your reply.

        Lots of things affect the judgments people make about numbers. As you say, people like double-digits. People also like round numbers, they pay attention to the gaps in finishes (with different reactions for close and not-close results), and they typically don’t do well with percentages, etc. I think you’re right that had the measure passed (barely) 52-48 in the way you outline, the headline would have been about Democratic defectors. (Though I suppose one could have written a headline like, “Hardliners hold on,” with a story that emphasized how the vote came out according to party lines because the bulk on both sides stayed pat. I don’t know, this probably isn’t very likely, but as I said before, I think that with these numbers, a lot is left to taste. Curse you vagueness!)

        How important do you think it is that in the case considered by Wald the final vote wasn’t really that close? And how important is it that the result would not have been different even if (as in your hypothetical) the percentages of defectors had been reversed?


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: