Plural of Data ≠ Anecdote

by Henry Farrell on August 29, 2006

“Mike Pence”:http://www.nytimes.com/2006/08/29/washington/29pence.html?pagewanted=2&ei=5090&en=b8427076a837addf&ex=1314504000&partner=rssuserland&emc=rss on evidence-based reasoning.

Mr. Pence argued that tax cuts help the poor by revving the economy. That may eventually prove true, but despite large tax cuts the poverty rate has risen in each of the last four years. “That’s anecdotal,” Mr. Pence said in an interview last fall. Then he offered an anecdote — a story President Reagan told about a pipe fitter pleased to see the rich prosper, “because I’ve never been hired by a poor man.”

{ 37 comments }

1

Ray 08.29.06 at 8:42 am

It’s not even about “anecdotes not being the plural of data”. It’s about ‘anecdote’ being more than “a derogatory term that you attach to ideas you don’t like”.

2

abb1 08.29.06 at 8:53 am

David Rees had this reaction to Mr. Pence’s year-old statement, made, apparently, in relation to the Katrina disaster.

3

Patrick 08.29.06 at 9:20 am

I love it!

I am going to use that from now on, whenever my advisor asks about any data or studies I haven’t looked at for my thesis.

4

soru 08.29.06 at 9:20 am

I don’t think a tradesman to the rich is likely to fall into any reasonable definition of ‘poverty’.

It’s a variant on the Marxist concept of class interests – the people you interact with economically are generally those who benifit from interacting economically with you. Consequently, they are glad you exist.

As a sampling technique, using those people as a source of even anecdotes, let alone data, is approximately as reliable as asking the fairies at the bottom of the garden.

5

lemuel pitkin 08.29.06 at 9:35 am

Not just not the plural, but the exact opposite, etymologically.

Data is latin for “given,” based on the formula used by Romans for dating their letters (date itself has the same origin.)

Anecdote is from Gk. anekdota “things unpublished,” neut. pl. of anekdotos, from an- “not” + ekdotos “published,” from ek- “out” + didonai “to give.” Procopius’ 6c. Anecdota, unpublished memoirs of Emperor Justinian full of court gossip — i.e. literally “not given.”

I love the Internet.

6

Richard Bellamy 08.29.06 at 10:03 am

Data = Any fact or facts that fit my pre-existing theory.

Anecdote = Any fact or facts that contradict my pre-existing theory.

7

Guest 08.29.06 at 10:14 am

My favorite economics anecdote is the one about the plumber who said, “Tax cuts for the rich help the poor because a stitch in time saves nine.” Really blows you away with its homely wisdom, huh?

8

tom s. 08.29.06 at 10:56 am

– that would be a Polish plumber, of course?

9

Andrew Edwards 08.29.06 at 11:30 am

One time this farm labourer said to me, “Andrew, you’re the best”. I therefore conclude that since I am obviously the best, I should be named President of the Universe.

10

dearieme 08.29.06 at 12:03 pm

But in social science, often “data” is the plural of bollock.

11

bi 08.29.06 at 12:27 pm

Guest: That’s a good one, even if I don’t know what that means. :) But since I don’t know what that means…

Tax cuts for the rich help the poor, because guns don’t kill children, children kill children.

Tax cuts for the rich help the poor, because Jesus saves.

Tax cuts for the rich help the poor, because absence makes the heart grow fonder.

12

Henry 08.29.06 at 12:35 pm

But in social science, often “data” is the plural of bollock.

Which stands in stark contrast to the sober and scientific approach to hypothesis testing adopted by non-social science academics when they “want to make arguments”:http://janegalt.net/cgi-bin/MT/mt-comments.cgi?entry_id=9425#110070 about complicated social phenomena ;)

13

derek c 08.29.06 at 12:59 pm

The pipe fitter anecdote, besides all its other problems (like having no evidence that the conversation ever took place: cf. “Friedman’s Cabbie”), has the problem that it doesn’t even make sense in its own universe.

If a pipe fitter wishes more customers, then he wishes more people who are *just* rich enough to want to employ him. He doesn’t need his current base of customers to become one that is half as large but twice as individually rich, or one tenth as large and ten times as rich. Consequently, a *real* pipe fitter would be wishing for the redistribution of wealth, not its consolidation.

If this was a science fiction story, I would say that it was not true to its premises.

14

derek c 08.29.06 at 1:05 pm

The pipe fitter anecdote, besides all its other problems (like having no evidence that the conversation ever took place: cf. “Friedman’s Cabbie”), has the problem that it doesn’t even make sense in its own universe.

If a pipe fitter wishes more customers, then he wishes more people who are _just_ rich enough to want to employ him. He doesn’t need his current base of customers to become one that is half as large but twice as individually rich, or one tenth as large and ten times as rich. Consequently, a _real_ pipe fitter would be wishing for the redistribution of wealth, not its consolidation.

If this was a science fiction story, I would say that it was not true to its premises.

15

Thomas 08.29.06 at 1:16 pm

What an odd story for the Times to reprint a second time (the conversation took place last September).

In the meantime we have more “data”–the poverty rate has fallen, the government says. I guess that means Pence was right to cut taxes last year!

Wow, this macroeconomics stuff is easy!

16

eudoxis 08.29.06 at 1:23 pm

When one picks a handful of data points out of many, it’s not exactly anecdotal, but not the whole picture, either. What is that called? Lying?

17

eudoxis 08.29.06 at 1:25 pm

At the limits, it’s clear that poverty would rise with reduced taxation. It’s been my understanding that the recent tax cuts are not offset by higher taxation of the poverty class, necessarily, since the government has increased borrowing.

18

Henry 08.29.06 at 1:28 pm

Thomas, you’re really not doing any favors to your political position by commenting here – you’re doing a far better job than I ever could of confirming a stereotype of the fact-shy conservative who’d prefer to debate anything other than the actual issues. There are smart conservatives out there with interesting and challenging arguments. Unfortunately, you’re not one of them.

19

Uncle Kvetch 08.29.06 at 1:35 pm

In the meantime we have more “data”—the poverty rate has fallen, the government says. I guess that means Pence was right to cut taxes last year!

WASHINGTON — The nation’s poverty rate was essentially unchanged last year, the first year it hasn’t increased since before President Bush took office.

The Census Bureau reported Tuesday that 37 million Americans were living under the poverty line last year — about 12.6 percent of the population. That’s down from 12.7 percent in 2004, but census officials said the change was statistically insignificant.

Wow, this macroeconomics stuff is easy!

Not nearly as easy as making shit up.

20

Thomas 08.29.06 at 2:00 pm

Henry, what are the actual issues you raised? I see a jokey post about a politican talking to a pressman, followed by the typical orgy of self-congratulation at CT. The post itself reveals you may be half as clever as a reporter for the NYTimes–your parents should be so pleased.

But, please, bring out the facts–looking at the string, I was the first to mention the facts, I’m happy to talk about them, and I’m not shy at all.

21

bi 08.29.06 at 2:06 pm

Thomas: Read. The. Title.

And stop talking about people’s parents, because if you don’t stop, then your mom’s a terrorist.

22

bi 08.29.06 at 2:13 pm

Speaking of data and anecdotes, here’s an interesting experiment to try out:

Take all of Thomas’s CT posts, and compute the percentage of posts that engage in name-calling. Do the same for other people’s CT posts. I’m not sure what that’ll show, but it should be interesting in any case.

23

Dæn 08.29.06 at 2:18 pm

Hah, this quote should be reworked as an 8th-grade standardized vocab item:

Mr. Pence argued that tax cuts help the poor by revving the economy. That may eventually prove true, but despite large tax cuts the poverty rate has risen in each of the last four years. “That’s (anecdotal/a conclusion based on the best statistical evidence we have available),” Mr. Pence said in an interview last fall. Then he offered an (anecdote/self-sufficiently persuasive economic argument in convenient homily form) — a story President Reagan told about a pipe fitter pleased to see the rich prosper, “because I’ve never been hired by a poor man.”

This would at least help us suss out whether Pence is genuinely treacherous in his term-twisting or merely ignorant.

24

Henry 08.29.06 at 2:57 pm

Thomas, your post consisted of a _No! No! Look over here!_ irrelevancy and a bogus factual claim. The Unteachable in Defence of the Indefensible. I know that lawyers are supposed to make bad cases sound good when they’re stuck defending them, but really, you should either do a better job than this or remain silent. We have pretty broad standards for contributions to argument here at CT, but you’re failing them badly. If I see comments as wilfully stupid as your first comment to this post in the future, I’ll delete ’em. If they make some minimal contribution to argument I’ll let them stand.

25

Thomas 08.29.06 at 3:54 pm

henry, that repeats your assertion that there’s some sort of argument here, but I still don’t see what argument you’re making. Are you saying that the fact of four years of increases in the poverty rate (followed, as it happens, by a statistically insignificant decline in the poverty rate) answers the question of whether tax cuts stimulate the economy? I’m not sure what the proper response to such an inanity is–surely it isn’t Pence’s response. Maybe eye rolling?

The standards for argument are certainly very broad when your framing up Pence’s response counts as an argument in itself. I thought my initial response–which focused on the concept of “data” you were using–was right on point. If it’s “evidence” that’s the core of your argument, have you reconsidered the argument’s merits today, in light of the Census report? How has your reaction differed from Pence’s?

26

John I 08.29.06 at 3:58 pm

There’s poverty, and there’s severe poverty which has been rising under GWB:

http://www.newswise.com/articles/view/522916/?sc=dwhp

Data is is parsable into more than anecdote.

via Ezra Klein:
http://ezraklein.typepad.com/

27

Henry 08.29.06 at 4:15 pm

Thomas – let me explain the post in terms suitable for the meanest understanding. I’ll try not to use too many long or complicated words. Mike Pence dismissed real data which was counter to his prejudices as “anecdotal.” He then proceeded to support his own point of view with an anecdote. There is a saying, common in the social sciences that the plural of anecdote is not data. Therefore, the joke that Mike Pence dismissed “plural” data as being an anecdote. Got it? It’s not a post on the relationship between tax cuts and poverty. It’s explicitly a post on the way in which Mike Pence treats evidence. And this discussion, which has gone on too long, is about the way that you’ve tried to derail argument by (a) introducing the irrelevant fact that the NYT has mentioned some of this in a previous story, and (b) making the bullshit claim that a statistically insignificant drop in the poverty rate proves that Pence was right all along.

28

a different chris 08.29.06 at 4:56 pm

Henry, your efforts have maybe given another angel it’s wings but are wasted on Thomas.

Anyway, there’s an even weirder thing about this anectdote. I dunno about “Old Europe”, but a lot of people seem to think a pipefitter is a plumber.

They are not, at least not the plumbers you find in the consumer Yellow Pages:

http://www.ua449.com/home.htm

They are a heavily unionized group of people that tend to work on big projects and have a very long historical lineage (hence the “Steamfitters”). I guess one of the “Reagan Democrats” *could* have said that, but it’s a little strange to think so.

29

Thomas 08.29.06 at 4:58 pm

henry, I got the joke. I even noted how clever you are.
I too made a joke, also about standards of evidence. After all, if, as the Times (and some others) suggest, the “data” on poverty is in tension with Pence’s views, then surely the new “data” will be relevant to the discussion–so I jokingly said that the new “data” showed Pence was right. Of course, just like the “data” cited in the article, the new “data” didn’t resolve the merits of Pence’s claims at all, as anyone can see. You just didn’t see that I was joking, which of course is my fault, not yours. I remain confused about your references to “facts” and “actual issues” and the question of whether “Pence was right all along” because those seem entirely unrelated to what you say your point with the post was, but I suppose I shall remain confused, because this has gone on too long.

30

Uncle Kvetch 08.29.06 at 5:50 pm

God help me, I shouldn’t…but I’m powerless to resist.

Of course, just like the “data” cited in the article, the new “data” didn’t resolve the merits of Pence’s claims at all, as anyone can see.

Thomas, Pence is claiming that tax cuts will help the poor. Five and a half years into the Bush administration, the poverty rate is higher than it was when Bush took office.

Please…no joking this time…just what kind of data would “resolve the merits of Pence’s claims”? I’m struggling as much as Henry to see what, if any, point you’re trying to make here. My sneaking suspicion is that as long as the poverty rate kept going up, or remained stable, you’d be telling us that it’s “too soon to tell” whether the tax cuts are going to work. Am I getting warmer?

31

Thomas 08.29.06 at 7:17 pm

UK, you’re wrong–Henry isn’t trying to understand anything about the merits of Pence’s claims at all. See his most recent post. But, tell me, did the “data” released today cause you to revisit the merits of tax cuts?

I didn’t think so. Following Henry’s lead, we can ignore the merits and safely conclude that you aren’t any better than Pence.

32

Walt 08.29.06 at 8:32 pm

Woo-hoo! Thomas has won! He has completely derailed the thread! Congratulations, Thomas! I’m sure there will be slaps on the back all around at the watercooler at whatever wingnut-billionaire think-tank you call home.

33

Uncle Kvetch 08.29.06 at 10:31 pm

I surrender, Thomas–I have no idea what the fuck your point is, or even if you have one. I guess that means you win.

34

+h0m4s 08.30.06 at 12:37 am

guys, it’s pretty simple: i overheard a conversation in a bar one time where you can prove using pictures that the government will always actually increase its revenue by cutting taxes. And then there are the leading indicators. I tried to do the same thing to Kieran’s post about demography. Apparently I seem to be the only one who doesn’t understand the point of blog posts. I don’t believe blogs exist, or if they do, they are purely anecdotal.

35

nick s 08.30.06 at 2:27 am

Perhaps if the title had been “You keep using that word. I do not think it means what you think it means” then thomas wouldn’t have bothered providing ample date to prove how stupid he is.

36

nick s 08.30.06 at 2:29 am

Gah. ‘Data’. My typist is now among the ranks of the unemployed.

37

Tim Worstall 08.30.06 at 3:26 am

“The Census Bureau reported Tuesday that 37 million Americans were living under the poverty line last year—about 12.6 percent of the population. That’s down from 12.7 percent in 2004, but census officials said the change was statistically insignificant.”

Err, no, that isn’t what was reported at all. Yes, what was reported was that 12.6% of the population was below the Federal Poverty Line. This is not, not at all, the same as saying that 12.6% of the population were living below the poverty line.
The poverty rate as reported is calculated before the impact of both the tax system and in kind benefits. Given that the major programs of poverty alleviation in the US are via the tax system (the EITC) and in kind benefits (Medicaid, housing vouchers and food stamps) the number below that poverty rate is not a guide to the number living in poverty.
It is a guide to the number of people who need help in order not to live in poverty, but it does not measure most of the help that they do receive.

Whether people should receive more help (as I’d guess most people here would insist they should) is an entirely different question. But to talk about poverty we should at least be talking about how much there is after whatever alleviation has been done, not about how much there is before, otherwise, how would we know how much more alleviation we ought to do?

For example, the US could use the same measurements of poverty as most of Europe (and I think codified by the OECD): post-tax, post-benefit, less than 60% of median income, adjusted for household size.

Comments on this entry are closed.