Framing Risk: The Fear of Living Dangerously

Kevin Drum writes: "The point of the series is that although the poor may be a bit better off in average terms than they were 30 years ago, they face a far greater risk of imminent disaster." And: "Conservatives will point to the folly of buying big screen TVs on incomes that can't afford it, but the bigger picture overwhelms this kind of pettiness." I think Kevin is failing to see how this so-called big picture looks petty when viewed in a properly conservative moral light.

Drum's post reminds me of a Frum bit I quote at the start of my evergreen Dead Right post. Frum explicitly makes the non-economic, socially conservative case that the state of affairs Drum is lamenting is, in fact, optimal.

The great, overwhelming fact of a capitalist economy is risk. Everyone is at constant risk of the loss of his job, or of the destruction of his business by a competitor, or of the crash of his investment portfolio. Risk makes people circumspect. It disciplines them and teaches them self-control. Without a safety net, people won't try to vault across the big top. Social security, student loans, and other government programs make it far less catastrophic than it used to be for middle-class people to dissolve their families. Without welfare and food stamps, poor people would cling harder to working-class respectability than they do not.   

Frum is not making the case that we need liquid labor markets, say, because liquidity means a rising tide raises all boats. The point is: stormy waters swamp all boats. And that's a good thing. Makes people huddle in the bilge.

Drum: "At some point, the risk-based society that the Republican party is trying to build will finally come crashing down on them. In the meantime, it's up to Democrats to make sure that when it does, the blame falls exactly where it belongs."

But surely Republicans will just laugh it off if we hint that they are thinking like Frum. Conservative eye for the hard-bitten guy! A reality TV show about a chubby liberal who buys a big screen TV, loses his job, his middle-class lifestyle, his ability to buy sugar, has to live off the land for a while - nuts and berries (shades of Survivor); becomes a bitter, tough, Republican-voting, penny-pinching conservative. It would never get good ratings. I admit it's a bit far-fetched.

But it seems to me that the fact that Frum was - and remains - tempted by this loser line points the way to some 'framing' stategies (if we must use the Lakoffian term.) But it's sort of important to get your facts tolerably straight before you start flinging rhetoric. Regarding social security privatization, say, I realize I don't know the degree to which Republican proposals are specifically open to the objection that they oblige people to live more dangerously than they might like. There are many other objections; and much Republican mendacity. I am just curious about this one point. Also, to what degree do the proposals allow living truly dangerously? To what degree is it true that folks might gamble away their retirement money? I realize these privatizations schemes are inchoate at this point. The sausage making has hardly begin; so answers must be speculative.

Comments

[foo]

Frum aside... I have a conservative relative (and a smart guy, who's also had some serious education in economics) who argues this point from the other side of Frum. When I (or someone else) point out severe income and wealth distributions, with heavy tails skewed to the high end... he comes back at me with the argument that the distribution hides the underlying volatility. In other words, incomes change rapidly, people in the top end fall out, people at the bottom rise up...

It's as if he has this mental model of wealth distributions that resembles some half-formed Markov chain, and what matters isn't the distribution of states but the mixing time. (There are obviously all sorts of problems with this, not least of which that any statistic he quotes to me to support this vastly under-determines the qualitative properties of the system whose existence he claims...)

So you can see why an article like this is interesting to me.... but my question is, (and I'm not asking it rhetorically, I'd seriously like to know) aren't most people who read this article taking risk to necessarily be something negative?

Here's a methodological quote from LAT:
The Times' basic finding is that the fluctuations in annual income that individual families have experienced have grown larger over the last three decades.

Based on the panel-study sample, The Times estimated the annual income swings, up or down, for 68% of all U.S. families — those who did not have the most extreme fluctuations. As a result, the newspaper's conclusions don't rest on cases outside the mainstream: the movie star whose career dries up overnight, say, or the hourly worker who wins the lottery.

So it seems to me, that this "risk" is more like the variance of the distribution of income swings, and that it can go up or down.

Now, I can (of course) see all sorts of reasons why higher variance at the low end of the spectrum can be a bad thing (I don't need anyone to point that out to me).

But I can also see how it fits into my relative's argument too: "you need higher variance at the lower end, so that people have a reasonable chance of moving into a higher income bracket in a reasonable time."

So is there some skew to the distribution of income movements (up or down)? Can we recast Frum's argument (right or wrong) into this language? ("Introducing safety nets would skew the distribution downwards; no safety nets would make the distribution more symmetric," or something?)

Anyway, that's always my complaint with the PSID -- I want the parameters of the corresponding markov-chain-like system, but they don't appear (I think) to be easily extracted in that form. Drat.

[jholbo]

The odd thing about Frum is that he seems to suggest that misery is a means to the end of morality, not upward mobility. I think there is some small hope that other conservatives think the same, so that it will be rhetorically effective to poke them with the 'why is high risk good?' stick and see what burbles up to the surface. (Very likely they will say what your relative says, which makes a good deal more sense, prima facie, foiling my rhetorical plans.)

I thought something like what you are thinking suggesting the article data, although I don't have anything like the stats know-how. I'm quite untutored in this important field, sadly, but it seems like it can be spun as you say.

This is sort of why I'm interested in whether Republicans are truly encouraging similar volatility at the geriatic end of the spectrum. Are they encouraging people to gamble with their pensions? If so, that's something to stick them for. Harder to see the utility of that. (But maybe these accusations against the privatization plans are not so solid, in which case there are solider points to fall back on.)

[foo]

Foiling of rhetorical plans aside (not something I'm intending to do, btw)...

I would object, a little, to the characterization of this "up or down skew" issue as "spin," necessarily. I mean, it strikes me that saying something like, "the variance of the distribution of the derivative of the income (or some reasonable finite stand-in) is inversely related to the size of the income" could either be

a) pretty much a mathematico-economic fact about income fluctuations in certain kinds of economic systems, and therefore a pretty vacuous statement, or

b) something that's not necessarily the case, in which case it's either
1) good for people at the low end, or
2) bad for people at the low end.

I saw this mentioned a day or two ago on Kevin Drum's site, and I feel like he made exactly the same assumption that the LAT does.... (a) it exists (which the PSID probably makes clear), and (b) it's undoubtedly in the B-2 case above.

But ... and maybe this is just my major lack of economics education shining through... I would actually like to see a reasoned explanation of this phenomenon, and not an article that's 4/5ths anecdote, as the LAT's was.

In the style of Brad DeLong: Why oh why can't we have better economics journalists?

On another tangent, this reminds me of the whole brou-ha-ha over Bill Bennett's gambling a few years ago, and his (insane) insistence that he had actually won money at slots.... which is, if you take an extremely simple statistical viewpoint, probably related to this income-volatility thing. He played, he lost, and it reminded me of a line from one of my fave stats books: in a game against an infinitely rich adversary the probability of an ultimate ruin is one if and only if the expected gain in a single trial is less than or equal to zero."

Maybe that's the answer then... all these reports are talking about the variance of the income-change distribution, but what we're really interested is the mean? Or maybe the stock-market's not a good approximation for Bennett's "infinitely rich (casino) opponent."

I'm sure someone with a more formal grasp of statistics will correct me. Time to go back to work...

[foo]

One more quick note... this is what I hate about statistical graphics, in particular the graph on Drum's page... what we're getting is a time-series, the variance over time. But this obscures two things: (1) membership in the different strata of the graphic is changing over time, and (2) it's only the variance. We should at least see an actual distribution of wealth changes, for a single set of people, in a single time-period... and we could, by eye, infer the top coupla statistical moments for ourselves (mean, variance, skew, etc).

[Matt Weiner]

I am so glad you posted on Frum because it gives me a chance to link to this highly relevant Onion piece:

What This Town Needs Is A Child In A Well

Though actually it seems from this as though Frum is making a much less credible point. Risk is good because it teaches people not to get ideas above their station. Or is it that risk does force people to stick with family and community, as Janet Casey argues?

foo, I agree that the statistics are massively unclear from the graph. What's being plotted is "maximum swings in annual income for most families at the 20th, 50th, and 90th percentile of the income distribution"? What's that mean exactly? Presumably they take the distribution of individual swings for all families in the percentile and then they winkle "maximum swings for most families" out of that distribution but (as not a statistics expert) I wish I knew how they'd done that.

[foo]

Matt, I'm not sure what it means, but there is a semi-extended (although mathematically imprecise) description of methods at the end of the LAtimes article.

And, now that I think about it, this sounds even worse. They're not even plotting "variance" (which is what most statistical people would probably start with).... if they're using, instead, some kind of "maximum jump," then things get even weirder.

The distributions of things like that are notoriously tricky...

[Doctor Memory]

Nothing for The Wire. Not a single nomination. For shame.

[Doctor Memory]

Oh good lord. Here we see perfectly demonstrated the hidden dangers created by the conjunction of tabbed web browsers and visually identical blog comment forms: the above comment was meant to go here.

Some days, I should not be allowed near a keyboard. John, do feel free to delete both of these (or not, depending on how amusing you find this).

[belle waring]

Nah, we're just glad you're keeping your focus on the important things, Doc M ;)

[Nicholas Weininger]

This is a nice reminder of the difference between the conservative and libertarian treatment of risk, and another reason to call bullshit on conservatives like this when they claim to have libertarian leanings. Frum doesn't even bother to mention any sort of argument like "risk is the inevitable corollary of liberty; the only way to reduce A's risk is to force B to take some of it on instead, which inevitably reduces B's freedom and is thus wrong, because B is not A's keeper and comparisons between A's and B's utilities are illegitimate." No, it's all about keeping people in their place. Sigh.

[foo]

Just out of curiosity... what do you think is the "conservative treatment" of risk, and how do you think it differs from the "libertarian treatment?"

[Nicholas Weininger]

I think the conservative treatment is pretty much what Frum described: risk is the stick that beats us into doing things the Good Old Traditional Way. The emphasis is on risk as a maintainer of good social habits, of the proper community order. The libertarian treatment focuses on risk as a necessary result of freedom. A conservative deplores social engineering to reduce risk because it will lead people to become dissipated, slovenly, immoral. A libertarian deplores such social engineering because it will interfere with the right of free individuals to do as they like with their own property, and frustrate all the innovation and want-satisfaction that proceed from that right.