Wednesday, December 22, 2010

Thinking rationally about unemployment

This (greatly) extends a comment I made here.

One of the dispiriting things about observing the US's economic problems from afar (specifically, a country that managed to avoid both the Global Financial Crisis and the Great Recession) is watching old and tired debates about unemployment getting another whirl. This piece from Salon provides examples of people blaming the unemployed for unemployment.

This all so "old" for me. My first job was in the (since abolished) Commonwealth Employment Service dealing with unemployed people in the 1982-3 recession. (I had never had a job before, so naturally the Australian Public Service thought the best use of my talents was advising folk on their job prospects.) I later worked in various labour market economics/stats areas. I am so familiar with all these arguments, and I find the “blaming unemployment on the unemployed” approach utterly tedious.

The basic three operating principles are:

(1) Unemployment shifts dramatically because of changes in economic conditions.

(2) Long-term trends in unemployment occur because of institutional (particularly regulatory) factors.

(3) There are sorting processes about who gets and stays unemployed, but this does not change (1) and (2).

Really, it is not so hard.

Unemployment in Australia is much lower than in the US or UK. (In my local suburb, most of the fast food places and cafes have "help wanted" ads in the window.) This is not because suddenly Australians are harder working than Yanks or Poms. It is because:

(1) our central bank, the Reserve Bank, handled monetary policy better than the Bank of England and (especially) the US Federal Reserve;

(2) our prudential regulation of our financial sector worked a lot better (and did not have to deal with a collapsing housing price bubble); and

(3) the Australian public debt position is much stronger [which improved both confidence generally and policy flexibility in particular].

This, added to a much more flexible economy due to almost three decades of economic liberalisation meant that we even manage to weather a dramatic drop in commodity prices.

Just as Europe needs to face the fact that its entrenched unemployment is largely due to the way it regulates its labour markets, the US needs to face the fact that bad decisions by the Federal Reserve is the most important reason for its massive levels of unemployment. Not any change in behaviour by the unemployed.

For if wage contracts are set according to certain expected trends in the value of money (e.g. around the 2% inflation the US has had for a couple of decades) and there is a sudden, unexpected increase in the value of money (i.e. the scarcity of money increases because the Fed suddenly adopts a tight monetary policy to get even lower, possibly zero, inflation) then there are suddenly a whole of labour contracts where the real cost of labour increases unexpectedly. Hiring plummets, firings surge, economic activity drops and you have a dramatic economic downturn and a massive increase in unemployment. (Now think what a sudden increase in the scarcity of money would do to highly leveraged financial institutions: Irving Fisher described this scenario back in 1933 [pdf] in his debt-deflation theory of the Great Depression of the 1930s.)

Which is not remotely the fault of the unemployed. Yes, of course interventions (such as extending unemployment insurance) affect trend levels of unemployment. Yes, of course there is a sorting process about who gets and stays unemployed. But that sorting process does not affect the level of unemployment: it occurs within a given level of unemployment. Which is a product of economic conditions (including public policy, especially monetary policy) and institutional structures (particularly labour market regulation and other interventions). Moralistic nonsense blaming the unemployed is precisely that.

Blame the Fed, they did it. Again:
Let me end my talk by abusing slightly my status as an official representative of the Federal Reserve. I would like to say to Milton and Anna: Regarding the Great Depression. You're right, we did it. We're very sorry. But thanks to you, we won't do it again.
Spoke too soon. But really, really the consequence of greatly increased unemployment is not the fault of the unemployed.

ADDENDA: On the US unemployment situation, this graph is deeply depressing. It comes from this highly informative post on the diabolical state of much of the US housing and housing finance market. Though this is not a good sign regarding Australian housing finance.


  1. I think the US tendency to blame the unemployed/poor for being so (and the concomitant tendency to punish them for it) is a side-effect of the Great American Dream that anyone can succeed if they try hard enough. It's a simple and obvious corollary that if you don't succeed, then you didn't try hard enough: which is your fault.

    Add to this the insidious "Prosperity Theology", where God intervenes in the world and rewards the Faithful on Earth. Again, the obvious corollary is that He punishes the wicked, and poverty becomes a moral failing: it is of itself evidence that one is both lazy (American Dream) and unworthy (Prosperity Theology), and the only way one can prove that one deserves aid in one's poverty is by no longer needing it.

  2. Yes, I agree, these things make the tendency more intense in the US. Though it is not as if there were no traces of the same "blaming the unemployed" in past Australian commentary, for example.

    One has to be careful, however, because behaviour does make a difference. If you complete high school, get and stay married and get a job, even at the minimum wage, your chances of staying poor in the US are very low. Hence my point about sorting processes.

  3. Indeed there are similar mindsets at work in Aus. I see them behind much of WorkChoices, for example.

    One should be careful not to give individual behaviour a power it does not have. Don't forget the insidious effects of social feedback: If you're from a poor family, then largely by virtue of that alone you are far less likely to complete high school, stay married or get/keep a job. And that's not just the US, that's a universal thing. That's how the Undeserving Poor class is perpetuated everywhere. It is really, really hard to fight one's way out of that tarpit. But without social safety nets, and with the aforementioned force of social opprobrium, internalised as well as imposed, the escape velocity is far higher in the US than in Aus.

    There are exceptions to this force, but that's the point: they're exceptions who left the ghetto and made good, and noteworthy for that fact. It is because almost all of their peers will live hard and often short lives that their freakishly unlikely success is so newsworthy.

  4. WorkChoices is not a good example: there was plenty problematic about it, but it was aimed at lowering the barriers to entering employment.

    It is also a myth that the US poverty performance is somehow much worse than that of Australia or other Western countries. The US poor are about as poor as the Swedish poor, for example, and about as proportionately numerous. The US has a lot more rich people, that is where its high rates of inequality show up.

    Moreover, there was a massive drop in poverty rates in the US from about 1945 to about 1965. Poverty rates since then have been more "stuck" but what constitutes poverty also changes. For example, poor people in today's US have levels of ownership of consumer goods that would have made them easily middle class in 1973.

    It is also true, however, that changing patterns of behaviour need not be looked at as a purely individual issue and there is still the issue of interactions with the actual incentives welfare (or, for that matter, government schooling) generates. One of the successes of welfare reform in the US is that it massively reduced welfare rolls without any adverse affect on poverty: indeed, there were some declines in poverty rates. One of the reasons California's fiscal position is so awful is that it refused to adopt the reforms. As a result, a State with 12% of the US population has 30% of the total national population on welfare rolls.

  5. Sweden? Why Sweden?

    Could it be to do with this? Because that was written in 2004, using data from 2001, and a lot of it refers to things which don't make the knock down points they think they do. (DVD players are cheap: it's not that poor people have more money, but that gadgets are much, much cheaper. And the figures they show for home ownership... that was in the middle of record cheap credit and a growing housing bubble. How do you think that's going now?)

    Or maybe there's this. To which this response seems to say what needs to be said.

    I wonder how much of the massive drop can be attributed to the Post-War boom. I might also point out that definitions of poverty have been under constant revision since then, to the point that one can be officially not poor, and at the same time not able to feed one's family.

  6. Actually, no, it was because I have seen OECD figures on incomes of lowest decile across OECD countries and the US-Sweden comparison was the one which stuck in my mind.

    Most of the drop can be attributed to the post-war boom: though that was a bit interactive, since changes in circumstances also contributed to the boom. The interesting thing was the movement out of poverty stopped before the boom did and has been resistant to much improvement since, regardless of economic conditions. LBJ's "Great Society" seems to have been counter-productive in that regard.

    The US poverty index has its problems, but not the one you are claiming.

  7. Actually, that describes exactly the problem I am claiming: that the official Poverty Line is, in fact, a fraction of the level below which life is a daily struggle, where there is no day-to-day security.

    “For one thing, as antipoverty advocates point out, since 1955 the proportion of family budgets devoted to food has fallen from one-third to one-fifth.”

    So for a meaningful approximation to the original concept (rather than going to great effort to miss the point), you'd have to multiply the Poverty Line by 5/3, which would in one stroke move the Poverty Line in the US for a family of four in 2009 from $21,954 to $36,590.

    Moreover, this function isn't the same across counties: Swedish poor people probably don't have the same 1/5th food/cost of living ratio as the US poor do, not least because they don't have to account for outrageous medical bills. Yay for socialised medicine. The poor in Sweden aren't as poor as the poor in the US because they are subsidised by the state. So to compare the rates so naively is apples and oranges.

    “The interesting thing was the movement out of poverty stopped before the boom did and has been resistant to much improvement since, regardless of economic conditions.”

    You know, that is interesting. Especially since from a quick look at the graphs here, there is very little visible relationship between poverty rates and recessions. There must be other economic and social forces behind what's going on. I can't really tell from that if there is anything which obviously is impacting, but I will note that the participation rate would seem to be a necessary addition to the total workforce size graphs to be meaningful in the context of poverty. (I know; teaching my grandmother to suck eggs.)

  8. People are poorer/more people are poor because food has got cheaper? That is too sophisticated and nuanced for me.

    I take it as read that poverty is meant to indicate struggle and a lack of security.

    The graph I remember was comparing incomes based on purchasing price parity, not poverty rates as such. Hence my comment that "US poor are about as poor as the Swedish poor". If you are poor in the US, you are likely to be covered by Medicaid, so the picture on health costs is complex.

  9. People are poorer/more people are poor because food has got cheaper? That is too sophisticated and nuanced for me.

    How about “People are poorer because the costs of the other necessities of life are relatively more expensive than they were.”

    Power, heating, water, rent/repayments on housing loan they should never have qualified for, telephony (a new requirement, but something which has become more necessary for anyone with a desire to become not-poor), transport, childcare, &c &c &c. That's just what I can think of.

    And as for Medicaid: “Medicaid does not provide medical assistance for all poor persons. Even under the broadest provisions of the Federal statute (except for emergency services for certain persons), the Medicaid program does not provide health care services, even for very poor persons, unless they are in one of the designated eligibility groups.
    It's much more restricted, and much more politicised than you may be thinking if you have Aus' Medicare in mind. And as well, much more bureaucratic, and with far greater stigma to its use. (One of those social forces again.)

  10. Relatively more expensive is not more expensive. In fact, most of those things have become cheaper over time, it is just that food got cheaper quicker. It still looks like "food got cheaper so poverty went up". The argument seems to be more that expectations have increased so what constitutes economic participation has expanded. Which begins to look like a way of penalising US society for being more successful and capable.

  11. "food got cheaper so poverty went up"

    Have the poverty rates decreased to the same extent as have relative food prices?

    Because... think about that. Food is cheap. Cheap as it ever has been in human history. The poverty line is literally defined on a really complicated and at the same time naive equation: how much it costs to buy enough food to eat multiplied by three. And yet the poverty level is still about 14%. Why is it, along with record low food prices, that there is not record low poverty?

    And, yes, poverty is relative. That's the whole reason for the poverty line being defined in terms of the local economy, yes?

  12. I have no problem with expanding the notion of poverty, understood in relative terms, to take into account expanding notions of economic participation. I am just deeply sceptical of any argument which claims that things getting cheaper makes poverty worse.

  13. If there really is any danger in the Oz housing market, let's pray we have the policy infrastructure already prepared, to make it as smooth as possible.

  14. Peter: that is a nice hope, but Swan offering to extend government guarantees to the "fifth pillar" or mortgage providers does not fill me with confidence.

  15. Lorenzo, yes, their policy implementation is even worse than its formulation. let's be honest: this mob could not organize a piss up in a brewery! :)

  16. Here, let me torch your straw man for you: The argument is not that "things getting cheaper makes poverty worse" as you pose it. It's that "using unrepresentative metrics by which to calculate the poverty line is leads to inaccurate measurements of poverty rates".

    I'll explain my understanding of this to you in very bare and simple terms, devoid of sophistication and nuance.

    Let's say that it used to cost three shillin's to buy food for a person for a week. Rent cost three shillin's. It also cost a shillin' to buy a weekly transport pass and a shillin' to heat and light one's humble abode.

    So to survive an uneventful week cost eight shillin's. To permit a person some leeway to be able to afford luxuries like medicine and shoes, we add another shillin' and call nine shillin's a week the "poverty line", but instead of adding it up like that, we go for the lazy shorthand of saying "three times wot their food costs".

    Time passes, food gets cheaper such that it is possible to buy food for a week for two shillin's. Rent gets expensive and now costs five shillin's. It also costs one and a half shillin's to buy a weekly transport pass, one and a half shillin's to pay the power and gas bills, not to mention that the tele-o-phone has become ubiquitous, and that's another half-a-shillin'. Shoes and medicine and other such fripperies take another shillin'.

    But the metric by which poverty is measured is more affordable. Food for a week costs two shillin's, and three times "cost of food for a week" is now a mere six shillin's. There are precious few people surviving in that bracket, so poverty has been reduced! Clearly people with almost twice that much money will be thriving!

    However, the fact is that people who are spending eleven and a half shillin's to maintain the same standard of living that nine shillin's used to pay for are still living in poverty. They just aren't counted as living in poverty because the metric by which poverty is calculated is not representative of all living costs.


    Your ability to find sources is better than mine, so I will express a keen wish to see whether the cost of food and the cost of rent and utilities have kept pace over the last few decades.

    I'd also be interested to know whether the (OECD) rates of people dying from hypothermia in their homes (or on the streets, or under bridges, for that matter) has increased while the number of deaths in from starvation has fallen in the last few decades.

    I'm betting they have.

  17. Your argument depends on other things getting more expensive. If basic cost of living items get more expensive, relative to income, then poverty gets worse. I have no problem with that claim, provided they have actually got more expensive, relative to income. Which is an empirical claim.

    In the US, for example, rents have risen relative to income in some places (e.g. California) but not others (e.g. Texas). I have posted on this and the reasons for it.

    Utility and transport costs, well you tell me. Since California capped power prices, for example, (and banned new power construction: utterly predictably, they had power shortages) I guess not in California. But cheaper food implies, amongst other things, that transport costs probably haven't gone up much.

    But here's another indicator: poor households level of ownership of consumer durables have increased significantly -- to levels that would have made them comfortably "middle class" 3 decades ago. Which suggests that what has happened is that, as food got cheaper, purchase of other things expanded.

    I agree that taking a simple multiple of food costs is a lazy way to measure poverty. It does not, however, follow that upping the multiple will be a good measure of poverty either.

  18. Yes. My argument depends on other things getting more expensive relative to the cost of food, which you keep repeating has been getting cheaper.

    Have basic cost of living expenses risen or fallen compared to low level incomes? Where and for whom? And what constitutes a basic living expense?

    I'm confident that I am paying more, as a portion of my income, for utilities, fuel, building materials and tradespeople regardless of food costs.

    Speaking of which: How closely is food transport linked to public transport costs in a city network which carries people exclusively?

    Also; I'm betting that people living under the poverty line thirty years ago had levels of consumer durables which would have made them comfortably "middle class" three decades before. And those of sixty years before that would be middle class by the standards of 90 years ago. I'm not sure what your point is, there.

    I have no idea who suggested that upping the multiple of food costs to estimate the poverty line would be effective. Certainly not me. The price of food is not a good measure of overall cost of living. Something akin to the Henderson poverty line which calculates the point at which income falls below the cost of living by (radical concept) estimating the actual cost of living instead of extrapolating from one factor.

    Unfortunately, the Henderson poverty line loses me at the point it suggests that $170/week will pay for accommodation for an unemployed sole parent with three dependent children. Feasible in some rural cities here, but poor people live in large cities, too, and... have you seen rent prices in Melbourne lately?

    I'm not suggesting that rent is the one true cost by which poverty can be measured, either, but the reality of that cost should be taken into account, not extrapolated from the cost of food.

  19. I don't think you can really mean to suggest that reduced food prices are more than vaguely linked to urban passenger-only transport network ticket prices.

    I did not suggest that upping the multiple is a useful way of measuring poverty. Just that using any one expense as the basis for estimating the total of all expenses is not going to provide an accurate result when the expenses in question vary wildly with regard to oneanother.

    The Henderson poverty index attempts to estimate basic living costs, (but my confidence in this measure is discouraged by the suggestion that sole parents with three dependent children will be able to find adequate housing for $170 a week) which seems to be a sane way to go about it if the estimates reflect reality.

  20. I don't think you can really mean to suggest that reduced food prices are more than vaguely linked to urban passenger-only transport network ticket prices. No, because they are not the only transport costs, even for the poor.

    Poverty indices based on living costs suffer the same problem that price indices do generally, aggravated by the fact that that they are (or should be) a basket of items. With both baskets AND price indices having mounting problems over time.