Wednesday, January 03, 2007

Poverty is absolutely relative

It struck me the other day that the concept of absolute poverty doesn’t make any sense. Poverty is necessarily relative.

Let me explain. Relative poverty, as commonly discussed, refers to being worse off than other people. Absolute poverty, as commonly discussed, refers to having a small amount of actual money. So if everybody suddenly had twice as much money, then relative poverty would remain unchanged but absolute poverty would be hugely reduced.

But this assumes that income increases wouldn’t be inflationary, which they doubtless would. Because if your income goes up by, say, 10%, but the cost of everything goes up 10% as well, you’re no better off.

Now, say Bob has an annual income of £4000. By any reasonable standard, we would class him as being in absolute poverty. But if we’re talking about the year 1907, then suddenly it’s obvious that he’s very rich, because everything was so very much cheaper then.

So poverty can’t possibly be defined as mere amounts of money. It has to take into account what that money can buy. But then you face the question: how do you decide what level of purchasing power is important?

All sorts of reports on changes in poverty rates – by government bodies, think-tanks and research organisations – make the standard distinction between relative poverty (below 60%, say, of median income in each year) and ‘absolute’ poverty. For the latter, they have to move the poverty line for each year to keep up not with earnings growth but with price rises. They typically use 1997 as a baseline to assess changes under this government, or 1979 as a baseline when considering the last government.

But what does it mean to talk about incomes today in comparison with the 1997 median? What do such comparisons tell us? Nothing. If we want to know about the 2007 income distribution, we need the 2007 median. (Alternatively, if we want to know how the poorest have fared over ten years, we need to look at the real income increases for that group and perhaps compare with other, richer groups.)

However we want to make our comparison, we have to set a baseline. And from that point, the relativity creeps in. Perhaps it’s not today’s poor relative to today’s well-off, but today’s poor relative to the poor of a decade ago.

Another source of relativity affects definitions of poverty that are supposedly nothing to do with income comparisons. The US official poverty line is an ‘absolute’ measure, defined in terms of being able to afford the food, shelter and clothing needed to maintain “healthy living”. Bar a few technical adjustments, this definition is unchanged since 1964. The assumptions that underlie it, though, are not: views of what counts as “healthy living” change all the time.

Advances in medicine, particularly diagnostics, influence public standards of healthiness. Shifts in government policy can strengthen or weaken the link between income and access to healthcare, affecting expectations of treatment. And lifestyle trends will have impacts on health standards – typically varying between different income groups.

(Futhermore, well-known research on British civil servants – all well clear of destitution – shows a clear connection between health and socioeconomic status, which of course is essentially relative.)

Even if you ignore comparisons with the median income and focus on the affordability of some basket of goods and services, society still has to take a view on what counts as a minimally tolerable standard of living. And as society changes, so will that view.


Praguetory said...

Would you prefer the UK to be relatively poor compared to other countries but fairly equal within its borders or for the UK to be relatively wealthy but fairly unequal within its borders? This is the choice facing the UK. Attempting to "solve" relative inequalities will lead to the continued disappearance of the most successful sectors, companies and people from the UK.

Praguetory said...

Oh and an ever-expanding group of dependent adults caught in the unemployment trap or some variation on it. There are 8 million economically inactive British adults of working age at this time. That is a disgrace.

Anonymous said...

It's a fair point that all definitions of poverty are relative in some sense. What we regard as a tolerable/dignified standard of living changes over time. And with geography - the minimum that would be tolerated in sub-Saharan Africa is very different to the minimum that would be tolerated in the States.

But there is still a distinction between measuring poverty relative to whatever is judged to be the minimum decent standard at the time, and measuring it relative to the median income. That is what I understand to be the distinction between absolute and relative poverty, not a comparison between something immutable across the ages and continents and something that shifts with every change in the economy.

All definitions of poverty are a statistical minefield, but relative poverty is relatively more elusive than absolute poverty. We talk about the median income, but is that per person or per household? Conventionally, it is per household, though it is not necessarily the business of the government to support ever lower occupancy levels and the fragmentation of families and society that goes with it.

Households range significantly in their needs, both geographically because of differences in the cost of living, and practically in terms of numbers and ages of adults and children in the house, and their circumstances. Should we do the calculation before or after allowing for housing costs, and are there other unavoidable costs (such as energy) that we should also allow for? Statisticians can allow roughly for these factors when converting raw sample data into equivalised income statistics, but it is both more difficult and more important to be accurate if one is trying to do the reverse - working from a statistical average to a real entitlement appropriate to each household.

Currently, 60% of the median household income would be around £11,000 p.a. That is barely viable for a multi-occupant household in London or the South East, but an excessively generous safety net for a single, childless adult in many parts of the North and West.

If this is the minimum that we will tolerate people having to live on, what happens if someone capable of work chooses not to, content with their £11,000 p.a? Will you withdraw the support? How will you distinguish between those who are able but unwilling and those who are willing but unable to work? The huge movement of those out of work from the ranks of the unemployed to the ranks of the incapacitated demonstrates the difficulty of making that judgment. But if you are not prepared to pull the plug on those suspected of abusing the state's generosity, at £11,000, you will find many who judge that unemployment offers the best return for their time.

Ultimately, the proof of the pudding is in the eating, and the effect of nearly ten years of a government committed to reducing relative poverty is that income inequalities have increased. With the best intentions in the world (and I accept that this Government has sincere intentions to tackle poverty), the targeting of relative poverty has not dealt with relative poverty. Because one effect of the sort of state generosity that is needed to address relative poverty is that poverty traps are increased, because generous benefits have to be withdrawn, and the withdrawal creates a very high effectve rate of marginal taxation. It is not critical of the poor to say that this reduces their incentive to work - who would work more if every extra pound you earned contributed only 10 pence to household income?

It is one of the classic cases of "unexpected consequences" that generous measures to tackle poverty can actually make matters worse for the poor. What they need more than state generosity, is the provision of sufficient to get by, and a low effective rate of marginal tax so that whatever they do earn is worth earning. That is the value of targeting absolute poverty rather than relative poverty. It is not about saying people should live on the minimum, but that they should keep more of what they earn.

My suggestion for how to square this circle is to replace most benefits with a Basic Income, and the various income tax rates and NI employee contributions with a flat tax. That would dramatically reduce the poverty traps in the current benefit system, but that's another story. Suffice it to say that, even if a Basic Income is no more than an idealistic thought experiment, it is useful to ask yourself whether it would be feasible and right to set it according to relative or absolute poverty levels. It has to be absolute poverty - relative poverty is simply not practical.

Tom Freeman said...

“Would you prefer the UK to be relatively poor compared to other countries but fairly equal within its borders or for the UK to be relatively wealthy but fairly unequal within its borders?”

If you insist on the trade-off, I’ll be bland and go for ‘somewhere in between’. I don’t think anyone not institutionalised wants total economic equality, but nobody seems to have a bright idea for, say, an optimal value for the gini coefficient.

Personally, I’m much more concerned with helping people at the bottom (and not just by chucking money at them) than restricting people at the top.

Oh, and I’m less interested in how well-off a country is than in how well-off the people who live there are.

As for the basic income idea, there’s definitely something in it, although it’s not without its problems. Set it too low and it’s – well, too low. But the higher a universal benefit gets, the more prohibitively expensive it becomes. Plus, if it’s unconditional, then the higher it gets, the more likely work disincentives are. And if the aim is to relieve destitution, a pity that most of the spending will go to boost the incomes people well away from poverty.

I guess it depends what we want a benefits system to do. If we want to encourage employment then it’s reasonable for JSA to be conditional upon actual job-seeking. And I hardly think it’s obscene that people on IB should go for occasional interviews to see whether there might be a way of enabling them to succeed in some line of work.

It’s true that tax credit withdrawal once in work can create too high marginal tax rates. But the basic income – which has different pros and cons – still, I think, leaves the circle not entirely square…

bgprior said...

If you set means-tested benefits too low then they're too low. And the higher you set them, the more prohibitively expensive they get. That's no argument for choosing one or the other of Basic Income or mean-tested benefits.

The higher means-tested benefits get the greater are the disincentives to work. Not only do those with no earnings have sufficient to live on and therefore may choose not to work (we may both agree that this isn't as common as all that), but because you are means-testing the benefits and therefore withdrawing them, the effective marginal rate of tax is high. The higher the benefit, the greater is the impact of withdrawal and the more people you drag into the trap. When we have effective marginal rates of taxation of 70-100% for many people at the bottom end of the scale, and (however Gordon tries to fiddle the statistics), a very high real rate of unemployment (unless you choose to believe that we are in the middle of a mental-health epidemic of catastrophic proportions), it is hard to argue that a Basic Income could be any greater disincentive to work than the present system. If the effective marginal rate of taxation is the biggest disincentive to work (which it is), a Basic Income is inherently superior to means-tested benefits by this criterion, because of the avoidance of withdrawal.

It is not true that a Basic Income would mostly go to boost the incomes of people well away from poverty. That depends entirely on what rates you set the Basic Income and tax-rate at. The two can be set so that those at the lower end benefit very much more than those at the higher end. Or vice versa. It is not a left vs right issue, it is an inefficient/micro-management vs efficient/laissez-faire issue. I do not remotely advocate this, but as a thought experiment, imagine a Basic Income set at £10,000 and a Flat Tax set at 60% to pay for it. That would be a huge disincentive to work, which is why it's much too high. But it demonstrates the possibility of a Basic Income system that benefited the poor more than the rich.

You can lead a JSA-recipient to interview, but you can't make them work (i.e. they may have to take a job, but they don't have to do anything to hold on to it). And if interviews are such an effective way of getting those on IB back into work, how come their numbers keep growing like Topsy? The only job that is worth having is the one the employee wants, the employer wants, and where the employee keeps a decent share of the proceeds of their labour. This _must_ be a pull-driven exercise from employers and potential employees, not a push-driven exercise by government. The best way to create that pull is to reduce the cost of employment to both employer and employee. That means no or low benefit-withdrawal rates, income tax rates as low as possible, and scrapping the minimum wage, on the basis that a Basic Income ensures that no one has to work for less than they are willing to.

I agree a Basic Income doesn't completely square the circle. Only two solutions do that - full-blown socialism with everyone being paid according to their needs, or (at the other end of the spectrum) anarcho-capitalism, with no welfare provision at all. They don't work, and people don't want them. So it isn't about finding the ideal solution, but the most efficient solution. A Basic Income could not be less efficient than the current system, and has every prospect of being significantly more efficient.