« Britblog Roundup # 75 | Main | Sgt. Mom »

July 23, 2006

Help Wanted

I’m not very good at searching through the various statistical databases out there so I wonder if I could ask for a little help? Perhaps Jim? (who knows all about these poverty statistics) Or Chris? The Angry Economist?

What I’m interested in doing is trying to compare the US and UK poverty rates. In the US this is normally shown as the number below the federal poverty line. This is where the "37 million in poverty" number comes from for 2004.

As long as people understand what the number is, it’s fine. It is before tax credits (the EITC), before tax itself, before food stamps, housing vouchers or Medicare (or any other non cash assistance).

It is, therefore, a measure more of those who require help to not be poor than anything else.

Here in the UK I think I’m right in saying that poverty is measured the other way, as in those still in poverty after the influence of the tax and benefit system (on a household size adjusted basis).

It is also true that the very definition of poverty is different. In the US it is one of absolute poverty (ie, it’s been upgraded for CPI inflation but not wage inflation over the years) and stands at a cash value of $19k and a bit for a family of four. Given that median household income is around the $40k level, this is close enough for my purposes to 50% of median income.

Here in the UK we say poverty is (after adjustments and the tax and benefit system) less than 60% of median income. However, there is a measure in use, set at 1999 (I think?) incomes, of absolute poverty, which is 50% of median income. This, again, I think, is calculated before the influence of the tax and benefit system.

Now, what I’m interested in is trying to compare like with like. Both US v UK poverty using the US system (absolute poverty, before amelioration and I have the US figures) and US and UK poverty after amelioration using the UK 60% of median relative figures. I’ve seen some US numbers which have sorta similar ideas but they’re still not quite right.

So, here’s the questions.

Can someone point me to those UK stats which give the 50% of median in 1999 figures before amelioration? I know they’re there somewhere but I can’t find them today (having spent some hours looking).

Can someone point me to US poverty figures using the same definitions as UK ones? Adjusted for household size, after the influence of the tax and benefit systems, less than 60% of median income?

Ta in advance!

July 23, 2006 in Economics | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Help Wanted:


US household income for two parents with two children is not $40,000, it's more like $60-$65,000.

Anyway this paper might help in some way.

Posted by: Matthew | Jul 23, 2006 8:42:54 PM


I'm responding to the second request. These pages won't, I think, accomplish all three adjustments simultaneously, but they can get you part of the way there:




Posted by: Craig Newmark | Jul 23, 2006 10:15:07 PM


This won't help you at all but i have to get it off my chest.

The U.K.'s definition of poverty is households that earn less than 60% of the median income. In other words, the UK doesn't measure poverty at all. What it does measure is the 'income inequality gap'.

A household (of four) earning up to $25,000 per annum is described as 'living in poverty'. Try telling that to a Sudanese family with a straight face.

Two examples show the absurdity of using a relative measure of poverty.

i) If Bill Gates were to decide to emigrate to Blighty, poverty would significantly increase.

ii) The government has a target of reducing poverty by 700,000 families. By far the most effective way to achieve this target is to engineer a recession such that umemployment increases and incomes fall.

Sorry to be of no help whatsoever but i feel better now.

Posted by: pommygranate | Jul 23, 2006 11:54:26 PM

I think you're wrong. If Bill Gates emigrated to Britain the "poverty line" would adjust by no more than if Tim moved back to Britain, ie by something like 1/100th of 1p. This would not be a 'significant increase'.

Tim adds: Bill Gates? Couple of quid I would have thought. Leave his wealth out of it (income is what counts here). He gets $100 million a year from Microsoft dividends alone. Add in his other investments to double that? 100 million quid a year then. 60 million in the UK, looks like someone coming in making that sort of sum would raise mean income at just under two quid doesn’t it?

Not sure what it would do to median though. Probably very little as Matthew says.

Posted by: Matthew | Jul 24, 2006 8:26:18 AM

[Not sure what it would do to median though]

You would add one household whose income was above the median, so the arrow pointing to the median would move from household number 22,345,678 to household 22,345,679 and the median income would thereby tick up by that increment. That's why Matthew said the effect would be the same as if Tim moved back to Britain.

Posted by: dsquared | Jul 24, 2006 9:51:58 AM

ok pedants - so i shouldnt have added the word 'significantly'. the point remains though.

Posted by: pommygranate | Jul 24, 2006 11:14:35 AM

I don't think it's pedantic. The measure has been explicitly designed so such a thing doesn't affect it, so to say it does and venture that as a criticism is quite strange.

Posted by: Matthew | Jul 24, 2006 11:17:59 AM

1st request - perhaps in this document?


Or via here: http://www.statistics.gov.uk/CCI/SearchRes.asp?term=household+income&x=33&y=16

There should be some kind of comparative study like this already done, but I don't know where yet, and haven't got time to find out. Work is very busy these days. Yes - I'm in the public sector and it does happen occasionally!

Posted by: angry economist | Jul 24, 2006 2:07:59 PM

I do know that the measure moved from 50% of mean income to 60% of median as the value is typically the same but the newer measure has less variance.

Posted by: David Gillies | Jul 24, 2006 5:09:15 PM

I had a discussion about this on my blog the other day, from various sources we deduced;

Figures for median household income are
(US $44,389 gross - 2004),
(UK $44,400 gross ($1.85 exchange) 2004)

US and UK taxes on income are roughly similar for the median earner. State income taxes vary from 5% to 9.5% - this is on top of federal tax rates of 10%,15% and 25% (compared to 10%,22% in UK). Property taxes are similar to UK as well. In New York the average is $1,900 p.a, the UK council tax average is around $1,850. This leaves;

(US $30k net for median household)
(UK $33k net for median household)

Remember that the US median earner also has to find $400+ a month for health insurance, as they get no NHS for the taxes they pay and also lose out by not getting many other services that are provided in the UK through taxation!

The US median earner should be much better off than the UK median earner but because of the level of inequality there, they are not!

The US has the advantages of a much bigger and more integrated market (economies of scale), they also have many more resources in terms of land and raw materials and are net exporters of food. The UK has a much higher density of population and has to import most of it's food. Hence the difference in land and food prices between the US and UK.

Posted by: Neil Harding | Jul 24, 2006 8:09:52 PM