Statistics question

[davywavy]

You know that figure of 'only 0.7% of the welfare budget is lost to fraud' figure that gets thrown around in comparison to tax avoidance/ evasion?

Does anyone have a citation with the data that's based on, please? I ask because a 0.7% incidence is lower than any statistically significant margin of error I've ever seen and I'd like to look at the original numbers to see how they came by it.

Comments

[matgb]

http://www.theguardian.com/news/datablog/2013/may/13/welfare-fraud-error-universal-credit

It appears to be taking the %age from total expenditure including pensions, and also doesn't include fraudulent underclaiming (which has to be an insane indictment of the complexity and stupidity of the current system).

[(Anonymous)]

Keeps a lot of people in thrall to the state though. That's got to be in someone's interest or it wouldn't happen.

[davywavy.livejournal.com]

yeah, I've looked that over and the DWP numbers it's based on and I see lots of estimates but unless I've missed it I can't see what the basis of those estimates are.

The DWP data says "percentage overpaid", but is that based on the percentage overpaid that they found out about? Or what? If not, what is the basis of those estimates? I genuinely can't see the information I'm looking for.

I recoil from any data which claims a degree of efficiency with a statistical significance >99% against a population size in the millions. Make sense?

[matgb]

Absolutely, I have no doubt that the amount lost has always been fairly small as a percentage and, given how bloody impossible it is to claim anything these days, has shrunk in more recent years. But I doubt any figure is ever going to be accurate, partially because how do you actually measure what people are claiming they shouldn't be? If the people doing the measuring know who's overclaming currently they'd stop the claim.

Of course, I'd scrap the lot and bring in a Basic Income, but that makes me a dangerous radical either too right wing or too left wing depending on who I'm talking to

[davywavy.livejournal.com]

I like the idea of a CBI, but can't see how it's affordable - by my calculations turning the entire welfare/pensions budget into a CBI would end up about £4k pa, which is enough to live on but would mean housing benefit being scrapped.

For both that reason and the entire bureaucracy involved in administering welfare being unnecessary and getting the can - I can't see it getting traction, tbh.

[drplokta]

You get a reasonable GBI by abolishing income tax allowances altogether, and having a flat rate of 40% or 50% on every penny earned in excess of the GBI.

[davywavy.livejournal.com]

Possibly, but who's going to vote for that?

[drplokta]

Everyone who would be better off, which is the majority of people?

[davywavy.livejournal.com]

I'd certainly take that bet. Tenner?

[whiskeylover.livejournal.com]

I would vote for that, and I suspect that if people realised (and believed) that they would be better off, they would vote for it too- however, that would require an intelligent electorate. For that reason, I can't see Dave losing that bet...

[davywavy.livejournal.com]

I don't make bets I think I'd lose :)

I also don't think people would be anything like as better off as I the proposal would suggest.

[(Anonymous)]

A well designed CBI system would incentivise work, which would mean everyone would be better off...

[davywavy.livejournal.com]

If you tax something you end up with less of it than you otherwise would have had.

Taxing work at 50% will result in less work being done.

As such I'm unsure that, overall, society would be wealthier, which wouldn't end up making the majority richer than they otherwise would have been.

But as I say, I'd like to see the numbers the assumption is being based on before passing coming to a conclusion.

[(Anonymous)]

Yeah, I didn't advocate taxing work at all. I think it's immoral.

[drplokta]

But at present we tax lots of work at an effective rate of 95% or more, due to benefit withdrawal. Overall, my proposal wouldn't change effective tax rates, taking benefits into account; it would just move them around.

[davywavy.livejournal.com]

That's rather why I'm asking if you've any links to studies on this (when I was a doley dosser getting my first job essentially involved a pay cut). My instinctive reaction is that it wouldn't work as well as proposed, because that's what tends to happen and I've learned to trust my instincts. However I've also learned that I can be wrong so if an idea sounds interesting I'll take the time to look it over - I've had enough 'well, I didn't expect that' moments over the years,

[davywavy.livejournal.com]

I spent some time googling this this morning and can't find anything. Do you have any links to calculations or numbers? I'm curious.

[matrixmann]

Interesting data.

[davywavy.livejournal.com]

You Germans have some very interesting/good ideas on how to run Welfare. I was looking into it when I was in Berlin.

[matrixmann]

But only if you take a look into times that are already past.
The rest is just a legal way to ensure the right to always be able to get cheap labour.