The "cost" bias

Simon's claim that fiscal austerity has been a disaster reminds me of a big bias in the media - the misuse of the word "cost".

Very often, when the press use the word "cost" to describe a policy, what they really mean is a transfer. For example, the BBC says that "offering four weeks paid paternity leave will cost £150m a year". No, it won't. This is a benefit for working fathers but a cost to their employers. This is a transfer, not a cost. Similarly, the Guardian writes: "A cut in the cap on tuition fees would cost just under £2bn." It might cost universities that, but it'll benefit future graduates. Again, it's a transfer - not a pure cost.

Of course, these might well be bad policies. And there might be all sorts of deadweight costs to transfers or dynamic disincentive effects. Once you've estimated these, you can reasonably describe your estimate as a "cost". What you should not do is claim that a transfer the same as a cost.

You might think I'm being pedantic here. I'm not. For one thing, wibble about "cost" can be used to hide the fact that some policies are a positive benefit. For example, in his interview with Natalie Bennett Nick Ferrari spoke of the "cost" of building houses. Ms Bennett should have replied thus:

There is no cost. It's a net benefit. The government can borrow at 2% but charge rent on the completed houses of 4-5% and have a capital asset to boot. If housebuilding is a cost, how come Bovis and Persimmon this week reported big profits? What's more, there are benefits to future renters who get somewhere to live, and to currently unemployed workers who'll get jobs. The only losers are existing home-owners and landlords who'll see lower gains in future because houses will be less scarce. Net, it's highly likey that housebuilding will be a benefit.

There is, though, something even more pernicious going on, which is where Simon comes in. When he estimates that austerity has cost 5% of GDP, he's using cost in the proper sense - of something that leaves us all worse off than we'd otherwise be. Similarly, when he claims that leaving the EU will cost at least 2.2% of GDP, John van Reenen is using the word properly - though you might disagree on his estimate.

The sums here are huge: 5% of GDP is £85bn - vastly more than the few hundreds of millions "cost" of some Labour policies.

And yet most of the media is much quieter about these genuine costs than they are about the spurious "cost" of other policies. Why?

It's not just because of an explicit Tory bias. Even the BBC talks about cost in the wrong sense. I suspect that two other things are going on.

One is a form of certainty effect. There is rough bi-partisan agreement upon the Exchequer cost of transfers to fathers, students or whoever whereas Simon's estimate lacks such consensus. In the world of "due impartiality", the truth is defined as "what people agree about" rather than what is actually the case. And so the word "cost" is misused.

Secondly, there's the fallacy of the "nation's finances"; politicians and the media routinely use this phrase to mean something utterly different - the government's finances. If you commit this error, then you'll believe the cost to the Exchequer is a cost to the nation - when it is no such thing.
Whatever the origins of the error, though, it is a deeply pernicious one. It tends to present any party wanting to increase borrowing as profligate - as imposing costs upon us. But in fact, the true costs are imposed by the parties that have trashed the economy.   

* The error I'm describing isn't confined to the right. Leftists also make it when they talk about the "cost" of tax-dodging. This too is a transfer - from the Exchequer to tax-dodgers.

Comments

[Pension60]

...Again, it's a transfer - not a pure cost. ...

Oh please, would you care to tell The Greens' Leader Ms Bennett and their 1 MP Ms Caroline Lucas of this?

The Greens unique lost universal and automatic Citizen Income
would have replaced all the massive welfare admin costs
and the ever reducing benefit to the starving entirely.

The Greens unique lost Full Citizen State Pension,
irregardless of NI history
would have replaced the 1 million of 60-64 age group
off benefit and granted a full instead of the tiny pro rata state pensions or even nil coming from 2016.

And helped women especially facing nil state pension for life.
See why, under my petition, in my WHY IS THIS IMPORTANT section, at:
https://you.38degrees.org.uk/petitions/state-pension-at-60-now

All the money now wasted by the hundreds of billions, both state and private contract, in welfare admin, would pour into the economy from being in the hands of the people entirely,
with absolute minimal admin, already within budgets
(The NI moved from Social Security department to Inland Revenue long ago).

So both policies would be entirely funding neutral.

There is no national debt (OECD says UK debt is only world average).

There is a £30 billion surplus in the National Insurance Fund since 2013 (source House of Lords Library).

www.anastasia-england.me.uk

[Luis Enrique]

OK but cost as in profits = revenues - costs is also a correct usage. It's valid to ask how much something will cost, even if you expect it to be profitable, this doesn't affect your point that the smarter answer would have ignored that and emphasized the net gain

[Luis Enrique]

also I can foresee some problems if the media got into the habit of reporting costs measured against unobserved counterfactuals

[Matt Moore]

I support labelling the money amount spent as a "transfer", since this highlights that it has been taken from one set of voters and given to another, rather than the government paying out of some magical government-produced money pot. But I think people generally understand what is meant.

More concerning is that people don't have any idea of the true cost of these programmes. Imagine if the BBC had to say something like: "Of course, transferring £4bn under the UK's tax and benefit system costs £1bn in foregone GDP / output, due to the deadweight loss of the taxes, the cost of administration, and the cost in disincentive effects for the recipients." (Not real data, obvs)

Would you support this in further interest of transparency?

[Iain]

I can accept the idea of a transfer as you describe it. But I think you'd need to accommodate the area of voluntary/involuntary transfer and policy implications.
Tax avoidance or evasion transfers money but it "costs" the state resources, which reduces the ability to deliver policy goals. And it substitutes private choices for public ones. That may "cost" all citizens in the form of poorer/more costly public services.

[Tim Worstall]

"For example, the BBC says that "offering four weeks paid paternity leave will cost £150m a year". No, it won't. This is a benefit for working fathers but a cost to their employers. This is a transfer, not a cost."

Not really, it's a transfer from all taxpayers to new fathers. At least I assume it is, if it works like maternity pay. When doing your PAYE form as an employer you calculate what should be paid over to HMRC from with holding. Then you look at what you've paid out in maternity pay and deduct 90% of that from what you've got to send HMRC. Employers only pay 10% of it....

[politics that work]

I knew a family growing up with three sons. Every year and Christmas, each of the sons would give each of the others a $20 bill. It was kind of a tradition. Then, one year, they decided to use $100 bills and the mom got all upset because they were not rich and couldn't afford $100 presents...

One example of this problem that drives me nuts in the US is that the healthcare debate treats government healthcare spending as a cost. What it really is is a different mechanism for the spending. If we increased taxes by $1 and government spent that $1 on our healthcare, and that means we can pay $1 less for private insurance, then that isn't a cost. Now, there certainly are legitimate arguments to be made on all sides about which mechanism is better, but that debate seems to take a back seat to the empty ranting about whether we can "afford" it or whether it is worth the cost.

[Roy Lonergan]

@Tim Worstall

If we're going to concentrate on the trees & not the wood, employers can usually reclaim 92% of employees’ Statutory Maternity (SMP), Paternity and Adoption Pay. But It's 103% if you qualify for Small Employers’ Relief. Still a transfer.

Roy

[Roy Lonergan]

And the maximum that can be reclaimed is the lower of £138.18 a week or 90% of their average weekly earnings. So, if you have an employer, like mine, that gives two weeks of paternity leave on full pay it's only a small fraction that doesn't come as a transfer from the employer.

[Kaleberg]

So if I buy a candy bar, it's actually a transfer. It costs me some money, but I get a candy bar in return? Naturally, I wouldn't buy the candy bar unless I got some net benefit in return. Maybe this is a bit pedantic, but it will add even more pleasure the next time I arrange for a transfer and get a candy bar in return.

[Roy Lonergan]

@kaleberg

Transfer if it's taking candy from baby A and giving candy to baby B. But when it's not like taking candy from a baby then you probably just bought it.

[Jamie]

"M-C-M and C-M-C+"

I think I got my Marxist terminology wrong. It should have been

C-M-C and M-C-M+

[Boursin]

My mother tongue (Finnish) uses the same verb, "maksaa", for both 'to cost' and 'to pay'. Just as with the German word "Schuld" meaning both 'debt' and 'guilt', this kind of thing probably either reflects cultural differences in economic thinking, or is at least liable to influence them.

[theOnlySanePersonOnPlanetEarth]

Roy - Tim Worstall isn't interested in the truth.

[Roy Lonergan]

OnlySane

I think that's probably a little unfair to Worstall. Polemical and seems to have a few unhealthy obsessions but he often makes interesting points.

Roy

[Bayard]

"The only losers are existing home-owners and landlords who'll see lower gains in future because houses will be less scarce."

Surprisingly, scarcity has very little effect on house prices or rents, so home-owners will not lose out, nor will most landlords. The ones who will lose out are those who are currently renting to those on housing benefit as these are the most likely future occupants of the new council houses.