The price elasticity of demand for rented accommodation

From The Evening Standard:

Charlie, 23, is studying for a masters in public policy at UCL. He spent his first year in London living in a poky two-bedroom flat in Battersea with two friends; they converted the living room into a third bedroom. “It meant we could live somewhere far closer to the centre,” he explains, “but it was horribly cramped. The flat was small anyway but it meant that when we had people around, we didn’t have anywhere to host them. There was always a queue for the shower and it was particularly busy when all our girlfriends stayed on the same evening.”

Charlie is part of Generation Rent —twentysomethings who bounce from lease to lease, squishing extra housemates into box rooms and sacrificing communal space for a Zone 2 postcode. It’s called “hutching up” and it’s exploding in the capital as young graduates flock to London to work but find starting salaries stretched to cover astronomical rents…

“Our surveys of tenants found that affordability is a key consideration but second to that is proximity to place of work or education,” explains Hudson. “So they’ll move down the line or bus route.”

The article is a long list of similar anecdotal evidence, but big picture wise it is true and this is how it works.

It is tenants competing against each other for the more convenient locations who dictate rent levels, the landlords just sit there and collect what they can get.

And tenants are very price sensitive. Price sensitivity is dictated by various factors, particularly relevant are items 5 to 9 on this list, all of which indicate that the demand for rented accommodation is price elastic.

So we observe that tenants respond to high rents by occupying less and less space, so the proportion of income spent on rent is not directly related to local rents in absolute terms - i.e. if the rent for a flat in London is twice as high as in Birmingham, London tenants do not spend twice as much per head - the tenant on a reasonable wage in Birmingham will rent the whole flat for himself or share with one other; the London tenant on a reasonable wage will share with two or three others.

All of which leads us to the conclusion that landlords cannot just "pass on" an increase in their costs or taxes on their income; they bear those entirely out of the rent, over which they have no influence whatsoever.

Where did they get the 65% withdrawal rate from anyway?

1. The suggested rate for the 'Single Unified Taper' has crept up over the past year:

Iain Duncan Smith's Centre for Social Justice suggested a maximum withdrawal rate of 55 per cent in September 2009 (page 303 of this pdf)

According to the Daily Mail a month ago"A withdrawal rate of between 60 and 65 per cent is being debated, according to sources."

The actual White Paper of last week says 65%.

2. These rates are misleading as they apply after PAYE is deducted, thus the effective rates after tax would be 69%, 73% and 76%. As I keep pointing out, 50% is the ideal rate because the administration would be very simple - it could be done via the tax system using K-codes for PAYE, i.e. a flat 50% of wages are deducted in lieu of income tax and National Insurance with no personal allowance.

What would the wider impact of this be?

3. Let's focus on people who earn the National Minimum Wage of £6 an hour (in round figures) of whom there are 747,000. This paper says that the price elasticity of the supply or labour in The Netherlands is 0.1 for men and 0.5 for women, which makes sense, because more men are in work than women, so it is easier to tempt relatively more women into work than more men. So let's assume that the price elasticity of the supply of labour is 0.3. What this means is that if wages offered increase by 10%, the number of people willing to work for that wage increases by 0.3 x 10% = 3%.

4. I explained in my previous post that with a 65% withdrawal rate, a worker also liable to PAYE keeps 23.8 pence for every £1 he earns if he earns above the income tax threshold, and HMRC and DWP between them take 90 pence. If we did the means testing using K-codes, the worker keeps 50 pence and HMRC takes 63.8 pence.

5. So let's assume those 747,000 people all work about 16 or 17 hours a week and earn £100 gross (to keep the numbers simple) and keep £23.80 net. If their net wages went up to £50 for the same hours and the same work, that's an increase of 110%, so we'd expect the number of people prepared to do such jobs to go up by 33%, so an extra 247,000 people would be prepared to take such jobs. This is perfectly plausible, there are about 8 million working age adults not in work in the UK, and the DWP claim that the Universal Credit system would reduce the number of workless households by "up to 300,000" (page 50).

6. Under DWP's plans, the amount of tax and benefit withdrawal that HMRC/DWP will be taking from those 747,000 people would be 747,000 x £90 a week = £67 million a week. If we reduced the Marginal Deduction Rate from 90 pence to 63.8 pence, then HMRC would be taking £63.80 a week from 994,000 people = £63 million a week, a fall of £4 million.*

7. Remember also that the DWP and HMRC between them are so staggeringly inefficient that it costs them about £34 per household per week to process benefit claims (including fraud and error), see earlier post. I think it's reasonable assume that my system, by which all means testing and benefit withdrawal is done at source via PAYE codes would get this cost down by at least a tenth (I would hope that it would get the admin costs down by about nine-tenths, but you can't budget for miracles).

8. So HMRC/DWP would 'lose' £4 million a week in PAYE and benefits clawed back, but the cost of administering the benefits for those 994,000 people also goes down by a tenth, call it £4 a week each, which will save HMRC/DWP £4 million a week.

9. But the most important thing about this is that instead of 747,000 people taking home £23.80 a week, we'd have 994,000 people taking home £50 a week. The total disposable income of those one million people would increase from £18 million a week to £50 million a week. On average, those people at the overlap between being welfare claimants and being workers would be £32 a week better off. That's £1.5 billion a year, which is still only one-tenth of one per cent of GDP, so not an unreasonable estimate (but well within the range of measuring error).

10. It's not going to get us out of recession at a stroke or anything, but it will make it absolutely clear that you are always better off working, so no need for these authoritarian gimmicks like forcing longer term claimants to sweep the streets every now and then. There are people who are happy to sweep the streets for whatever wage the council pays them, why not leave them to it?

Here endeth.

* For all these calculations, I've assumed that the workers concerned earn over the personal allowance for PAYE. Maybe they don't, in which case their net wages would go up from £35 a week to £50 a week, and the number of additional people prepared to do minimum wage jobs would only go up by 96,000. In this scenario, the DWP/HMRC would 'lose' £3 million a week and these 843,000 very low paid workers would end up collectively £16 million a week better off. Certainly still worth doing, because all that £16 million will go back into the economy as spending or rent, so the extra taxes generated will be at least £3 million.

There is a more pessimistic view, that the employers who currently pay £7 or £8 an hour and know that the bulk of their workforce claim Working Tax Credits would simply drop the wages they pay to soak up some of the tax saving. There will be some element of this, but not enough to distort the whole picture; in any event, the best thing we can do for employers (apart from scrapping VAT) is to reduce Employer's National Insurance and all the stupid employment regulations, which would benefit employers in particular but workers in general; is it so terrible if we adopt a measure which benefits workers in particular and employers in general?

Google search "effect of tax on unemployed"

Somebody stumbled across my blog via the above search, not that I can track down where I was on the list.

Two things I can offer up in evidence, if he or she ever comes back ...

1) "Families and work" by Annette Walling, ONS, 2005. To summarise, married/cohabiting mothers whose partner is in work, who by and large are not entitled to benefits, and thus face an average tax/benefit withdrawal rate of 25% up to an average salary of £25,000 have an employment rate of 80%. Married/cohabiting mothers whose partner is unemployed, who are entitled to fairly chunky benefits, but face an average tax/benefit withdrawal rate of 74% up to an average salary have an employment rate of 44%.

2) "Labour supply and marginal tax rates"by A J de Bruin, that boils it down to the simple formula that say "the amount of paid labour would increase by [around 3%] ... if the tax pressure on labour income was reduced by 10% (i.e. a reduction in the marginal tax rate from 40% to 36%)".

Seeing as most basic rate taxpayers face an effective marginal tax/NI rate of 41% (including Employer's NI), that means we could more or less eradicate unemployment by scrapping Employer's NI and getting the marginal rate down to 33% (22% basic rate tax plus 11% Employee's NI).

And if not, it must be worth a f***ing try!