Friday, 30 June 2017

I know it sounds good but will it work? Thoughts on big rises to minimum wages.

Be careful what you wish for - or so they say:
The study, commissioned by the city government of Seattle and published by the National Bureau of Economic Research, found that Seattle’s law incrementally raising its minimum wage — to $13 an hour last year, en route to $15 — resulted in low-wage workers’ earning less money rather than more. This surprised many in Seattle, who had been assured by all the best economists, including Paul Krugman, that such a thing would not come to pass.
The debate about minimum wages is a really important one. And, like all such debates, one that is dominated by righteous assertion - 'how can you not support people being paid a basic minimum, they have to eat' - rather than through considered analysis. It is probably time to take the pain and ask again whether minimum wages work.

The theory (at least as I understand it) tells us that, under all but the rarest of circumstances, if you raise the price of something the demand for that something will decline. All we need to examine is the magnitude of this effect - how quickly or slowly that demand falls as the price rises. It is stretching credulity to breaking point if we suggest that these effects do not apply when the price is the price of an hour's labour. So if we mandate a higher price for certain types of labour the result will be less demand for that labour either through the work concerned being forgone, reduced amounts of the work or else through automation (or, in extremis, through the business that used the labour in question ceasing to exist).

You all know this of course but how come we are all so keen on raising minimum wages (for good and noble reasons about people having a 'living wage' or whatever)? Partly this is because there have been a lot of studies suggesting that minimum wage effects are, at worst, small. The classic - if that's the word - in this genre is a study of fast food restaurants in New Jersey following a minimum wage hike:
Our empirical findings challenge the prediction that a rise in the minimum reduces employment. Relative to stores in Pennsylvania, fast food restaurants in New Jersey increased employment by 13 percent. We also compare employment growth at stores in New Jersey that were initially paying high wages (and were unaffected by the new law) to employment changes at lower-wage stores. Stores that were unaffected by the minimum wage had the same employment growth as stores in Pennsylvania, while stores that had to increase their wages increased their employment.
So much for supply and demand theory! Minimum wages rises increase employment! The problem is that things don't change quickly and reduced employment is built into business plans rather than implemented immediately - employment is sticky (partly because most business managers and especially small businesses really don't like laying folk off).

But there is a middle ground:
Thus, allowing for the possibility of larger job loss effects, based on other studies, and possible job losses among older low-skilled adults, a reasonable estimate based on the evidence is that current minimum wages have directly reduced the number of jobs nationally by about 100,000 to 200,000, relative to the period just before the Great Recession. This is a small drop in aggregate employment that should be weighed against increased earnings for still-employed workers because of higher minimum wages. Moreover, weighing employment losses against wage gains raises the broader question of how the minimum wage affects income inequality and poverty.
What's being suggested here is that in a market like the USA that generates lots of work all that increases in minimum wages do is anticipate rising wages - it's not that there's been a reduction in the numbers of jobs (and it doesn't really matter whether this is 100,000 or 1,000,000) but rather the impact on overall employment. If raising minimum wages doesn't raise unemployment this can only be because the rate at which the economy is generating new demand for labour is greater than or equal to the reduced demand for labour consequential on a higher minimum wage. And in a recession the wage floor might mean a more rapid shedding of labour.

In the UK we have seen surprising labour market resilience - employment has risen despite relatively slow growth and perhaps because of stagnating levels of productivity. A lot of these jobs have been minimum wage jobs but so far the effects of a rising minimum wage (the cutely tagged 'National Living Wage') haven't been seen in overall levels of employment - indeed quite the contrary as record employment and record immigration tell us.

UK wage distribution 2016. Source:ONS

This graph shows the effect of gradual increases in minimum wages. The question is whether the current approach of gradually increasing the minimum wage will continue to protect wage levels for low paid workers without the negative effect of reducing levels of employment for those low paid workers. A minimum wage is no use at all if its consequence is merely more unemployment or fewer hours of work for those low paid workers.

This for me is the reason why the Seattle findings are so important. It's not that we shouldn't consider minimum wages but that the impact of a massive rise in minimum wage is simply to increase unemployment and income among the lower paid. In Seattle, by 2021 the new minimum wage will increase wage costs for employers of minimum wage labour by 60%. It seems to me that the City got itself ahead of its labour market - here are the key findings from the University of Washington study that caused all the fuss (bear in mind that the study found the usual very small effect from the first minimum wage hike in 2015):

It does seem that Seattle has discovered that demand curves slope downwards and we should take lessons from this unsurprising revelation. The best way to improve wages for the least well paid isn't to require that employers pay them more - 18% more in the case of Seattle's 2016 minimum wage rise - but for economic growth to increase demand for labour.

For the UK our ever more surprising ability to create new employment masks these minimum wage effects:
...retail executives concede that sharp increases in the minimum wage have lent urgency to their efforts to use workers more efficiently, by investing in technology that makes many low-skilled jobs obsolete. “On a personal level, we all want to pay our workers more,” said a top executive at one large retailer that has recently announced redundancies. “But there’s going to be unintended consequences from what the government is doing. Automation that used to be too expensive is now cheaper than the people it can replace.”
The question is really how much longer can our economy absorb the workers shed as firms reduce labour levels and introduce automation? So long as the workers leaving Sainsbury or John Lewis can get a job elsewhere with relative ease the rising National Living Wage won't be an issue. But when that rising wage results in job losses the economy can't fill and sees reducing hours worked it becomes counterproductive. Labour's policy of an across the board £10 an hour minimum wage is great politics but such a huge increase in wage bills results in the sort of system shock seen in Seattle. And the people who would feel this shock will be the very low paid workers the policy is designed to help.



Jonathan Bagley said...

The the UK, tax credits and housing benefit subsidise the income of many wage earners to well above the current minimum wage. Tax credits for families with children are effectively taxes on the childless. Increasing the minimum wage will increase the net income of the childless and make little difference to that of families with children. Maybe we should try it. With the population increasing by 500 K a year, we have no need for more reproduction.

Sobers said...

How many jobs have been cut out by the use self service tills in shops? A fair few I'd guess. Thats the long term result of ever increasing wages - employers find ways of replacing people with machines. Indeed raising the cost of the person makes the machine more competitive. Look at those robot warehouses that are coming on stream. Whats the likely effect in the long run of making people more expensive, more or less robots? Yes, the immediate effect may not be that negative (although it will be negative) the long term one definitely will be considerably more obvious.

We appear to have decided to make human labour far more expensive precisely at the point that technology is providing the means to replace those humans with automatons. And the people whose jobs are replaced by robots are not the sort of people who will be able to find work as robot engineers and the like. A cleaner replaced by a robot isn't going to retrain in robotics. They're just going to be out of work on benefits.