Imagine that you didn’t know how cold weather and turning on the heat affected the temperature of a room.
You’d watch the boiler burn more or less gas, and you’d watch the weather get colder and hotter, but if the thermostat and central heating were working properly there’d be no change in the temperature of the room. If you didn’t already know, you’d learn nothing – you would probably even conclude that boilers and weather had no effect on how hot rooms were.
This is Milton Friedman’s “thermostat example” and, although he used it to explain Federal Reserve policy and inflation, it’s applicable to the minimum wage as well.
Imagine you have a skilled body setting minimum wages in order to try to help those at the bottom without causing unemployment. If they’re doing their job correctly, you will observe no link between hikes in the minimum wage and employment – precisely because they only increase the rate when they don’t think it’ll put anyone out of work.
But even being aware of this paradox is not enough. Firms do not for the most part hire labour and machines on a job-by-job basis; they make plans that last for months, years and, in some cases, decades. Even if a minimum wage hike makes employing certain staff uneconomic in the long run, in the short run they cannot simply sack them all tomorrow and immediately switch to fewer, higher-skilled employees, or robots.
They will adjust slowly and over time. If minimum wages have an impact, they will have an impact steadily, reducing employment growth, rather than causing firms to immediately ditch workers.
We also have to drill down into the groups we think a minimum wage impacts. Only around 5 per cent of the UK workforce is currently paid at the minimum rate. The median annual wage is about double what working full time at the minimum would earn you. It would be very surprising if hikes had any negative impact on those paid well above the base.
Finally, minimum wage impacts are not linear. Taking the minimum from zero to £2 per hour will not have the same impact as taking it from £7 to £9 per hour. We would expect rapidly increasing downsides from each extra pound we add as we get higher.
All of these difficulties make the specific pay floor rate, like the Bank of England’s base interest rate, a poor candidate for party-political tussle, and until 2015’s Autumn Statement this was widely accepted. The Low Pay Commission (LPC), staffed with industry figures and labour economists, set the minimum wage, raising it as little as 4p at times when it thought bigger hikes would be risky.
As I show in my new Adam Smith Institute paper, out today, this all changed with the National Living Wage, which replaces the old minimum for over-25s. The LPC has been relegated to determining the exact path of increases while then-chancellor George Osborne chose the goal, an arbitrary 60 per cent of median earnings by 2020.
No, we have not seen immediate job losses, but then again few of the sceptics predicted them. But over time, the path for employment will be worse: the OBR, the official economic forecaster, reckons 60,000 fewer jobs will be created under the system. The bigger worry is that the issue becomes a political football. While the LPC decided independently, politicians could not fight a “race-to-the-top”, offering visible wage hikes to voters, hoping they wouldn’t notice their invisible costs. Now, that is a very real prospect.
The only upside is that the thermostat example will no longer be applicable. If government is setting minimum wages with no thought at all for their wider and long-term labour market impacts, then we could well see a relationship between hikes and unemployment. That might be nice for economists like me; I suspect the unemployed whose jobs never get created might disagree.