Saturday, September 29, 2018

How economists lost their fear of minimum wage rises

Do rises in the minimum wage come at the expense of jobs? If you listen to the employer groups, they certainly do. But this is a question on which economists have changed their tune.

So much so that the latest issue of the Reserve Bank’s Bulletin includes an article by one of its researchers, James Bishop, concluding there’s no evidence that modest, incremental increases in minimum award wages have an adverse effect on hours worked or the rate of job destruction.

There’s no way the Reserve would have said such a thing 20 years ago.

For decades, most economists did believe increases in the minimum wage would cause employment to be lower than otherwise if they took the wage rate above where market forces would have set it – the “market-clearing” price at which the quantity supplied and the quantity demanded were equal.

Their certainty came from elementary economic theory. Their simple “neo-classical” model of markets, the bedrock on which most economists’ thinking is based, told them that if you raise the price of something without any change in its supply, you’ll cause less of it to be demanded.

That was as true for the price of labour as it was for the price of bananas or anything else.

This continued to be the conventional wisdom among economists until 1994, when two American economists, David Card and Alan Krueger, published the results of their “natural experiment” in which they compared what happened in 410 fast food restaurants in two adjoining states after New Jersey raised its minimum wage by 19 per cent but Pennsylvania didn’t.

To much amazement, they found that the rise in the price of labour actually led to a small rise in employment, not a fall.

In other words, they checked the theory against the real world and found it wanting.

This implied that the simplified model of demand and supply might be good for predicting the consequences of a rise in the price of bananas, but it isn’t much good at predicting developments in a market where every unit of labour is different and comes with a human attached.

A model that could predict the outcome Card and Krueger found is one that assumes employers have a degree of market power over wages, allowing them to fix wage rates below where a free market would put them, until the government intervened.

Card and Krueger’s challenge to the conventional wisdom set off decades of empirical studies throughout the developed world trying to replicate or refute their findings. Not surprisingly – since academic economics is riven by ideological conflict – they found both.

Bishop says that, on balance, the weight of evidence is that “modest and incremental increases in minimum wages do not have significant adverse effects on hours worked and job loss”.

But Australia’s system of minimum wages is very different to other countries’ systems, and there hasn’t been much empirical testing here.

Countries such as Britain, Germany and New Zealand set a single national minimum wage; in the United States it varies by state.

In Oz we, too, have a national minimum wage, but we also have more than 100 industrial awards covering particular industries or occupations, each of which sets a number of minimum wage rates for particular job classifications covered by that award.

Awards cover those aspects of employees' pay and conditions that they’re permitted to cover by the national Fair Work Act. Awards are awarded by the Fair Work Commission after submissions from unions and employer groups and they have the force of law.

Pay someone less than the minimum amount specified in the relevant award and you’re breaking the law.

It’s true, of course, that many workers’ pay – a good third of all employees – is determined by their enterprise agreement rather than their award. The wage rates specified in agreements are usually a fair bit higher than those in the award.

Roughly 40 per cent of employees are covered by “individual arrangements” between the individual and their employer, which may be formal (written) or informal. These wage rates need to be at least as high as provided in the individual’s award.

Not a huge number of workers depend on the national minimum wage (of $18.93 an hour, $719 a week and $37,406 a year), but many workers are paid according the much higher minimums set out in their award.

And here’s the trick: when, after a public hearing, the Fair Work Commission decides by how much it will increase the national minimum wage on July 1 each year, it increases the thousands of minimums set out in awards by the same percentage. (The highest award minimum is $171 an hour.)

So our minimum wage directly affects the wages paid to about a quarter of all employees. That’s a much higher proportion than in the other rich economies.

What’s more, the minimum wage increase probably affects many more workers indirectly, particularly those on individual arrangements.

Our national minimum wage has long been among the highest in the rich countries, both in its absolute level and relative to the median wage.

Consider this: while the wage price index has been rising by only about 2 per cent a year in recent years, the annual increase in the minimum wage was 3.5 per cent this year, 3.3 per cent last year and 2.4 per cent in 2016.

All these are the reasons it was important for Bishop to study our minimum wages to check that the broad conclusions reached in other countries also apply to us.

He did, and they do. He finds that our minimum wage increases “appear to have no discernible adverse effect on hours worked or job loss”.

But minimum wages being the contentious topic they are, he’s quick to add some qualifications.

“The results do not necessarily generalise to large, unanticipated changes in award wages. There will always be some point at which a minimum wage adjustment will begin to reduce employment significantly,” he says.

And here’s a worry: “It is possible that the adverse consequences of higher wage floors may be borne by job seekers, rather than current job holders.”