As a
general interest follow-up to the IMF World Economic Outlook, I decided to test the IMF’s hypothesis
in Australia specifically.
PART-TIME EMPLOYMENT IS SLOWING INFLATION
One of the key findings of the IMF was that the modern proliferation of part-time employment is hindering wage growth and making it harder for Central Banks around the advanced world to achieve their inflation targets[1], even a decade after the GFC.
Well, a simple comparison between Australia’s inflation rate over time, and the share of the labour force that is full-time employed, seems to support that notion. Australia’s full-time share of employment has fallen from 85% in 1978 to 68% most recently in September 2017, with the remainder being part-time workers. And correspondingly, Australia’s inflation rate has fallen from commonly over 8% per annum up to 1990, to an average of 2.5% since March 1993.
Over this period, the R-squared between these two variables is 0.63 – a fairly strong relationship, where the increasing proliferation of part-time work does seem to be associated with falling inflation.
So the IMF appears to be on to something.
Note that I’m using headline consumer price index (CPI) inflation above, instead of core CPI inflation. Headline CPI inflation is an estimate of the rate at which the goods and services purchased by the average household are increasing in price. Core CPI inflation excludes items such as food and fuel which, due to the volatility of their prices, distract from the underlying trend. So core CPI inflation is arguably a better measure of inflation, but I used headline CPI inflation above because the data provides a longer time series.
But even when using the shorter core CPI inflation time series (since March 1983), the whole period still returns an R-squared value of 0.59 – fairly strong still. But what using core CPI inflation does reveal – in further support of the IMF’s findings – is that the relationship between inflation and the full-time employment share has become even stronger since the GFC (an R-squared of 0.72 since March 2008).
That is, the modern proliferation of part-time work really does seem to be increasingly associated with lower inflation.
CORRELATION VS. CAUSATION
Now, I’m an economist, so I understand the risks of assuming causation from correlation. As the meme below rather humorously suggests, you could just as easily conclude that there is causation between Nicolas Cage movies and pool drownings.
But even if these Nicolas-Cage-pool-drownings numbers are correct (I admittedly haven’t check), there’s no reason – short of subliminal messaging – that would lead us to conclude that Nicolas Cage movies are causing pool drownings (or the other way around).
But there is logic behind the IMF’s theory. Part-time workers arguably have less bargaining power to demand higher wages than their full-time counterparts – quite simply because an employer risks losing more labour hours if they don’t meet the needs of a full-time worker. Think of it like one person asking for a raise under the implicit threat of quitting, versus two workers asking for a raise. Who’s more likely to get the raise? The two workers because together, they have more labour hours with which to bargain. It’s the same logic behind trade unions improving worker bargaining power by acting on their collective behalf, rather than individually.
So the IMF’s logic does go beyond mere correlation.
But there are a few points to consider.
THE BALANCE BETWEEN HIGH AND LOW INFLATION
First, we don’t want inflation to be as high or as volatile as it was before the 1990s. True, inflation that is too low is a risk because the closer it gets to negative territory, the greater the chances of a self-fulfilling deflation-led Depression where households, businesses and even government indefinitely delay their major expenditures and investments in the expectation that these expenditures and investments will be cheaper in the future, thereby creating a self-fulfilling prophecy and causing the economy to crash. Deflation also increases the real burden of debt, by putting downward pressure on household wages, business revenue, and government tax revenue while the absolute level of debt remains constant. As Keynes observed, high inflation can cause problems, too, but at least it encourages spending, while the expectation of deflation can “inhibit the productive process altogether”. Some Central Banks are even considering increasing their inflation targets to around 4% to provide them with a greater buffer from zero, given how close inflation rates around the advanced world have been/are to the dreaded negative territory.
But we also don’t want inflation to be too high. High inflation isn’t necessarily bad, per se, as long as it’s predictably high. If we know that inflation is going to be 10% (say 9.5-10.5%) with as much confidence as we used to expect it to be between 2% and 3%, then that’s probably okay. Everyone will simply adjust their expectations and behaviour to account for 10% inflation. Theoretically, a million per cent inflation would be fine, as long as it’s consistently a million per cent (even though there’s no reason we’d want it that high). Remember, inflation shouldn’t have any real consequences if its predictable – they’re just numbers[2].
Unfortunately, high inflation can be (and historically has been) also very volatile. And that’s a problem. Wage contracts are harder to negotiate because it’s harder to predict the future cost of living, resulting in either excessive wages, reduced business profits and increased unemployment, or insufficient wages, rising inequality and slower economic growth. Businesses have a harder time justifying major investments in property, machinery, labour, etc. because they don’t know what the price of their goods or services will be in the future. The finance industry also has a harder time lending money if it doesn’t know what interest rate to charge to offset future inflation[3]. Even governments need to do cost-benefit analyses on their investments, which becomes harder when future prices are unpredictable. But even if high inflation weren’t also volatile, we still wouldn’t need it to be too high to be a safe distance from negative territory.
So while we want a little inflation, if worker bargaining power is the key, we may not want to push it too far.
THE RESERVE BANK AND AUSTRALIA’S AGEING DEMOGRAPHIC ALSO AFFECT INFLATION INDEPENDENTLY OF THE FULL-TIME EMPLOYMENT SHARE
Second, there are definitely other factors at play here. From the early 1990s to the GFC, the correlation between inflation and the full-time employment share virtually disappears. As mentioned above, despite some short-term volatility, inflation was relatively flat around a 2.5% average from March 1993, while the full-time employment share continued to decline. And there’s a reason why this correlation could have weakened during this period. In March 1993, the RBA unofficially adopted what would eventually become it’s official and explicit target of 2-3% inflation over the business cycle. The RBA was also given independence from government to achieve this target with all the tools at its disposal[4]. And to their credit, since March 1993, Australian inflation has averaged precisely 2.5% – not bad! So the RBA became much more effective at controlling inflation, apparently regardless of what the full-time employment share was doing – at least for a while.
Thirdly, Australia’s population continues to age and the workforce as a share of that population continues to shrink. This will be an ongoing drag on economic growth and therefore, probably inflation, and could already be partly responsible for recent sluggish inflation, independent of the rise in part-time employment.
Fourthly and finally, (though these four factors are not by any means exhaustive), as Milton Friedman once said, “inflation is always and everywhere a monetary phenomenon”. While we may like to scapegoat “greedy businessmen, grasping trade unions, spendthrift consumers, Arab sheikhs, bad weather, or anything else that seems remotely plausible” as the cause of inflation, none of these entities can affect ongoing inflation because none of them possess a printing press.
THE COMMON ANCESTOR – DEMAND
Personally, I think both the modern proliferation of part-time work and low inflation, rather than driving each other directly, are both driven by insufficient demand. Simply, I think the solution isn’t going to be Central Bank-driven or monetary policy-driven. It’s going to be demand-driven. As I’ve mentioned before, while Central Banks still have only limited scope to further stimulate their economies with interest rates and liquidity, and while global interest rates are still so low, infrastructure spending by governments should be used. Happily, this should drive both inflation and full-time employment.
THE IMF OFFERS VALUABLE INSIGHT BUT WE MUST PROCEED WITH CAUTION
So no doubt the share of full-time employment isn’t the only factor driving the inflation rate, certainly not for Australia. This is not to say the IMF’s findings were wrong – never would I be so bold[5]. Perhaps the attempts of Central Banks to control inflation independently of the full-time employment share only succeeded temporarily, and recent years have simply seen a return to the long term downward trend. But it is an important lesson in the difference between correlation and causation, and in the risk of applying general findings for the advanced world to specific countries without considering their individual (perhaps unique) circumstances.
PART-TIME EMPLOYMENT IS SLOWING INFLATION
One of the key findings of the IMF was that the modern proliferation of part-time employment is hindering wage growth and making it harder for Central Banks around the advanced world to achieve their inflation targets[1], even a decade after the GFC.
Well, a simple comparison between Australia’s inflation rate over time, and the share of the labour force that is full-time employed, seems to support that notion. Australia’s full-time share of employment has fallen from 85% in 1978 to 68% most recently in September 2017, with the remainder being part-time workers. And correspondingly, Australia’s inflation rate has fallen from commonly over 8% per annum up to 1990, to an average of 2.5% since March 1993.
Over this period, the R-squared between these two variables is 0.63 – a fairly strong relationship, where the increasing proliferation of part-time work does seem to be associated with falling inflation.
So the IMF appears to be on to something.
Note that I’m using headline consumer price index (CPI) inflation above, instead of core CPI inflation. Headline CPI inflation is an estimate of the rate at which the goods and services purchased by the average household are increasing in price. Core CPI inflation excludes items such as food and fuel which, due to the volatility of their prices, distract from the underlying trend. So core CPI inflation is arguably a better measure of inflation, but I used headline CPI inflation above because the data provides a longer time series.
But even when using the shorter core CPI inflation time series (since March 1983), the whole period still returns an R-squared value of 0.59 – fairly strong still. But what using core CPI inflation does reveal – in further support of the IMF’s findings – is that the relationship between inflation and the full-time employment share has become even stronger since the GFC (an R-squared of 0.72 since March 2008).
That is, the modern proliferation of part-time work really does seem to be increasingly associated with lower inflation.
CORRELATION VS. CAUSATION
Now, I’m an economist, so I understand the risks of assuming causation from correlation. As the meme below rather humorously suggests, you could just as easily conclude that there is causation between Nicolas Cage movies and pool drownings.
But even if these Nicolas-Cage-pool-drownings numbers are correct (I admittedly haven’t check), there’s no reason – short of subliminal messaging – that would lead us to conclude that Nicolas Cage movies are causing pool drownings (or the other way around).
But there is logic behind the IMF’s theory. Part-time workers arguably have less bargaining power to demand higher wages than their full-time counterparts – quite simply because an employer risks losing more labour hours if they don’t meet the needs of a full-time worker. Think of it like one person asking for a raise under the implicit threat of quitting, versus two workers asking for a raise. Who’s more likely to get the raise? The two workers because together, they have more labour hours with which to bargain. It’s the same logic behind trade unions improving worker bargaining power by acting on their collective behalf, rather than individually.
So the IMF’s logic does go beyond mere correlation.
But there are a few points to consider.
THE BALANCE BETWEEN HIGH AND LOW INFLATION
First, we don’t want inflation to be as high or as volatile as it was before the 1990s. True, inflation that is too low is a risk because the closer it gets to negative territory, the greater the chances of a self-fulfilling deflation-led Depression where households, businesses and even government indefinitely delay their major expenditures and investments in the expectation that these expenditures and investments will be cheaper in the future, thereby creating a self-fulfilling prophecy and causing the economy to crash. Deflation also increases the real burden of debt, by putting downward pressure on household wages, business revenue, and government tax revenue while the absolute level of debt remains constant. As Keynes observed, high inflation can cause problems, too, but at least it encourages spending, while the expectation of deflation can “inhibit the productive process altogether”. Some Central Banks are even considering increasing their inflation targets to around 4% to provide them with a greater buffer from zero, given how close inflation rates around the advanced world have been/are to the dreaded negative territory.
But we also don’t want inflation to be too high. High inflation isn’t necessarily bad, per se, as long as it’s predictably high. If we know that inflation is going to be 10% (say 9.5-10.5%) with as much confidence as we used to expect it to be between 2% and 3%, then that’s probably okay. Everyone will simply adjust their expectations and behaviour to account for 10% inflation. Theoretically, a million per cent inflation would be fine, as long as it’s consistently a million per cent (even though there’s no reason we’d want it that high). Remember, inflation shouldn’t have any real consequences if its predictable – they’re just numbers[2].
Unfortunately, high inflation can be (and historically has been) also very volatile. And that’s a problem. Wage contracts are harder to negotiate because it’s harder to predict the future cost of living, resulting in either excessive wages, reduced business profits and increased unemployment, or insufficient wages, rising inequality and slower economic growth. Businesses have a harder time justifying major investments in property, machinery, labour, etc. because they don’t know what the price of their goods or services will be in the future. The finance industry also has a harder time lending money if it doesn’t know what interest rate to charge to offset future inflation[3]. Even governments need to do cost-benefit analyses on their investments, which becomes harder when future prices are unpredictable. But even if high inflation weren’t also volatile, we still wouldn’t need it to be too high to be a safe distance from negative territory.
So while we want a little inflation, if worker bargaining power is the key, we may not want to push it too far.
THE RESERVE BANK AND AUSTRALIA’S AGEING DEMOGRAPHIC ALSO AFFECT INFLATION INDEPENDENTLY OF THE FULL-TIME EMPLOYMENT SHARE
Second, there are definitely other factors at play here. From the early 1990s to the GFC, the correlation between inflation and the full-time employment share virtually disappears. As mentioned above, despite some short-term volatility, inflation was relatively flat around a 2.5% average from March 1993, while the full-time employment share continued to decline. And there’s a reason why this correlation could have weakened during this period. In March 1993, the RBA unofficially adopted what would eventually become it’s official and explicit target of 2-3% inflation over the business cycle. The RBA was also given independence from government to achieve this target with all the tools at its disposal[4]. And to their credit, since March 1993, Australian inflation has averaged precisely 2.5% – not bad! So the RBA became much more effective at controlling inflation, apparently regardless of what the full-time employment share was doing – at least for a while.
Thirdly, Australia’s population continues to age and the workforce as a share of that population continues to shrink. This will be an ongoing drag on economic growth and therefore, probably inflation, and could already be partly responsible for recent sluggish inflation, independent of the rise in part-time employment.
Fourthly and finally, (though these four factors are not by any means exhaustive), as Milton Friedman once said, “inflation is always and everywhere a monetary phenomenon”. While we may like to scapegoat “greedy businessmen, grasping trade unions, spendthrift consumers, Arab sheikhs, bad weather, or anything else that seems remotely plausible” as the cause of inflation, none of these entities can affect ongoing inflation because none of them possess a printing press.
THE COMMON ANCESTOR – DEMAND
Personally, I think both the modern proliferation of part-time work and low inflation, rather than driving each other directly, are both driven by insufficient demand. Simply, I think the solution isn’t going to be Central Bank-driven or monetary policy-driven. It’s going to be demand-driven. As I’ve mentioned before, while Central Banks still have only limited scope to further stimulate their economies with interest rates and liquidity, and while global interest rates are still so low, infrastructure spending by governments should be used. Happily, this should drive both inflation and full-time employment.
THE IMF OFFERS VALUABLE INSIGHT BUT WE MUST PROCEED WITH CAUTION
So no doubt the share of full-time employment isn’t the only factor driving the inflation rate, certainly not for Australia. This is not to say the IMF’s findings were wrong – never would I be so bold[5]. Perhaps the attempts of Central Banks to control inflation independently of the full-time employment share only succeeded temporarily, and recent years have simply seen a return to the long term downward trend. But it is an important lesson in the difference between correlation and causation, and in the risk of applying general findings for the advanced world to specific countries without considering their individual (perhaps unique) circumstances.
[1] The
rate at which the goods and services in an economy are increasing in price
which, in Australia, the RBA targets at 2-3% per year.
[2]
Though it would still be a hard political sale for a Central Bank, which has
spent so much time building up its credibility around a 2-3% inflation target,
to suddenly announce that they want 10% inflation.
[3] Also,
other countries might get annoyed if they have lent money to a country that is
now producing higher inflation and paying back their debt in increasingly
worthless currency, without them necessarily being able to charge a higher
interest rate to compensate. While decreasing the real value of their debt may
be beneficial for the country that is in debt, it risks starting a
currency/trade war with the lender country.
[4]
Their official objectives under the Reserve Bank Act 1959 are: a stable
currency; full employment; and the economic prosperity and welfare of the
Australian people over the medium term. And it achieves these three objectives
partly through maintaining a low and stable inflation rate.
[5]
Until I start talking about their role in Europe’s austerity efforts.
No comments:
Post a Comment