Why does Australia have mini-recessions?
By Scott Sumner
Both here and at MoneyIllusion I’ve occasionally done posts on the odd lack of mini-recessions in the US. I define these as periods where unemployment rises by between 1.0% and 2.0% and then falls back, although you could set the lower bar at 0.7% if you removed the brief 1959 steel strike. If you smoothed the unemployment series with a three-month moving average you could also get a wider range. In other words, when unemployment starts rising in America, it rises by a lot before falling back. The unemployment rate graph here looks “smooth”, not jagged like a stock market index graph that follows a random walk. In a stock market graph, small changes are more common than medium size changes, which are more common than large changes.
But other countries like Australia do have mini-recessions. You can see them on the Australian unemployment graph during 2001-02, 2008-09 and 2013-14:
Australia’s last two actual recessions occurred around 1982-83 and 1990-91, when unemployment rose by about 5.0%. Both were large recessions. During the three mini-recessions, the unemployment rate rose by 1.0% to 2.0%, before falling back. In contrast, America has had 5 full-blown recessions since 1980, three mild ones and two big ones. What allowed Australia to replace America’s three mild actual recessions with three even milder mini-recessions?
My basic view of the business cycle in big economies is that most recessions are caused by unstable monetary policy. In Australia, both the 1982 and 1991 recessions led to a big drop in inflation, so they might be regarded as in some sense “intentional”, not a policy mistake. They were the unfortunately price that must be paid to get inflation under control. The same is true of America’s 1982 recession. But most US recessions do look like policy mistakes—dramatic slowdowns in NGDP growth leading to a fall in RGDP (due to sticky nominal wages), which have no benefit to society. Money was simply too tight.
So why was monetary policy better in Australia than in the US? I believe that reliance on interest rate targeting, combined with a backward-looking focus on past macroeconomic data, makes Fed policy too procyclical. Policy is too tight during recessions and too easy during booms. We should rely more on market prices available in real time.
The exchange rate is not an optimal policy indicator, especially for a big economy like the US. But unlike GDP data, it is an important economic indicator that is available in real time. And whereas falling interest rates might reflect monetary easing or tightening, falling exchange rates are quite likely to reflect easing.
While it’s true that even exchange rates are a somewhat ambiguous indicator, reflecting various types of shocks, at least with exchange rates there’s nothing like the liquidity, income and Fisher effects problem that you have with interest rates. Interest rates are perhaps the worst possible monetary indicator.
So let’s say we are entering a slowdown that threatens to become a recession. Suppose the Fed cuts interest rates, believing it has eased monetary policy. But the economic slowdown has quietly reduced the equilibrium (natural) interest rate, so in fact the Fed has not eased policy as they hoped and expected, despite the rate cut. Eventually they see their mistake and ease policy more aggressively. But by then it’s too late; we are in a mild recession.
Now assume the same shock hits Australia. The Reserve Bank of Australia (RBA) eases policy and looks at exchange rates as an indicator that they have eased enough. They cut interest rates enough to significantly depreciate the Australian dollar. The exchange rate tells the RBA whether the monetary stimulus they are attempting has actually been enacted. Unlike in the US, the RBA doesn’t fail to enact the desired stimulus due to an over-reliance on interest rates and past macro data, because they have a real time market indicator to guide their policy response—exchange rates. This quick and effective monetary response to the economic slowdown results in a mini-recession, rather than the full-blown recession experienced in the US.
[In my view, even mini-recessions might be prevented if central bankers relied on NGDP futures prices, but that’s another issue.]
If I’m right, then you should expect to have seen the RBA reduce the exchange rate of the AUS$ during their mini-recessions, in order to prop up aggregate demand. And that seems to be the case:
The Australian dollar depreciated during 2000-02, 2008-09, and 2013-14, which are the periods where mini-recessions occurred. That Aussie dollar depreciation was the way that the RBA knew that they had actually implemented monetary stimulus, not the phony stimulus you often see in the US when the Fed’s interest rate cutting lags the fall in the equilibrium interest rate, and monetary policy does not actually ease.
In 1991, the RBA did not sharply depreciate the AUS$, presumably because they were willing to bite the bullet in order to get inflation down to a more acceptable level.
The bottom line is that the “price of money” approach to monetary policy is far more powerful and unambiguous than the “rental cost of money” approach to policy. In the US, FDR used the price of money approach during 1933, and the effects were immediate and dramatic.
PS. RBA monetary policy is currently too tight.
PPS. I’m no expert on Australian macro history, so please let me know I’ve gotten any facts wrong.