Friday, August 23, 2013


Talk to VCTA Teachers Day, Melbourne, Friday, August 23, 2013

Often when I talk to economics teachers I focus on helping them keep up to date with the latest thinking on some topic, believing they need to know a lot more background information than their students do and leaving it for them to decide how much of what I’ve said they need to pass on to their kids. But this time I’m going straight to the classroom to give you answers to what I imagine are frequently asked questions by your students - and maybe even by you. The full version of my speech is a lot longer than I’ll have time to talk to today, so make sure you get a copy. Even so, I’m sure there are many more FAQs than I’ve had time to write about - or even think of. So if you’ve got questions I didn’t answer, I’d be grateful if you’d write them down and give them to me - or send me, if you think of them later - and I’ll use them for another talk or bear them in mind for my Saturday column, which has high school economics students as primary target audience.

Can we trust the official unemployment figures?

Short answer: yes and no. Yes we can trust the figures in the sense that, contrary to a widely believed urban myth, there was no time in the past when some government - Labor or Liberal - doctored the figures to make them look better. The figures are calculated by the Bureau of Statistics, which is not a government department but, like the ABC, has a high degree of independence of the elected government and doesn’t let politicians tell it how to measure things. The bureau, which is regarded as one of the best statistical agencies in the world, sticks closely to the statistical conventions laid down by the UN Statistical Commission, the IMF and, in the case of the labour force survey, the ILO. The definitions it uses to decide who is employed, unemployed or ‘not in the labour force’ haven’t changed significantly for many decades.

Remember that the labour force figures come from a sample survey conducted every month by the bureau, using a sample of 26,000 households - up to 20 times those used in media opinion polls. Even so, this does mean it is subject to sampling error, and the results jump around from month to month, meaning it’s best to look at the ‘trend’ (smoothed seasonally adjusted) figures.

Many people assume that the number of people said to be unemployed by the bureau is the same as the number on the dole. This isn’t true. You can be on the dole but not counted as unemployed in the survey (say, because you picked up a few hours of casual work during the week) or you can be counted as unemployed by the survey but not on the dole (say, because your spouse’s job gives you too much income to be eligible). Some old people have ideas in their heads that are a hangover from the time before 1978, when the Fraser government paid to have the labour force survey moved from quarterly to monthly, so that it replaced the old method of measuring unemployment as the number of people registered with the Commonwealth Employment Service.

I suspect some people’s false memories of the government fiddling with the figures stem from their memory of controversies over governments changing rules about how much work you can do and still be eligible for disability benefits or the dole. It’s sometimes claimed that a government has tried to hide some of the unemployed by putting them on training schemes. But people have been making such claims for years and the claim implies the training schemes are phoney, that they’d be of little value to the job seeker and are motivated only by a desire to fudge the figures. Whether a person is classed as unemployed depends not on how they’re classified by a government department, but on what answers they give to the bureau’s interviewers.

So, yes, we can trust the official figures in the sense that they haven’t been fiddled. But, no, we can’t trust them in the sense that they don’t give an accurate picture of the extent of unemployment. It is true - and has been for decades - that, under the international convention, someone who’s done as little as an hour’s work in the previous week is classed as employed, not unemployed. This means the official definition of unemployment is too narrow, making it too hard to qualify as unemployed and thus understating the full extent of joblessness. Note that very few people actually work only a few hours a week. It’s also true that the majority of people working part time (ie less than 35 hours a week) are happy with the number of hours they’re working. Many full-time students, young mothers and semi-retired people don’t want to work full-time.

Even so, a significant number of part-timers do wish they could get more hours, so we have a significant problem with under-employment. I suspect this measurement problem has arisen because the decision to call someone employed if they worked for only a few hours was made long ago when part-time and casual employment was quite rare. As it has become increasingly more common, the original definition of unemployment has become increasingly misleading.

The bureau has tacitly acknowledged this by calculating the rate of underemployment and adding this to the official unemployment rate to get the rate of ‘labour force underutilisation’. This broader measure of unemployment is calculated every quarter and published with the monthly labour force survey. From July 2014 the bureau plans to calculate and publish the broader measure monthly. Let’s hope this will prompt economists and the media to give it more attention.

In May 2013 the trend unemployment rate was 5.5 pc, while the underemployment rate was 7.3 pc, giving an underutilisation rate of 12.8 pc. Note that the measure counts as underemployed not just people working part-time who’d prefer to be full-time, but also those part-timers who’d like only a few more hours. So to that extent its definition of unemployment is probably a little too broad.

For many years I’ve used the rough rule of thumb that the easy way to correct the official unemployment rate is to double it. If you’re making comparisons with the past, however, you have to remember to double both the starting point and the end point. And remember that even if the level of the official rate is too low, it should still give a reasonably reliable indication of whether unemployment is rising, falling or staying the same.

Does the RBA still control interest rates when the banks can do as they please?

Short answer: yes it does. The RBA uses market operations to keep the overnight cash rate under very tight control. The cash rate has acted - and still acts - as the anchor for all other short-term and variable interest rates. Of course, all the other interest rates - from bank bill rates to mortgage interest rates - are a margin (or ‘spread’) above the cash rate because they involve riskier lending, but for several years before the global financial crisis world financial markets were very steady and those margins changed little. This gave people the impression mortgage interest rates always move in lock-step with the cash rate. After the turmoil of the crisis, however, many of the margins widened. The banks passed this increase in their cost of funds on to their borrowing customers. In the case of people with home loans, the banks did this by increasing their mortgage interest rates by more than any increase in the cash rate, or by failing to pass on the whole of any cuts in the case rate. Note that the banks increased the rates they charge their business borrowers by a lot more than they increased the politically sensitive mortgage rates.

For a brief period during the GFC the overseas financial markets in which our banks borrowed a high proportion of the money they lent to their customers ceased to operate. When trading resumed their margins were a lot higher. Realising the extent of our banks’ over-dependence on overseas ‘wholesale’ markets, the share market, the credit rating agencies and the official regulators put pressure on our banks to borrow more of the funds they needed from domestic depositors, whose deposits tended to be ‘sticky’ (slow to move away in search of higher returns) and thus more dependable. The resulting sudden surge in all the banks’ demand for deposits forced up the interest rates they paid on deposits, particularly term deposits, raising them from below the cash rate to above it. This, of course, was a great benefit to Australian savers, but the banks passed this higher cost on to their borrowers.

Could the banks have absorbed these higher borrowing costs? They could have - their profitability (not just the absolute size of their profits, but the rate of their profits relative to the value of their total assets or their shareholders’ capital) is very high by world standards or by the standards of other Australian industries - but they chose not to. And the limited degree of competition between the members of the big-four banking oligopoly gave them the pricing power to pass their higher costs on to borrowers and preserve their rate of profitability.

But don’t confuse the rights and wrongs of the banks’ actions with the quite separate question of whether their behaviour has robbed monetary policy of its effectiveness. It hasn’t. Why not? Because although the RBA uses the cash rate as its instrument, what does the real work of monetary policy are the market interest rates actually paid by businesses and households, so the RBA focuses on getting market rates where it wants them to be. If the independent actions of the banks cause market rates to be higher than where the RBA wants them, it simply cuts the cash rate by more to achieve its desired result. In other words, the fact that the banks’ margin above the cash rate is now wider than it was before the GFC simply means the RBA has had to cut the cash rate by more than it otherwise would have to get markets rates to where it wants them.

Does monetary policy still work?

Short answer: yes. When the share and property markets were booming in the late 1980s, the RBA spent several years raising interest rates to get the boom under control. The rise in rates didn’t seem to be working, and it became fashionable to say that monetary policy had become ineffective. I was still wondering whether this could be true when the economy started the slide that became the recession of the early 90s, the worst recession since the Depression, in which unemployment got close to 11 pc. Then all the smarties started saying interest rates had been held ‘too high for too long’.

There could be no better experience to cure me of ever doubting that monetary policy was effective. And yet we hear such claims whenever people observe a delay between the RBA starting to move the cash rate and making clear its desire to speed up or slow down demand but nothing seems to be happening. When the RBA cuts the rate but there’s a delay before demand picks up, people use an old Keynesian phrase that using interest rates to try to stimulate demand is like ‘pushing on a string’. But that analogy is appropriate only when the economy is in a liquidity trap - which the North Atlantic economies may be in at present, but we certainly aren’t.

In 40 years of watching the management of the Australian economy I can’t recall any time when monetary policy has failed to move demand in the desired direction. The problem is just that, as you well know, monetary policy operates with a lag that’s ‘long and variable’. Another thing that makes the process slow and adds to people’s impatience is that the RBA almost invariably moves in baby steps of 0.25 percentage points. Clearly, a single 25 basis point change isn’t likely to have a big effect on decisions about borrowing and spending. It’s probably true, too, that the response to a monetary tightening or loosening episode isn’t proportional or linear. That is, you may adjust rates several times without getting much effect, but then anther click finally has a big impact. The RBA uses the rule of thumb that most of the effect of a monetary policy on demand occurs within two years, with maybe two-thirds of the full effect occurring in the first year. The effect on inflation - which, of course, runs via the effect on demand - is longer again.

Would a big cut in the cash rate produce a fall in the dollar?

Short answer: no. This question has been asked a lot in recent times as trade-exposed industries such as manufacturing have be hard hit by the high dollar associated with the resources boom.

The first point to understand is that, in practice, economists don’t have a good handle on what factors determine movements in the exchange rate over short periods of less than a year of so. There are rival theories, but no particular theory always gives a convincing explanation of why the exchange rate has moved - or not moved - as it has in recent weeks. Instead, one theory tends explain recent events better than another does at a particular time, so economic practitioners tend to switch between the rival theories depending on which one seems to be working better at the time. I think the reason no theory seems to work well at all times is that the global foreign exchange market isn’t nearly as rational as the perfect market hypothesis assumes.

In the old days, a common theory was that the currency of a country with an excessive current account deficit would tend to depreciate, so as to help bring it back to equilibrium and, similarly, the currency of a country with an excessive current account surplus would tend to appreciate. These days, you rarely hear this theory relied on because there’s little if any empirical support for it. I think it was a hangover from the days of fixed exchange rates, when it was clear the authorities’ decisions on whether to devalue or revalue the currency were determined by pressures on their current accounts. In these days of floating currencies and the removal for foreign exchange controls, it’s clear the ‘driver’ of floating exchange rates has switched from the current account to the capital account - that is, from trade flows to capital flows.

These days, and particularly from an Australian perspective, there are three main, rival theories to explain exchange rate movements. The first is that the biggest influence over our exchange rate is our terms of trade, and particularly world primary commodity prices. There is much empirical support for this view if you look at a graph of the two over the years, though you can see the correlation breaking down over some shorter periods. The second theory is that the biggest influence over our exchange rate is our ‘interest-rate differential’ - the size of the difference between our official interest rate (or short-term commercial rates) and those of the major developed economies, particularly the United States. The higher our rates are relative to the others, the more our exchange rate is likely to be high and rising, and vice versa. Note that this is very much a capital-flows driven theory. The third theory is a kind of combination of the first two: countries with strong economic prospects relative to the major developed countries should have strong currency, whereas countries with weak prospects relative to the majors should have a weak currency. This theory makes a lot of sense and often seems to be pretty true, but there are times when it’s far from true.

Australia’s very strong exchange rate over most of the past decade is commonly explained by the resources boom and our exceptionally favourable terms of trade as a result of record high prices for coal and iron ore. Its rise can not be explained by any increase in our interest rates relative to the major economies, even though their rates have been at rock bottom since the global financial crisis. But this has not discouraged people adversely affected by the high dollar from convincing themselves the high rate is the product of currency market speculation or our relatively high rates since the GFC, and then arguing the RBA should make a big cut in our cash rate with the express purpose of engineering a big fall in the dollar.

Our terms of trade began falling in about September 2011, but the dollar didn’t start to fall until April 2013. This delay probably encouraged people to switch to a different theory. They may have thought the RBA was being too cautious in the speed at which it was bringing rates down.

Although no one can be too dogmatic about these things, the RBA does not believe the interest rate differential has very much effect our exchange rate. And this is despite the signs we see that expectations about whether the RBA will or won’t move rates haves an immediate effect on the bill rate. These effects are very temporary. During the period in which the RBA was lowering rates and openly expressing its hope that the dollar would fall to a more appropriate level, many people concluded it was cutting rates in the hope this would lower the exchange rate. It wasn’t. Rather, it was loosening monetary policy because the exchange rate wasn’t coming down. That is, it was trying to ease pressure on the tradeables sector as a substitute for a lower dollar.

Although the Aussie stayed high for about 18 months after commodity prices had fallen sharply, it has fallen by about 10 per cent since April 2013. Some people may attribute this to steady easing in policy over most of that time, but the BRA doesn’t agree with them. A much more likely explanation is that the Aussie finally began falling when Wall Street began worrying that the long-awaited pickup in the US economy would prompt the Fed to start ‘tapering’ the size of its quantitative easing. QE - the central bank’s purchase of bonds and other securities which are paid for merely with bank credits - puts downward pressure on a country’s exchange rate.

The point to note is that the exchange rate is a relative price - the value of my currency relative to the value of yours. So it shouldn’t be so surprising that changes in the level of our exchange rate need to be explained in terms of changed conditions in the US as well as changes in Australia.

Why are our interest rates always higher than other people’s?

Short answer: because we’re riskier. It’s true our interest rates are almost invariably higher than those in the major economies. This has been true for many years. It wasn’t hard to understand before the mid-1990s - when our inflation rate was still well above everyone else’s - but it remains true even when you compare real interest rates.

The explanation seems to be that, as a nation of perpetual net borrowers from the rest of the world (we run a persistent current account deficit), we are required to pay our foreign lenders a significant risk premium on top of the going international rate to compensate them for the extra risks they run in lending to a country that already has a very large net foreign debt and that, being a relatively small economy, is perceived to be more volatile (even though that’s not always true).

Another way of putting it is that Australia always has higher interest rates because we’re a country with an abundance of potentially profitable investment projects relative to the major economies. Our projects have to be relatively profitable or we wouldn’t be able to continue borrowing despite the high risk premium foreign lenders require us to pay.

Does a budget deficit mean fiscal policy is expansionary and a surplus mean it’s contractionary?

Short answer: no they don’t. Life would be very simple for students of macroeconomics if they did, but unfortunately they don’t. Why not? Because what macro economists focus on is not the level of economic activity, but the change in the level - that is, whether the economy has been/will be expanding or contracting. That means they’re interested in determining whether the budget - fiscal policy - is making a positive or negative contribution to economic growth. So it’s the change in the budget balance - and the direction of the change - that matters when assessing whether a particular budget is expansionary or contractionary.

These days the RBA and most market economists assess the stance of policy adopted in a particular budget simply by looking and the direction - and size - of the expected change in the budget balance between the previous year and the budget year. An expected reduction in a deficit or increase in a surplus is regarded as contractionary; any expected increase in a deficit or decrease in a surplus is regarded is expansionary. As a guide, the change needs to be equivalent to at least 0.5 pc of GDP to be significant. A change of 1 pc or more is extremely significant.

Strict Keynesians, however, define the stance of fiscal policy differently, distinguishing between changes in the cyclical component of the budget balance (caused by operation of the budget’s automatic stabilisers as the economy moves through the business cycle) and changes in the structural component (caused by governments’ explicit changes to taxes and spending programs). So they define the stance of policy adopted in a budget according to the direction of the expected change in the structural component arising from the net effect of the spending and taxing changes announced in the budget. They ignore the change in the budget balance caused by the economy’s effect on the budget, focusing on the change caused by the budget’s effect on the economy.

Note, changes in the stance of fiscal policy will be only one of the factors contributing to whether the economy is expanding or contracting. Other factors include: the stance of monetary policy, movements in the exchange rate, changes in the world economy and in confidence.