Wednesday, May 5, 2010

Does the economy depend on population growth?

Talk to University of Sydney Political Economy Society
May 4, 2010

There are a hundred political economy points I could and would like to make about immigration and population, but time doesn’t permit so I’m going to focus the more strictly economic question: does the economy depend on population growth?

I’ll start by stating upfront where I’m coming from on population: I believe we should do what we can to limit the growth of our population, and do that by focusing largely on immigration. Net immigration has accounted for about half our population growth over recent decades, with natural increase (births minus deaths) accounting for rest. Immigration (and the subsequent children of immigrants) would account for well over half of the 60 per cent growth in the population, from 22 million to 36 million over the 40 years to 2050, as mechanically projected by Treasury - the projection that’s stirred up so much debate.

The reason for focusing on immigration rather than natural increase is that fertility is much harder and more controversial for governments to attempt to influence. In any case, the fertility rate is running just below the 2.1 babies per woman needed just to hold the population constant over the longer term. Arithmetically, some net migration would be necessary to stop the population starting to fall by around the middle of the century. So immigration is the ‘swing instrument’, so to speak, and I’ll focus on it from here on.

What are my reasons for favouring limiting immigration to limit our population growth? It’s mainly my concern about the damaging ecological effects of population growth, as much from a global perspective as from a local Australian perspective. But this concern is augmented by my belief that economic growth (ie increase in material standard of living, as conventionally measured by the real growth in GDP per person) does nothing to increase subjective wellbeing (happiness) in developed countries. If so, why pay a social or environmental price to pursue it? But this isn’t true for developing countries, which is why I believe the rich countries need to limit both their population growth and their growth in GDP per person, to leave more ecological space for the understandable material aspirations of the poor countries. All this is discussed in my new book, The Happy Economist, which will be out in August.

OK, let’s get down to it: what’s the relationship between population growth and economic growth? This needs to be unpeeled like an onion. First, it’s clear that if you have a growing population - more people producing and consuming goods and services - you’ll get a bigger economy. But in narrow economic terms, what’s so good about having a bigger economy? Well, just about all business people, politicians and even economists think it sounds pretty nice. Business people like it simply because it gives them a bigger market to sell to and profit from - a much easier way to grow your business than trying to pinch market share from your competitors. To take an obvious example, the home-building industry wants to boost the demand for new houses. What business wants the politicians generally want, and they probably also think that in a growing economy voters are likely to be more content with the way things are going. As for economists, I think many of them are so conditioned to believe in growth that they’ve long ago stopped inquiring into the whys and wherefores.

But now the second layer of the onion. For a rigorous economic analysis it’s not good enough to simply assume that bigger is better. Why exactly is it better? The conventional answer is that bigger is better if it brings us a higher material standard of living - if it makes us more prosperous. But for this to happen - not necessarily for each individual, but on average, and for the community as a whole - the economy must grow faster than the population grows ie there must be an increase in real GDP per person.

But there’s a third layer: even if increased population does lead to higher GDP per person, who shares in that increase? Conventional economics is about self-interest, so for immigration to be justified economically it has to be shown that the pre-existing population benefits from the decision to increase the population. If instead all the benefit went to the immigrants, then the immigration program would be merely an act of charity.

So, from a narrow, strictly economic perspective, those are the questions to be answered when asking what the relationship is between economic growth and population growth: does population growth lead to higher income per person and, even if it does, do the people who agreed to let in more immigrants gain from that action?

The most recent official attempt to answer those questions came in a report prepared by the Productivity Commission in 2006, Economic Impacts of Migration and Population Growth. Now, the Productivity Commission is a body of impeccable credentials in economic orthodoxy, it’s one of the leading advocates for economic growth and you’d expect it to be very favourably disposed to the belief that immigration makes us better off materially. Which makes its findings all the more significant.

It sought to answer these questions the way economists do, by commissioning some economic modelling. Such models are built on a host of simplifying assumptions, they are driven by the modellers’ beliefs about how the economy works, and so their findings should be viewed with caution. The key assumptions driving the results need to examined, and the whole exercise can be subject to a lot of critical scrutiny. The proposition the PC modelled was the effect of a 50 pc increase in the level of skilled migration over the 20 years to 2024-25. It found that this did cause real GDP to be 4.6 per cent bigger than otherwise in 20 years time. And, yes, this did lead to an increase in real income per person, but the increase was pathetically small: 20 years later real income per person would be 0.7 per cent higher, or $380 a year. The PC found that ‘the distribution of these benefits varies across the population, with gains mostly accrued to the skilled migrants and capital owners. The incomes of existing resident workers grow more slowly than would otherwise be the case’.

The PC concludes that ‘factors other than migration and population growth are more important to growth in productivity and living standards’. Indeed, growth in income per person from technological progress and other sources of productivity growth, and long-term demographic changes, could be expected to be about 1.5 pc per year, or more than $14,000 a year by 2024-25.

So that’s an end point of $380 a year from immigration versus $14,000 a year from technological advance. On this evidence, a rational economic rationalist would have little enthusiasm for population growth. From my perspective, it leaves me confident my opposition to immigration-fed population growth on ecological grounds would not come at any great cost in terms of our material standard of living (or our happiness, for that matter).

But let’s look at why the PC’s modelling exercise came up with conclusions so at variance with what almost all business people, politicians and economists would have expected. It’s because the effects of immigration on the economy are complex, with some positive and some negative, so you have to try to determine the net balance, and the two pretty much cancel each other out. (PC2006report, from p115)

The first positive effect on GDP per person is that immigration leads to an increase in the proportion of the population that’s in the workforce producing things. The second positive effect on GDP per person from an increase in skilled migration is that the workforce is now a little more highly skilled on average, making its production more valuable. The third positive effect is that, eventually, consumer prices don’t rise as much as they would have, which increases incomes in real terms.

But offsetting those three positive effects - according to the PC’s very conventional analysis - are three negative effects. The first is that when the country suddenly gets more workers, those workers have to be supplied with additional physical capital (machines) to work with. That is, immigration leads to a need for ‘capital widening’. If the extra equipment isn’t forthcoming, we suffer a problem called ‘capital dilution’ - the amount of capital available per worker falls, which means the economy’s ratio of capital to labour falls, which means the productivity of labour falls. To the extent this happens, real income per person falls.

The second negative effect arises from the likelihood that a far bit of the extra physical capital our businesses need to avoid capital dilution will end up being supplied by foreign investors. The return that has to be paid to these foreign investors - in interest and dividends - represents a loss of income to Australian residents. So immigration will have the effect of adding to our current account deficit and foreign debt. The third negative comes from the model’s assumption that the bigger economy involves more exports and more imports, but while the prices we pay for those imports are unaffected, to sell more exports we have to accept slightly lower prices, meaning a deterioration in our terms of trade, which reduces our real national income.

That’s all very technical and hard to understand, and based on all the assumptions of the neoclassical model, many of which are wrong or misleading. For instance, I doubt that it takes sufficient account of the effect of the extra pressures migration creates for the public sector: the extra public infrastructure needed to meet the needs of the bigger population and the greater demands on the budget for services provided to immigrants and their families. This implies a need for higher taxation - paid by the original residents, not just the immigrants. And any delay or foul-up in providing the extra housing, roads, public transport, utilities, schools and hospitals etc could have significant negative effects on road congestion and other aspects of our amenity.

Even more significant, conventional economic analysis abstracts from the effect of economic activity on the natural environment, essentially assuming the environment to be a free good. Only when specific effort is made to ‘internalise’ environmental externalities - such as through an emissions trading scheme - do they enter into the model’s calculations. So these modelling results would take no account of the increased environmental costs generated by immigration-fed population growth: the increased emissions of greenhouse gases, the greater pressures on water, land quality, fish stocks and the destruction of species. All these very real costs - which eventually would feedback disastrously into GDP - are ignored in conventional analysis.

Now let’s take a different tack. When you ask why we in the developed countries should continue pursuing economic growth when the evidence says it does nothing to increase our subjective wellbeing, the best answer you get back is that we need continued economic growth to create the additional jobs needed to cope with a growing population. That is, if the population’s growing the economy needs to grow or we end up with ever-rising unemployment. This is a strong argument, but it loses its force in our world of an ageing population and a fertility rate that’s below the replacement rate of 2.1 babies per woman.

But in the present population debate the argument coming from the pro-growth side is the reverse: rather than arguing we need economic growth to cope with population growth, people such as the prominent demographer Professor Peter McDonald of ANU are arguing we need population growth to keep up with economic growth. The economy is growing strongly as we seek to exploit the super-high prices China and the world are willing to pay for our coal and iron ore. This growth is increasing employers’ demand for labour at a time when the unemployment rate is low and we’re close to full employment. High immigration is filling that demand, as well as helping to supply the growing labour needs of the mining states without them having to bid their wages up to persuade workers in other states to move to the backblocks of Western Australia and Queensland. In other words, if the economy’s demand for labour is outstripping the local population’s ability to supply that demand, but the government were to decline to allow more workers into Australia, the result would be a wage explosion as employers sought to attract the workers they need by bidding them away from other employers, which would soon lead to rapid inflation - which the Reserve Bank would respond to by greatly increasing interest rates so as to avoid inflation and trying to keep the economy comatose.

So McDonald’s argument is: the government doesn’t control the level of immigration, the economy does. Over the years the rate of immigration has gone up and down, and you can see a strong correlation with the ups and downs of the business cycle. More people come (and are let in) when the economy’s booming; fewer people come (or are wanted) when the economy’s in a slump.

It’s good to be reminded that economic growth is essentially endogenous. Governments use their fiscal and monetary policies to smooth the rate of growth, not to cause it. The micro-economic policies they pursue can encourage or discourage growth to some degree. But, fundamentally, the economy grows because businesses in free markets are always seeking out new ways to make a quid.

All this says that, as the economy is presently configured, it’s difficult for governments that have long courted economic growth to refuse to provide to the economy the immigrant labour it needs to avoid serious overheating.

But to acknowledge this difficulty is not to detract from my earlier point: all of this argument has proceeded on the conventional assumption that the environmental consequences of our economic actions can be safely ignored, to be thought about another day. So if our economy is presently configured in such a way that we can’t keep it functioning stably without doing additional damage to our natural environment - without exceeding the land’s carrying capacity - then the economy needs to be re-configured to put it onto a basis that’s ecologically sustainable. If it’s presently working on a basis that’s unsustainable then, by definition, things can’t continue the way they have been.