Book review: The Skeptical Economist

In his 1987 book On Ethics and Economics, Nobel Prize winning economist Amartya Sen wrote that "economics, as it has emerged, can be made more productive by paying greater and more explicit attention to the ethical considerations that shape human behaviour and judgement". That is essentially the goal of Jonathan Aldred's 2009 book, The Skeptical Economist. The ethical underpinnings of economic models are largely left unexamined, so Aldred (and before him, Sen) makes a strong point that it is useful to realise that economic models and theories are not entirely value free, despite many (or most) economists' hopes (or assertions) that economics is a positive science.

However, despite the importance of the topic area, I found Aldred's treatment to be somewhat uneven. There are good parts to the book, but also some parts that are not so good. Unfortunately for the reader, the bad parts occur earlier in the book. For example, the start of Chapter 2 paints economics as 'the dismal science', a common trope, noting that:

One reason is that many economists are profoundly cynical about human behaviour and the motivation that underlies it.

However, to paint economics as the dismal science for such a reason is to ignore the real origins of the term. In his 1849 book Occasional Discourse on the Negro Question, Thomas Carlyle labelled economics the dismal science because economists were not in favour of reintroducing slavery in the West Indies (see Wikipedia on this point). It is unlikely that Aldred doesn't know the origins of the term, given the other explorations of economic history in the book, so it is clearly being used to cast aspersions at the field.

Sadly, from there the book proceeds to outline a number of the (well-known) limitations of the assumption of rational choice. However, I am never persuaded by authors who launch into these critiques. Often, as is the case there, the argument is essentially that because rational choice can't predict every person's behaviour at all times, there is nothing that can be learned from it. However, any alternative that is presented can equally not predict every person's behaviour at all times, so it seems to me that we start from something that works well in the aggregate, and noting its limitations.

Aldred has a particular dislike of the idea of consumer sovereignty, and he does make some compelling points. However, not everything that he says is accurate on this point either. For example, he argues that economics says that advertising cannot change consumer preferences, which are fixed. As a counterpoint, I offer this post of mine, which uses the tools of consumer choice theory with the assumption that advertising does change consumer preferences.

Finally, Aldred extensively critiques cost-benefit analysis. Again, some of his criticism is warranted, including that cost-benefit analysis is based on excessive quantification. However, his proposed alternative is the precautionary principle, where the focus would be on choosing "the best worst-case outcome". But, how is one to know the best worst-case outcome, without in some way ranking the alternatives in terms of their outcomes, which brings us right back to quantification of the outcomes?

Despite those gripes, the book has lots to offer, and I have made a number of notes of things to add or change in the papers I teach. I especially liked the argument that Aldred proposes against the 'ownership principle' in relation to taxes and earnings. Aldred also does an excellent job of making the reader think about the ethics underlying the economic theories and models. That was the purpose of the book, and if it was stripped of the unnecessary critique of rationality at the beginning, and focused more on presenting the ethics, I think it would have been a much better book.

Overall, I think economics critics would benefit more from this book than the average reader. However, students with concurrent interests in philosophy and economics would probably benefit the most. The book may not necessarily persuade, and it did not persuade me, but it does make you think.

Adrian Orr on New Zealand's Phillips Curve

In macroeconomics, the Phillips Curve (named after the New Zealand economist Bill Phillips) depicts the relationship between inflation and unemployment. In the traditional macroeconomic textbook view, this relationship is downward sloping: for a given set of government policy settings and consumers' expectations about future inflation, lower unemployment is associated with higher inflation. This is shown in the diagram below. Say that the economy starts at some point A, where unemployment is equal to UA and the inflation rate is πA. If unemployment decreases to UB, the economy moves along the Phillips Curve in the short run to point B, and inflation increases to πB.

However, there are a couple of things to realise about the Phillips Curve. First, it doesn't show a causal relationship. Lower unemployment doesn't cause higher inflation. This is an empirical correlation that can be explained through other mechanisms. For example, if aggregate demand increases (such as from increased consumer demand, increased investment by businesses, increased government spending, increased exports, and/or decreased imports), then the domestic economy is producing more. To produce more, firms need more workers, so employment increases (and unemployment decreases). This increases the demand for workers, which pushes up wages (or, alternatively, workers have relatively more bargaining power than before, and can demand higher wages). Wage increases lead to increasing costs for firms, who pass on those costs to consumers. This increase in prices leads to higher inflation. So, as you can see, there isn't a direct relationship between inflation and unemployment. It is changes in one or more of the components of aggregate demand that cause changes in both inflation and unemployment, and make them appear to be related.

The second thing to realise about the Phillips Curve is that, in the long run, it is vertical. That is because as firms' costs rise (because of higher wages) they cut back on production and employment. So, in the long run, there is no trade-off between inflation and unemployment. All that happens is that the economy returns to the natural rate of unemployment (which is UA in the diagram above). However, consumers' expectations about future inflation may now have increased, leading them to ask for greater wage increases in future. In that case, the short-run Phillips Curve would move upwards.

That all brings me to this article from the New Zealand Herald earlier this week, which outlines the Reserve Bank governor Adrian Orr's views of the future trajectory for the New Zealand economy:

Orr warned that the interest rate hikes needed to beat inflation would mean higher unemployment.

"Returning to low inflation will, in the near-term, constrain employment growth and lead to a rise in unemployment," he said.

"The actual extent of this trade-off remains unclear, however, given the significant labour shortages globally and the very different means of employment being adopted post-Covid."

"Importantly, it is highly unlikely that we are at maximum sustainable employment if inflation is still high and variable," he said.

As the Reserve Bank increases the Official Cash Rate (OCR), that will reduce aggregate demand (both through reducing consumption, and reducing investment). Orr clearly expects this to have the opposite effect that I described above, decreasing inflation but at the cost of higher unemployment. The Reserve Bank needs to act fast, which is why we've seen a succession of increases in the OCR. The longer the Reserve Bank takes to act, the more that higher inflation will seem like the norm, and the more likely it will be that the economy will end up back at the natural rate of unemployment, but with semi-permanently higher inflation.

Progressive taxation as an automatic stabiliser

Fiscal policy is the umbrella term used to refer to the government's plans for taxing and spending. It has an impact on the macroeconomy, because government spending becomes income in the hands of households and businesses. So, some people argue that the government can use changes in taxes and spending to counteract the business cycle. This is referred to as countercyclical fiscal policy. Under this approach, when the economy is in recession, the government should tax less and spend more, but when the economy is in an expansion, the government should tax more and spend less. If this worked, then the economic fluctuations of the business cycle would be dampened.

However, it may not be necessary for the government to be quite so interventionist. The economy has some automatic stabilisers, that automatically (hence the name) adjust to increase household incomes when the economy is in recession, and decrease when the economy is in expansion. One example of an automatic stabiliser is the unemployment benefit. In a recession, more people are unemployed and this increases the number of unemployment benefit claims, reducing somewhat the negative impact of the high unemployment on spending. When the economy is in expansion, fewer people are unemployed and there are fewer unemployment benefit claims, reducing the amount of stimulus provided by the unemployment benefits.

To be fair, the unemployment benefit doesn't provide enough stabilisation on its own to substantially reduce the size of business cycle fluctuations. But the unemployment benefit is not the only automatic stabiliser. Progressive income taxation may be another.

A progressive income tax is one where the marginal tax rate (the amount of the next dollar that is paid in tax) is greater than the average tax rate (the proportion of total income that is paid in tax). A typical tax schedule that leads to progressive income tax is a graduated income tax, like that employed in New Zealand, where there are specific income bands that have different marginal tax rates, with higher marginal tax rates within higher income bands.

That looks something like the following graph, which uses the current income tax rates for New Zealand. The blue solid line shows the marginal tax rate, and the red dotted line shows the average tax rate. Notice that the marginal tax rates jumps up at regular intervals (this is a graduated income tax). Above the first income tax threshold, the average tax rate is always below the marginal tax rate. This is a progressive income tax. Also, notice that no one pays 39 percent income tax (which is a point that I have made before). Even at the highest income shown in the diagram ($250,000), the average tax rate is just over 31 percent. That's because, even for those at the highest incomes, their first $14,000 is taxed at 10.5 percent, the next $34,000 at 17.5 percent, and so on. Only each dollar above $180,000 in income attracts a marginal tax rate of 39 percent.

Anyway, coming back to progressive taxation as an automatic stabiliser, when the economy is in an expansion, incomes rise, and more taxpayers will find themselves paying more tax (because they move to the right along the diagram above). In fact, because of the progressive nature of the tax system, the percentage change in taxes is bigger than the percentage change in income. A taxpayer moving from $50,000 to $55,000 in income (a 10% increase) will go from paying $8,020 in tax to paying $9,520 (an 18.7 percent increase). This also works in reverse. When the economy is in a recession, incomes fall, and more taxpayers will find themselves paying less tax. So, a taxpayer moving from $50,000 to $45,000 in income (a 10% decrease) will go from paying $8,020 in tax to paying $6,895 (a 14.0 percent decrease). The same applies to other levels of incomes and income changes that we might consider.

So, when the economy is in recession, progressive taxation removes less from household incomes, and when the economy is in expansion, progressive taxation removes more from household incomes. This will act to reduce the size of economic fluctuations of the business cycle, making progressive taxation an automatic stabiliser.

[HT: John Quiggin, in his book Economics in Two Lessons, which I reviewed here]

Read more:

• Tax cuts, and misinterpreted average and marginal tax rates

More on social security and work disincentives

Duncan Garner had an interesting article in the National Business Review yesterday (gated), on work disincentives associated with social welfare (or social security). And interesting timing, given that I had just written about this a few days ago (see here). Garner wrote:

Until recently, Eric was on the DBP with three kids and was paid $850 a week by Work and Income. They lived in a state house and paid $125 a week because it’s income-related rent. 

Life wasn’t easy but they could get by and Eric could drop off and pick up his kids before and after school and he was in control. Sure, the struggle was real but the state was there for him. 

But he hated the example it set his kids and wanted to show them he went to work each day and paid his way...

So Eric picked up a 40-hour truck driving job and was slowly removed from the welfare system. 

He was paid just over $30 an hour for the truckie job, which is well above the minimum wage and the new job took him all over Auckland. But then his state house rent went up by close on $200 because his income had gone up too. 

Then came the killer blow. How was he to pay for the kids after-school care? In reality he’d never paid a cent for care before because it was always his job, as a solo dad on the DPB. 

But now it could add another $200 to his weekly outgoings and, once you add the extra housing costs, it soon showed he was worse off working, by about $200 week.  

He was better off signing back on to the DPB. He hasn’t done that and wants to make paid employment work.

This again illustrates the problem of high effective marginal tax rates. The effective marginal tax rate (EMTR) is the amount of the next dollar of income a taxpayer earns that would be lost to taxation, decreases in rebates or subsidies, and decreases in government transfers (such as benefits, allowances, pensions, etc.) or other entitlements. In this case, Eric earns more as a truck driver, but then gives up more in lost welfare and entitlements (including the new obligation to pay after-school care) than what he would gain in higher income. On a purely monetary basis, he is worse off.

Interestingly, Eric notes that there are substantial non-monetary benefits from working, and those offset the net monetary loss. Not everyone would feel that way, and that's why high effective marginal tax rates provide such a disincentive for working. I don't necessarily agree with all of the broader points that Garner makes in his article, but on this we do agree:

We need to redesign welfare so these perverse outcomes don’t take hold.

Read more:

• Welfare programmes, effective marginal tax rates, and the US marriage penalty
• Work disincentives, and the income and substitution effects in social security