How important is the 'size of government' component of economic freedom indexes?

A few months ago a guest blogger on ‘The Baseline Scenario’ blog, StatsGuy, wrote a post entitled ‘Good Government Versus Less Government’. It was described as a ‘must-read’ in a post by Tyler Cowen on ‘Marginal Revolution’ and received a great deal of attention on a range of other blogs including Scott Sumner’s (here).


StatsGuy draws attention to the fact that the size of government component of the Heritage Foundation index of economic freedom is negatively correlated with the other components of this index. He concludes that the Heritage Freedom index is really a composite of measures that get at two different things: good government, and less government. His bottom line:

‘Overall, the Good Government factors tend to dominate, and drive a lot of the correlation with good economic and quality of life outcomes. When one splits out the factors, the case for Less/Weaker Government weakens substantially, and the case for Clean/Non-Corrupt/Efficient government strengthens considerably’.

Some other researchers have similarly objected to the inclusion of size of government in economic freedom indexes. For example, Peter Lindert describes this as ‘guilt by definition’ on the grounds that it tends to make big government and the welfare state look bad merely by describing this national attribute as contributing to lower economic freedom (‘Welfare states, markets and efficiency: the free lunch puzzle continues’, 2007: 6).

At least one contributor to the discussion of StatsGuy’s post made the point that if economic freedom has two different dimensions, a lack of correlation between those dimensions does not necessarily mean that one of them is irrelevant. For example, it is possible for both size of government and quality of government to be important to economic growth.

I recently had an opportunity to test whether this is so in preparing a background paper for the 2025 Taskforce, which was established by the New Zealand government to advise how average incomes in that country could be raised to equate those in Australia by 2025. The analysis provides some support for the view that size of government is an important component of economic freedom indexes.

The analysis uses the Fraser Institute’s index of economic freedom because this provides a consistent measure of institutional quality over a longer time period than the alternatives. The data set relates to ‘advanced economies’ as defined by the IMF – this data set includes high income jurisdictions with small governments, such as Hong Kong and Singapore, as well as OECD countries. The regression, based on panel data, explains average per capita GDP growth in each decade in terms of several variables including two components of economic freedom at the beginning of each decade, the size of government index and ‘other economic freedom’. The relevant regression results are presented in Table A 2.3, p 43 (the right hand column).

The coefficients on both the size of government and ‘other’ economic freedom variables were significantly greater than zero - suggesting that smaller size of government has a positive effect on economic growth. The magnitude of the estimated coefficient on size of government is about half that on ‘other’ economic freedom, but that is about twice as large as I had expected it to be on the basis of the weight of size of government in the economic freedom index (20%).

One fairly obvious question that might be asked is that if size of government is so important, how is it that some countries with big governments, an obvious example is Sweden, have managed to maintain relatively strong economic performance. I have attempted to answer this question in the chart below which compares per capita incomes in Sweden and Australia (Penn World Tables, rgdpch with some extrapolation using IMF growth estimates). The results of the simple analysis presented in the chart suggest that if Sweden had not undertaken substantial economic reforms (including some improvement in the size of government component of economic freedom as well as other components) it would have performed poorly. The chart also suggests that Sweden’s economic growth performance could have been much better if it had a smaller government.

This analysis doesn’t tell us that every country could become a paradise if only it had a small government, or that countries with big governments are dreadful places to live. It just suggests that big government is not a free lunch . The lack of correlation between the size of government and other aspects of economic freedom is interesting, but it doesn’t mean that size of government doesn’t matter.

Postscript:

A colleague has suggested that I should have mentioned that the size of government of most of the countries included in this ‘advanced’ country data base used in the regression analysis reported above is now much larger than it was prior to the 1960s.

It might also be worth noting that recent World Bank research suggests that in low income countries government spending on infrastructure, education etc. can have a positive effect on economic growth if (a big if) the countries concerned have favourable institutional characteristics.

Can New Zealand catch up to Australia?

Is New Zealand disadvantaged by economic geography to such an extent that it cannot hope to catch up to Australia’s average income levels, even with further improvements in institutions and policies? That is probably the most important question considered in the second report of the 2025 Taskforce that was released a few days ago.


The 2025 Taskforce was set up by the New Zealand government after the 2008 election to recommend how the gap between average incomes in Australia and New Zealand could be closed. Incomes of New Zealanders have generally risen less rapidly than those of Australians over the last 40 years, resulting in a gap between average incomes of around 35 percent in recent years. After the 2008 election, the NZ government committed to closing this income gap by 2025.

Since the Taskforce presented its first report last year, Philip McCann - an economist with expertise in economic geography – has advanced the view that New Zealand’s geographical disadvantages prevent it from becoming a high productivity economy. McCann has implied that structural features that are advantageous in the current era of globalization differ so much from those exhibited by New Zealand that this economy could not reasonably be expected to have relatively high productivity. He suggests ‘this is true irrespective of the degree of flexibility in the domestic labour market, the degree of transparency in the local institutional environment, or the levels of cultural aspirations for success’ (‘Economic geography, globalisation, and New Zealand’s productivity paradox’, New Zealand Economic Papers, Dec. 2009: 299).

The particular aspect of geography that McCann considers to be most disadvantageous to New Zealand is its relative lack of agglomeration economies associated with large cities. These agglomeration economies arise from knowledge exchanges, better networking and coordination, a nursery role for new enterprises, improved labour market matching processes and greater competition.

McCann argues that agglomeration economies can explain the decline in New Zealand’s per capita incomes relative to Australia because of the way the world has changed. One strand of the argument has to do with the increasing importance of knowledge-intensive activities that can often be undertaken at lower cost where face to face contact is possible among the various participants. Another strand is that with closer economic integration between Australia and New Zealand the economy with relatively larger agglomeration economies, i.e. Australia, has become a relatively more attractive location for capital investment and employment of highly skilled workers.

McCann sums up: ‘ ... although New Zealand underwent fundamental institutional reforms in the 1980s and 1990s, at exactly the same time as this was taking place the landscape of global economic geography was shifting in favour of other places. It may well be that the deregulatory reforms limited some of the most adverse aspects of these shifts, thereby minimising the productivity gap. Yet the point still remains that the world changed, and the world of the late 20th and early 21st centuries is very different from the world that provided New Zealand with almost a century and a half of productivity advantages’ (p. 300).

How does the Taskforce respond? The Taskforce acknowledges that both New Zealand and Australia have been disadvantaged by geography. It notes that according to recent OECD research the impact of greater distance to markets is equal to around 10 percent of GDP per capita for both countries. However, it judges the evidence in support of the view that New Zealand’s small population limits the potential to obtain agglomeration effects to be weak. In particular, Auckland’s position within the regional hierarchy of Australasian cities is not declining – the population of Auckland has been growing faster than the populations of Sydney and Melbourne. The Taskforce also points out that there is no evidence that New Zealand suffered an adverse shock from globalization during the 1980s; that migration from New Zealand to Australia is disproportionately of highly skilled workers as agglomeration theory implies; or that the relative performance of small countries has declined in the past 20 years.

The Taskforce concludes: ‘... modern growth theory provides stronger support for the importance of institutions and policy than it does for geography, especially in the deterministic interpretations of economic geography’ (p. 41).

Sitting in Australia, current concerns in public policy discussions about the emergence of a two-speed economy in this country make the agglomeration theory of relative decline in New Zealand’s economic performance seem rather odd. Rather than a concern that agglomerations centred on Sydney and Melbourne are leaving the rest of Australia behind, the main concern is that New South Wales and Victoria (along with other states) are being left behind as economic growth steams ahead in Western Australia and Queensland, as a result of rapid expansion of the minerals sector and related industries. There is also reason for concern that, over an extended period, the particularly poor performance of the New South Wales government has detracted from the substantial location advantages that Sydney should enjoy.

If we reject the idea that Australia’s alleged agglomeration advantages make it impossible for New Zealand to close the income gap, where does that leave us in terms of explaining New Zealand’s relatively poor economic performance? The Taskforce pours cold water – correctly in my view - on another geographical explanation, namely Australia’s good luck in having plentiful supplies of mineral resources to export to rapidly growing markets in China and India. It is only in the last few years movements in Australia’s terms of trade have been much more favourable than in New Zealand. Moreover, New Zealand also has substantial mineral and hydrocarbon resources.

I think that leaves us with having to explain New Zealand’s relatively poor economic performance in terms of policies that are less favourable to economic growth. That also poses a problem because the impression given by various international comparisons of institutions and policies is that since the mid-1990s there has not been much to choose in overall terms between the economic policy environments in New Zealand and Australia. It seems likely, however, that New Zealand has not performed so well in the areas that have mattered most from a growth perspective. For example, one major problem discussed by the Taskforce is the effect of relatively high levels of government spending in discouraging investment in export industries - via impacts on the real exchange rate as well as tax rates.

The Taskforce has expressed the view that closing the gap in average income levels by 2025 will require policies that are superior to those in Australia in their focus on growth. It seems to me that those who believe that New Zealand has geographical disadvantages should logically be strong supporters of that view (unless they reject the objective of closing the income gap). The greater the geographical disadvantage, the greater the policy superiority New Zealand will need in order to meet the objective of closing the income gap by 2025.

How much does over-work affect happiness?

The results of a survey conducted recently by the Australia Institute apparently shows that half of Australians (61 per cent of those working overtime) were prevented from spending enough time with family in the preceding week as a result of over-work. According to the press release (which is the most detailed description of the study I could find) a lot of people don’t have time to exercise, eat healthy meals or go to the doctor when they should.

If we take the results of this survey at face value it would appear that over-work is a huge problem in Australia. I suspect, however, that the problem or over-work is not as widespread as the Australia Institute suggests. I also suspect that over-work has a much smaller adverse impact on happiness than does under-work.

Cartoon by Nicholson from "The Australian" newspaper: www.nicholsoncartoons.com.au

The results of a study by Bruce Headey, Ruud Muffels and Gert Wagner, based on a long-running German panel survey, shows working hours to be one of the factors that has a long-term impact on life satisfaction. One of the things I like about the study is that the variable used is a measure of the extent to which respondents achieve their preferred tradeoff between work and leisure, rather than divergence of working hours from some arbitrary standard chosen by researchers. The relevant variable was the gap between the number of hours a week respondents said they would prefer to work and the number of hours per week they actually work. Those who worked over 3 hours per week more than they preferred were treated as overworked and those who worked over 3 hours per week less than they preferred were treated as underworked (‘Long running German panel survey shows that personal and economic choices, not just genes matter for happiness’ PNAS, 2010).

The results indicate that the negative impact of under-work on life satisfaction was about four times greater than the negative impact of over-work. The authors suggest that this ‘is presumably because lost consumption rankles worse than lost leisure’. (It would seem that the regression analysis does not control for income levels.) The study suggests that the negative effect of unemployment is much worse than that of either over-work or under-work (about four times greater than for underwork).

Some of the other results of the study might help further to put these findings into perspective. The study shows that social participation – a measure of frequency of meetings with and helping out friends, relatives and neighbours – has a substantial positive effect on life satisfaction of around the same magnitude as the negative effect of under-work. The positive effect on life satisfaction of frequent exercise is of about the same magnitude as the negative effect of over-work. The adverse effect of having a neurotic personality is about ten times greater than that of being overworked, but having a neurotic partner has only about half the adverse effect of being overworked.

What should we make of these findings? One obvious qualification is that it isn’t clear to what extent they might apply outside Germany. Leaving that aside, it seems to me that the most important implication is the importance to individual happiness of having the opportunity to work as many hours as the individuals concerned want to work. Under-work is not as bad as unemployment, but it is likely to be a much worse problem for the individuals concerned than is over-work.

It is hard to see how anyone could argue that overwork could be a huge problem when people are free to choose among jobs on the basis of hours of work along with other employment conditions. Some individuals may make bad choices, allowing themselves too little time for social participation and exercise, but that is not a systemic problem.

Update:
The Australia Institute report has now been published. It has some features that I like e.g. over-work and under-work are judged relative to respondent's desired hours of work. However, I think the findings on percentages prevented from spending enough time with family tend to overstate the extent of the problem. The percentages who always feel rushed or pressured for time are relatively low. It seems that having children living in the household is as great a source of time pressure as working overtime (Figure 4). Finally, the study doesn't tell us much about the emotional well-being of those who work more or less than their preferred number of hours. This seems to me to be an area where data on life satisfaction can be useful.

Postscript:
I have posted more information on this subject, including subjective well-being data for Australia, in a more recent post entitled: Is work-life balance a big problem in Australia?

How should we respond to a small risk of catastrophe?

I try to remember to pay house insurance premiums. Otherwise, I tend to avoid thinking about small risks of catastrophe. There are plenty of other things to worry about.

This avoidance strategy usually helps me to maintain a positive state of mind - atleast, until someone manages to ambush me with the thought of how dreadful it would be if one of those catastrophes actually occurred.

The last time such an ambush had a lasting impact on me was in March this year when I was reading ‘The Science of Liberty’ by Timothy Ferris. This book contains an excellent discussion of the historical links between liberty and the advance of knowledge. I tend to trust Tim Ferris’ judgement on scientific matters. (I have previously written about his book here.)

Tim Ferris’ discussion of the science of climate change begins in a fairly low key fashion until he reaches the point where he suggests that greenhouse gases generated by human activity constitute the most plausible explanation for the gradual increase in the earth’s average temperature since the beginning of the 20th Century. Then, in the following paragraph, he proceeds to suggest progressively less benign consequences of further global warming until, at the end of the paragraph, he mentions the possibility of runaway warming. The next paragraph reads:

‘On which point it may be useful to contemplate Venus, the brightest planet in the sky. Venus is virtually Earth’s twin – the two planets have the same diameter and the same mass – but while much of the earth’s carbon is bound up in its oceans and plants and in fossil fuels like coal, oil and natural gas, the carbon on Venus resides in its atmosphere. The surface temperature on Venus is 457 degrees Celsius, hot enough to melt led. Should the earth be pushed into runaway greenhouse warming, it might wind up resembling the Venus of today’ (p. 282).

This wasn’t the first time I had heard about the possibility that Earth’s future could be like Venus. On previous occasions, however, it was obvious that scare tactics were being employed and my defences were activated well before the Venus card was played.

Cartoon by Nicholson from "The Australian" newspaper: www.nicholsoncartoons.com.au

I was reminded of Tim Ferris’ invitation to contemplate Venus while reading ‘The Rational Optimist’. Matt Ridley, the author of this book, adopts a very different position. He begins the discussion by mentioning Martin Weitzman’s argument that if there is some possibility of a huge disaster resulting from global warming, the world should take steps to avoid it. He then suggests that the problem with this reasoning is that it applies to all risks, including the remote possibility of collision with a large asteroid.

I agree with Robin Hanson’s view, in his review of Ridley’s book, that some action may be warranted to reduce the potential impact on human well-being of any potential catastrophe. How we should respond should depend on the nature of the potential catastrophe, the probability that it will occur and what can be done to avoid it.

How should we respond to the small risk of runaway global warming? A fairly obvious answer is to put a tax on carbon emissions in order to provide incentives for development of technologies that generate less CO2, accompanied by an appropriate subsidy for activities that remove CO2 from the atmosphere. In many countries it would be possible to do this at little or no economic cost by substituting a carbon tax for other taxes that impose greater economic distortions. It is important to emphasize, however, that the main aim of the exercise should be to put in place incentives for development of better technologies.

Matt Ridley makes a strong case that climate mitigation is currently being mismanaged and that this mismanagement could be more damaging to human well-being than climate change itself. By encouraging a return to the medieval practice of using biofuels as an energy source, governments have added to misery in poor countries by raising the price of grains and have provided incentives for the further destruction of rainforests. In addition, incentives for greater use of costly wind and solar technologies are raising the cost of electricity substantially and/or requiring substantial subsidies from taxpayers, for little benefit in terms of reduction in CO2 concentrations in the atmosphere.

It is possible that solar technology will become competitive at some time in the future, but subsidizing use of current solar panel technology will not make that happen. If solar panel technology ever becomes competitive it will not need to be subsidized to enable scale economies to be achieved.

I think the current mismanagement of climate mitigation is attributable to scare tactics and panic. Some of us have grown so accustomed to environmental scare tactics that we find it difficult to take seriously the idea that a small risk of catastrophe is worth considering. Others have been too easily panicked into supporting costly policy responses that seem to be directed toward reducing CO2 as rapidly as possible irrespective of cost. The outcome of these conflicting forces in Australia has been a policy to encourage increased use of existing renewable energy technologies that are still highly inefficient. This policy achieves only a small reduction in CO2 emissions per additional dollar spent. (My test of how genuinely concerned enironmentalists are about reducing CO2 emissions in the short term is whether they are in favour of the nuclear power option, which seems to be the best alternative to use of fossil fuels that is presently available. )

There are signs now emerging that people in Australia are becoming concerned about the cost of rising energy prices attributable to the silly policy of encouraging greater use of high cost renewable energy. Hopefully similar concerns in other countries will result in adoption of sensible strategies to encourage development of less costly technologies over the next few decades.