Spending and Growth A response to David Laws David Howarth
David Laws has recently received much favourable publicity in the Conservative press for advocating further spending cuts and tax cuts. He wrote: Future UK governments should consider a further substantial real rise in the personal tax allowance, along with lower marginal rates of tax at all income levels. This can be paid for over time by continuing to reduce the share of public spending in GDP [E]ven after the existing fiscal consolidation, state spending will account for some 40% of GDP, a figure that would have shocked not only Adam Smith, Gladstone and J.S. Mill, but also Keynes and Lloyd George. The implication of the state spending 40% of national income is that there is likely to be too much resource misallocation and too much waste and inefficiency. (David Laws, The Orange Book: Eight Years On, Economic Affairs 32:2 (June 2012) 31 at 34). Perhaps the enthusiasm of readers of the Tory press might have been dampened if their journalists had allowed them to read some of the other things Laws wrote, for example: If economic liberalism has proved itself over time as the best guarantor of wealth creation, it has proved rather less successful in delivering the society of opportunity that many liberals would like to see. Too often, free market capitalism has been associated with gross inequalities of wealth, income and opportunity. No liberal can be content to live in a society where life chances are determined more by family background and parental income than by natural ability. Milton Friedman claimed in his famous book Capitalism and Freedom (1962) that capitalist societies would be meritocracies, in which social mobility would be high and in which everyone would enjoy opportunity. While it is true that most liberal societies are increasingly meritocracies where people are judged on their personal worth and not on their race, class, creed, sex or sexuality, the sad fact is that the chances of acquiring merit are grossly unequal. (ibid.) But leaving to one side the ideological debate about meritocracy, Laws assertion that public spending at 40% of GDP leads to too much resource misallocation and inefficiency, although itself rather imprecise and politically calibrated for serious analysis ( too much for what and compared to what?), does prompt the question of what we know, as a matter of empirical fact, about the relationship between public spending and economic growth. Is Laws right to imply that reducing the overall share of public spending in GDP will lead to greater prosperity?
In 2007, I wrote: Despite much research effort, the fundamental position is still as Jonathan Temple stated it in 1999: In political discussion it is common to hear claims that a high ratio of social security transfers to GDP and a high level of government consumption can be damaging to growth prospects. The evidence is not strong. Some researchers find a negative link between government consumption and growth, but overall studies disagree, and it would be wrong to argue that a correlation between small government and fast growth leaps out from the data. Work continues on whether particular types of government expenditure are likely to have positive or negative effects on conventional economic growth, but industrial subsidies seem to be a more likely source of economic failure than social security payments. (Duncan Brack and Ed Randall (eds), Dictionary of Liberal Thought (2007) at 103) The research in the area has moved on since then. The overall assessment that the correlation between reducing public expenditure and encouraging economic growth is not as clear as the political right would have us believe still stands, but there is more to say, and some emerging themes of the research, especially that of Norman Gemmell, Richard Kneller and Ismael Sanz in their so far unpublished study Does the composition of government expenditure matter for economic growth?, is worth reporting to a wider audience. The first point is that, although the relationship between higher public spending and lower growth rates remains weak and not statistically significant, stronger relationships appear if we differentiate between different ways of funding higher spending. Roughly speaking, if higher spending is funded by taxes on consumption, no adverse effect on long term growth rates can be demonstrated, but if it is funded by direct taxes an adverse relationship does appear. Deficit funding, significantly for present political purposes, also has an adverse long-term effect, but a much lower one than that of direct taxation. Clearer results are also starting to emerge about different forms of public spending. Again roughly speaking, public expenditure on transport and communications infrastructure raises long term growth rates, as does expenditure on education. That much has been suspected for a while, but researchers are now starting to find evidence that public expenditure on health is also associated with higher long term growth, although the effect is only about a third of that of expenditure on education.
Spending on housing and defence, however, has no statistically significant long-term effect. The only statistically significant negative relationship for a category of public spending found by Gemmell, Kneller and Sanz is for welfare spending (as we have been browbeaten into calling social security spending, in obeisant imitation of the US Right), but it is interesting that the effect is quite small a loss, they estimate, of 0.04% in annual growth of GDP for every additional percentage point of GDP spent on social security, assuming that spending on all other functions falls proportionately by an equal amount. Some work has also been done on breaking down the categories of expenditure into more specific types. For example, the work of Vandenbussche, Aghion and others seems to show that education spending works better for countries operating at the boundaries of existing technology (that is to say, highly developed countries) if it is spent on higher education and research, whereas for countries that are not so technologically advanced, and which aim to grow through imitating the innovation produced by others rather than through innovating themselves, expenditure is best directed at lower levels of the education system. What are the policy implications of these emerging themes? Economic policy is not, of course, only about economic growth. Other policy goals such as the environment and social justice are important, and GDP itself is far from a perfect measure of welfare. But even if we look only at the growth issue, although some of these results reinforce David Laws position, in particular his desire to reduce direct taxation, they suggest that he is not right to imply that public spending itself necessarily kills off growth. Some forms of public expenditure are good for growth, regardless of the theoretical objection that public expenditure not guided by the market risks misallocation. Those forms of expenditure include not only transport and communications infrastructure a form of spending both David Laws and David Cameron seem to favour and not only education spending of which Laws also approves but also health spending, a result that gives more comfort to a different strand of Liberal Democrat politics. Another problem with Laws position is that the policy he has promoted of reducing spending on higher education in favour of raising it on school education especially through the pupil premium turns out to be just the kind of misallocation of resources he denounces. In a developed country such as Britain, it amounts to redistributive spending that will eventually cut economic growth.
If we want to finance redistributive spending such as the pupil premium without adversely affecting growth, the options should be restricted to reducing other forms of redistributive spending, reducing spending on relatively economically unproductive functions, such as defence, or by raising expenditure taxes. Unless his long-term strategy is to move Britain back from the frontiers of technology and towards developing country status, the idea that higher education spending is merely a subsidy for the middle-class makes no economic sense. Instead of the rather odd strategy of cutting overall public expenditure while increasing the purely redistributive elements of education spending we might start to think about a different strategy for the period beyond the current era of austerity. That strategy would start with determinedly shifting the composition of public spending towards those forms of spending that tend to produce growth, which includes not only transport and communications infrastructure and higher education, but also health. We could fund increases in spending on those areas without damaging growth, and without increasing the deficit, by increasing consumption taxes. We could further enhance the positive impact of such a financing strategy by concentrating on expenditure taxes that promote other important policy goals, for example environmental taxes. Another option for financing increases in spending on growth-enhancing functions would be to reduce spending on other, less economically productive, areas. The current government is targeting social security spending, and we should acknowledge that the evidence shows that such reductions would probably enhance growth. The question, however, is whether one could justify their effects in terms of poverty and inequality in the light of the size of the gains available in long-term economic growth rates. If 1% of GDP off social security brings in only 0.04% on the growth rate, we are not being confronted by an overwhelmingly impressive return. On those numbers, even if the whole of the budget of the Department of Work and Pensions were to be cut, even assuming, contrary perhaps to reality, that the effects on growth would the same for the last billion as for the first, the long-term rate of growth in GDP would rise by no more than 0.4%. More realistically, the extra 10bn in welfare cuts demanded by George Osborne would change the annual growth rate by an imperceptible 0.027%.
If, by the time the country is in a position once again to consider a long-term economic policy, there remain any growth-strangling non-environmental subsidies for production (perhaps there might be some in defence), they should, of course, go. We should also finally ditch Britain s imperial pretensions and face down the defence industry lobby. Defence spending is not good for long-term growth and those expenditures would be better directed elsewhere. There might also be specific examples of redistributive spending that we might want to suppress, either because they do not work or, more importantly, because they redistribute in an irrelevant way for example, as in the case of some general benefits, to people who are comfortably off anyway, or, as in the case of some benefits for pensioners, from the working poor to the retired upper middle class. But in the end, if we are to spend to enhance long-term growth without starting another cycle of debt, the choice will come down to cutting redistributive spending or raising indirect taxes. Liberals should choose the latter. David Howarth is a Reader in Private Law in the Department of Land Economy and a Fellow of Clare College at the University of Cambridge. He was the Liberal Democrat Member of Parliament for Cambridge between 2005 and 2010.