It is widely believed that governments of more open countries are forced to reduce taxes and public expenditure to attract foreign investment and human capital, or at least to prevent domestic firms from relocating abroad. Although this is certainly possible, the data tell a different story. As first shown by David Cameron (1978), and then by Dani Rodrik (1997, 1998), more open countries have bigger governments. In Epifani and Gancia (2009), we provide more evidence on this puzzling stylised fact. Using a panel of 150 countries observed over half a century (1950-2000), we find a strong positive correlation between trade openness and government size, not only across countries but also over time. In particular, a 1% increase in openness (the ratio of imports plus exports to GDP) is associated with a 0.15% increase in government share of GDP. Together with the fact that openness increased, on average, by 42.5% in the second half of the last century, the growth of trade volumes alone can explain a 6.5% increase in the size of the public sector, more than one-half of its total increase (12%).
So far, the main explanation for this observation is due to Rodrik, who argues that, in more open countries, firms and workers are more exposed to external risk and therefore demand more public insurance. Although plausible, this explanation fails to convince. In particular, if trade openness increases the demand for public insurance, this should show up in a surge of public transfers for social security and welfare. Our evidence shows however that, unlike government consumption, this kind of expenditure is uncorrelated with trade openness. Moreover, terms-of-trade volatility, which has been suggested as a measure of external risk, does not seem to (robustly) affect any kind of government expenditure.
Economic openness and the cost of public spending
We propose a different explanation. More open countries have larger public sectors because the cost of providing public goods is lower the higher a country’s involvement in foreign trade. The basic idea is that an expansion of the public sector crowds out private production, thereby reducing the domestic supply of exports. As long as the world demand for domestic products is downward sloping, a fall in domestic exports brings about a terms-of-trade improvement that partly compensate the increase in public expenditures. In other words, the rise in export prices shifts some of the costs of the public sector onto foreign consumers. This effect is stronger in more open economies, because the real-income effect of terms-of-trade movements is proportional to the volume of trade. Moreover, such a terms-of-trade improvement materialises independent of whether a country is large or small, provided that it produces differentiated goods. For instance, given that Nokia’s mobile phones are perceived as different from Motorola’s, even a country as small as Finland is a price setter in world markets. Hence, insofar as the price of a Nokia mobile reflects high domestic taxes, every unit sold to foreigners provides a subsidy to the Finnish welfare state.
In our data, we find that the positive correlation between openness and government size holds strongly only for countries producing differentiated products. Moreover, simple calibrations suggest that the mechanism illustrated above is quantitatively relevant, as it can explain the entire empirical correlation between openness and government size, and between one-third and one-half of the overall average increase in public spending over the second half of the last century.
That the growth in governments’ size is driven by terms-of-trade considerations may appear implausible at first. It is less so once it is understood that terms of trade and relative wage are closely related in open economy, for any policy that increases the relative demand for domestic labour also affects positively the terms of trade. This is indeed the case with public expenditure. A shift in the composition of expenditure from private to public raises the relative demand for domestic labour, because public goods and services are produced almost entirely locally, whereas private goods are partly imported. Evidence on this is compelling. In a sample of developed and developing countries with available data, we find that the average import share in government consumption is 1%, versus an economy-wide import share equal to 50%.
Hence, the logic behind our results is closely related to the Keynesian view that public expenditure can be used to sustain demand for domestic labour. The current crisis offers numerous examples of how globalisation affects this logic. For instance, when asked how to give a stimulus to the US economy, a PIMCO spokesman said: "Consumers will just buy more Chinese goods with stimulus package money, more of the same. What is needed is public investment to fill demand void of private sector". This is the mechanism we stress in action. In a world of integrated product markets, sustaining demand via public spending is considered more effective than a tax cut.
Restraining fiscal policy externalities
What are the normative and policy implications of our findings? From a welfare standpoint, our explanation is fundamentally different from Rodrik’s. While the expansion of the welfare state to provide more public insurance can be optimal in the presence of higher risk, the terms-of-trade externality we emphasise leads to welfare-reducing overspending. This happens because, in the absence of international coordination of fiscal policies, governments do not internalise the costs they impose on the world economy. What are then the remedies to this inefficient by-product of globalisation? One option could be to extend the WTO rules, currently limiting terms-of-trade manipulations arising from non-cooperative trade policy, to correct the externalities arising from fiscal policy as well. However, expanding the scope of the WTO rules in this direction would be a very difficult task, because fiscal policy is widely perceived as a matter of national sovereignty over which no country is readily willing to accept foreign-imposed constraints. This may explain why market integration and political cooperation have proceeded at an uneven pace.
The EU may appear in this respect an exception, as it imposes constraints on fiscal policy to member states and provides an appropriate institutional framework to achieve fiscal policy coordination. Yet, its rules do not provide a full solution to the fiscal externalities due to globalisation, as they impose limits on governments’ budget deficits and debt, and therefore at most only indirectly alleviate the inefficiency arising for an excessive level of public spending.
Finally, a dismal implication of our results is that globalisation may increase the temptation for policy makers to use public resources for their private aims. This is because the welfare cost of wasteful public spending may be lower in open economy, partly offset by terms-of-trade movements. This suggests that, contrary to the firm belief that globalisation is efficiency-enhancing, it may actually protect or even promote public inefficiency.
Cameron, David R. (1978). "The Expansion of the Public Economy: A Comparative Analysis," American Political Science Review, 72, 1243-1261.
Epifani, Paolo and Gino Gancia, G. (2009). "Openness, Government Size and the Terms of Trade," Review of Economics Studies, 76, 629-668.
Rodrik, Dani (1997). Has Globalisation Gone Too Far?, Institute for International Economics, Washington, DC.
Rodrik, Dani (1998). "Why Do More Open Economies Have Bigger Governments?" Journal of Political Economy, 106, 997-1032.