A service of the

Download article as PDF

Against the backdrop of the considerable increase in protectionist measures that could be observed during the recent financial crisis, this article investigates the following questions: Are “Buy National” clauses a suitable instrument to internalise the externality that arises because the stimulating effect of “general”, non-discriminatory government spending is distributed among many countries, while the cost of the stimulus is borne by only one country? Are short-term trade policies, like import tariffs, export subsidies and “Buy National” policies, suitable policies to mitigate an economic downturn?

When the financial crisis began in 2008, as economic activity collapsed and unemployment rose, protectionism risks intensified. Many countries have resorted to “Buy National” clauses in stimulus packages and have raised trade defence measures, tariff measures and export taxes, among other trade barriers.

This trend is very well documented in Figure 1, which shows the cumulative number of protectionist measures taken, starting from January 2008. As we can see, before the outbreak of the financial crisis in September 2008, starting with the failure and merging of a number of American financial companies, only very few protectionist measures were taken. Since then, we have seen a sharp increase, specifically in 2009 and 2010. In mid-2011 the number of protectionist measures started to stabilise. The most widely used protectionist measures are, in descending order: bail out / state aid measures, trade defence measures, tariff measures, export taxes or restrictions and non-tariff barriers.

Figure 1
Cumulative Number of Protectionist Measures Taken
Lechthaler Abb-1.ai

Data Source: http://www.globaltradealert.org/.

The countries implementing the most protectionist measures since January 2008 in descending order are the Russian Federation, Argentina, India, China and Brazil. The countries most affected by protectionist measures since January 2008 in descending order are China, the United States of America, Germany, France and the United Kingdom. Hence, the countries implementing protectionist measures are not necessarily the ones most heavily affected by protectionist measures. This suggests that retaliation might not be the most important issue. We therefore concentrate our analysis on one-sided protectionist measures but also briefly discuss retaliation.

Additionally, there has been a rise in inward-looking “Buy National” policies, which require fiscal stimulus packages to be concentrated on domestically produced products. Such policies are assumed to be very effective for the protection of local jobs and enterprises from foreign competition. Even very prominent economists, like Paul Krugman, have argued for “Buy National” policies. In a New York Times column1 Krugman argues that there are major policy externalities that can be internalised by “Buy National” policies. The externality arises because the stimulating effect of “general”, non-discriminatory government spending is distributed among many countries, while the cost of the stimulus is borne by only one country. Hence, without international policy coordination, each country’s optimal fiscal stimulus will be less than it would optimally be from a world perspective. According to Krugman, using “Buy National” policies would alleviate this problem because the stimulus is concentrated on the country which is paying the bill. This is the claim.

The questions we want to investigate here are the following:

  • Are “Buy National” clauses suitable to internalise the policy externality?
  • Are short-term trade policies, like import tariffs, export subsidies and “Buy National” policies, suitable policies to mitigate an economic downturn?

Concerning question two, we show that trade policy should not be used as a business-cycle instrument. Concerning question one, according to Krugman, “Buy National” clauses should increase the fiscal multiplier as compared to general government spending. We will show that this is not the case. Actually, the fiscal multiplier is higher for general government spending than for “Buy National” fiscal stimuli. The reason for the negative answer to both questions is that trade policies lead to trade distortions. They induce changes in relative prices which hurt the country initiating the policy. This makes them a less suitable instrument for stimulating an economy during a recession than general government spending.

The Analysis

Before we describe our findings in detail, let us first briefly describe the main features of the underlying theoretical model. For our analysis, we refer to the model of Ghironi and Melitz2 and extend it to take account of certain protectionist measures. This model is a dynamic version of Melitz3, the most prominent theoretical model in the trade literature.

Each quarter new firms will innovate and enter the market. However, setting up a new business is costly and it is thus assumed that entry is subject to an exogenously given entry cost. Hence, new firms enter as long as they can earn positive profits (zero-profit condition). After entering the market, firms learn their productivity and decide whether to serve only the domestic market or whether to additionally export to foreign countries. Since serving the export market is more challenging and more costly than serving the domestic market, only the most efficient firms find it profitable to export.

In the long run trade has to be balanced, i.e. no economy can permanently consume more than it produces. However, in the short run trade may not be balanced. This gives countries the opportunity to react to temporary shifts in economic fortunes. A more detailed description of the theoretical model is given in Larch and Lechthaler4.

We use the model just described to analyse a set of different protectionist measures. Specifically, we analyse a temporary increase in tariffs, in non-tariff trade barriers and in export subsidies and “Buy National”. We compare these protectionist measures with the consequences of traditional, general government spending which does not discriminate between domestic and foreign products. All policies, if necessary, are financed via distortionary taxes on labour income.

The main results are shown in Figure 2. It illustrates the development of GDP and other relevant variables initiated by a temporary increase in each of the measures considered. The effects are illustrated in terms of percentage deviations from the development which would have taken place had there been no stimulus whatsoever. So a point in the positive region for a certain quarter means that the measure increases GDP in that quarter.

Figure 2
The Effects of Protectionism
Lechthaler Abb-2.ai

It is immediately clear that protectionism does not pay off. Although “Buy National” and export subsidies tend to stimulate the economy, traditional government spending is more successful in doing so. Erecting barriers to restrict trade even reduces output and consumption, thus harming the economy instead of protecting it.

Let us discuss each measure in turn. Government spending concentrated on domestic products (“Buy National”) stimulates the economy. However, the effect is smaller than the effect of non-discriminatory government spending. This implies that discrimination against foreign products actually hurts the economy. The reason is simple: “Buy National” hurts consumers and exporters. Consumers are hurt for two reasons. Concentrating the stimulus on domestic products makes the stimulus more expensive, since cheap foreign goods are ignored. Thus one dollar buys less in the case of “Buy National” than in the case of general government spending. Further, the stimulus concentrated on domestic products considerably pushes up the price of domestic products. Consumers react by reducing their consumption expenditure (an effect also known as consumption crowding out).5 Similar to raising trade barriers and tariffs, government spending concentrated on domestic goods distorts international prices. Domestic products become more expensive than foreign products, lose competitiveness and as a result exports go down. Summarising, we find that government spending increases GDP but concentrating the expenditures on domestic products reduces the positive effects and, thus, is not beneficial.

What about other protectionist measures – do they work better? On the contrary. By far the worst measure is raising trade barriers. Although this protects domestic firms from foreign competition and increases domestic production for the domestic market, it considerably hurts the economy as a whole. The reason is twofold. It directly hurts consumers because foreign products in the home country become more expensive. This makes consumers poorer and they react by reducing their consumption of both domestic and foreign products. The distortion in trade also implies changes in relative prices, making domestic products more expensive than foreign products on the export market, thus decreasing their competitiveness. Naturally, this leads to a considerable decrease in exports. It turns out that the two negative effects outweigh the positive protectionist effect and GDP shrinks. The effects of increases in tariffs are very similar, but at least tariffs generate income. This income can be used to reduce taxes or to finance transfers to consumers. The net effect of tariff increases is still negative but less harmful than increasing trade barriers.

If all these measures are so harmful, why are so many politicians still tempted to use them in times of recession? One possible explanation is that this kind of policy creates both winners and losers. While the policy is bad for the economy as a whole, some domestic firms (firms that do not export) will gain from these measures and earn higher profits. If these firms have powerful lobbies, this can lead to surges in protectionism. Especially during times of crisis, politicians seem to be prone to such arguments.

Note that there is also one measure that actually has beneficial outcomes. While export subsidies are considered as protectionism because they favour domestic products over foreign products, they tend to stimulate trade, which is beneficial. Consumers in both countries are happy because their imports become cheaper (imports of the foreign country become cheaper because of the subsidy, imports of the home country become cheaper because of the adjustment in relative prices which makes foreign goods cheaper). As a result, consumption and GDP rise in both countries. Note, however, that the stimulating effect of export subsidies is smaller than the effect of general government spending and very short-lived, turning into a contractionary effect after a short period of time.

Our analysis has not considered retaliation by trading partners. It was our intention to show that even when the trading partner does not retaliate, protectionism does not pay off. Note, however, that with the exception of export subsidies all measures hurt trading partner countries and are therefore likely to lead to retaliation. Taking retaliation into consideration would make matters even worse.6


In this paper we have addressed the question whether protectionism can be a suitable instrument to stimulate an economy during a recession. This question was motivated by the considerable increase in protectionist measures during the recent financial crisis. The proponents of protectionism have popular supporters, among them Nobel Prize winner Paul Krugman.

The argument pro-protectionism is that it protects domestic firms and workers and stimulates domestic production by redirecting demand towards domestic products. Additionally, it internalises policy externalities, since general, non-discriminatory government spending is distributed among many countries, while the cost of the stimulus is borne by only one country. Hence, it is claimed that by using “Buy National” clauses this externality could be internalised. We demonstrate that this argument falls short. It ignores the adjustments in the terms of trade, the relative price of domestic and foreign goods. These terms of trade adjustments counteract the shift in demand, distort trade and hurt exporters. In sum the negative effects prevail.

Mario Larch, University of Bayreuth, Germany.
Wolfgang Lechthaler, Kiel Institute for the World Economy, Germany.

  • 1 Available at http://krugman.blogs.nytimes.com/2009/02/01/protectionism-and-stimulus-wonkish/.
  • 2 F. Ghironi, M. Melitz: International Trade and Macroecoenomic Dynamics with Heterogeneous Firms, in: Quarterly Journal of Economics, Vol. 120, No. 3, 2005, pp. 865-915.
  • 3 M. Melitz: The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity, in: Econometrica, Vol. 71, No. 6, 2003, pp. 1695-1725.
  • 4 M. Larch, W. Lechthaler: Why “Buy American” is a bad idea but politicians still like it, in: Canadian Journal of Economics, forthcoming; M. Larch, W. Lechthaler: Is there a short-run case for protectionism? Trade policy and the business cycle, Kiel working paper 1717, 2011.
  • 5 General government spending also implies consumption crowding out but to a lesser extent, because the increase in domestic prices is smaller.
  • 6 For further details see M. Larch, W. Lechthaler: Why “Buy American” is a bad idea…, op. cit.

Download as PDF

DOI: 10.1007/s10272-011-0384-5