Why trade policy matters for firms’ R&D investment

Andreas Moxnes, Karen-Helene Ulltveit-Moe, Esther Ann Bøler, 18 July 2012



As the crisis continues to ricochet around the world and unemployment stays elevated, the threat of protectionism rises. As the latest WTO report on G20 trade measures (WTO 2012) and the GTA database confirm (Evenett 2012), the threat is real. Trade barriers are rising. This is, therefore, an excellent time to refresh the evidence that openness to trade and investment comes with substantial benefits. Perhaps most important is the role of trade in fostering productivity growth.

The effects of trade reform on productivity have been widely studied (Harrison et al. 2011). Recent literature has focused on the role of input tariffs in generating productivity gains. For example, Amiti and Konings (2007) find that a ten percentage point fall in input tariffs leads to a 12% productivity gain for importing firms. Although it is usually not possible to empirically discern the exact channel that gives rise to these productivity gains, it is common in the literature to attribute it to either learning (Keller 2002), variety, or cost/quality effects.

In a recent CEPR discussion paper (Bøler et al. 2012), we address a related question that has received little attention. Do cheaper imports affect firms’ incentives to innovate? Moreover, how do R&D investment and imports interact in determining firms’ productivity growth? We will start by looking at a set of stylised facts about firm-level R&D spending and international sourcing.

Stylised facts on innovation and imports

Our analysis is based on novel firm-level data for Norway which allows us to merge data on firms’ performance, international trade , investment in R&D, and R&D personnel. Our data is a biannual panel of Norwegian manufacturing firms with more than 50 employees from 1997 to 2005. After dropping firms with either zero employment, missing capital stocks or missing value added, we get an unbalanced panel of roughly 850 firms per year.

Five basic facts about innovating firms and their sourcing behaviour are worth noting:

  • Fact 1: Only a subset of firms innovates. Among innovating firms, almost all firms import. More than 40% of firms do not invest in R&D. Among those who do, as much as 98% source inputs from abroad. As for those who do not invest in R&D, 13% are non-importers.
  • Fact 2: Innovating firms are larger, source more foreign products, and have higher import share and labour productivity. Innovators have more than 50% as many employees, import twice as many products, have a 60% higher import share of intermediates, and a 13% labour productivity advantage compared to non-innovators.
  • Fact 3: Importers are also larger and have higher labour productivity. Importers are roughly three times larger and 30% more productive (labour productivity) than non-importers. This remains true when comparing firms within industries.
  • Fact 4: There is a positive correlation between R&D investment and import participation, import share, number of imported products and labour productivity, which also holds within a given industry and after controlling for firm size.
  • Fact 5: Firms that start to innovate grow faster, increase their import share and increase the number of imported varieties compared to all other firms. Within each industry, R&D starters grow faster than the other three groups. Furthermore, R&D starters clearly shift their sourcing strategy - in an absolute sense and relative to non-innovators - as they start to import a larger number of products and increase the value of imports as well as the share of imports relative to total costs.

Complementarity between R&D investment and imports

Guided by the stylised facts on innovation and imports, we build a quantitative model with heterogeneous firms to analyse the relationship between investment in knowledge and imports of intermediate goods. Both innovating and importing intermediates are subject to fixed costs. Complementarity arises since R&D on average increases future profits and revenue, thereby making it more profitable to cut costs by sourcing inputs internationally.

We confront the model with data on Norwegian firms' innovation activities and their sourcing of imported inputs. Our structural estimates show that investment in both knowledge and foreign sourcing drive down marginal cost. A firm that performs R&D in every period has on average 30% higher revenue compared to a firm that never invests in R&D. A firm in the upper quartile in terms of the number of products sourced from abroad has roughly twice the revenue compared to a firm in the lowest quartile. This translates into substantial measured productivity differences across firms. 

Our results have important implications.

First, since both innovation and foreign sourcing reduce marginal costs, and the two activities are complementary, we need to control for both factors when estimating the impact of R&D or imports on measured productivity.

Second, input trade liberalisation promotes technological change since falling import prices raise the incentive to innovate. Both mechanisms generate higher measured productivity. Hence, our work offers a new mechanism through which imports increase productivity, which may contribute to explaining why a number of studies find firm-level productivity gains to be associated with input trade liberalisation.

Third, an R&D policy reform that reduces firms’ costs of R&D raises the returns to both R&D and imports of intermediate inputs, thereby promoting not only technology upgrading but also imports and enhanced productivity improvement.

A policy experiment

In the early 2000s, a tax credit for R&D projects was introduced in Norway, lowering the marginal cost of R&D by 20%, but only for a subset of firms. This policy reform lends itself as a natural experiment. Our evidence suggests that lower marginal R&D costs had a large impact on investment in both knowledge and imports of foreign intermediates, consistent with our model.

To what extent can our proposed theoretical mechanism match observed imports and productivity growth in the data? Comparing our evidence with a simulation exercise based on our structural estimates provides some guidance. We find that roughly half of the import surge that occurred in the aftermath of the policy change can be attributed to the proposed theoretical mechanism. Furthermore, one fifth of average measured productivity growth among new innovators came from sourcing more products, while the remaining four fifths came from technical change, illustrating how trade can amplify productivity gains.

In sum, both R&D and trade policy may have additional benefits in terms of productivity growth, due to complementarity between R&D investment and international sourcing.


Amiti, Mary and Konings (2007): “Trade Liberalization, Intermediate Inputs, and Productivity: Evidence from Indonesia”, American Economic Review, 97(5), 1611-1638.

Bøler, E. A., Andreas Moxnes and Karen Helene Ulltveit-Moe (2012): “Technological Change, Trade in Intermediates and the Joint Impact on Productivity”, CEPR Discussion Papers 8884.

Evenett, Simon (2012), Débâcle: The 11th GTA report on protectionism, CEPR.

Harrison, Ann, Leslie Martin, Shanthi Nataraj (2011), “Learning vs stealing: How important are market-share reallocations to India's productivity growth?” VoxEU.org, 22 March.

Keller, Wolfgang (2002): “Geographic Localization of International Technology Diffusion”, American Economic Review, 92(1), 120-142

WTO (2012) “Report on G20 trade measures”, 31 May, Geneva.

Topics: International trade, Productivity and Innovation
Tags: imports, productivity, R&D

PhD candidate, University of Oslo

Andreas Moxnes

Assistant Professor of Economics, Dartmouth College

Professor of Economics at the Department of Economics, University of Oslo and CEPR Research Fellow

Most Read