Endogenous Technology Adoption, Government Policy andTarifficationJosh EderingtonUniversity of KentuckyPhillip McCalmanUniversity of California-Santa Cruz∗January 2006AbstractIn this pap e r we integrate government policy into game-theoretic models of endoge-nous technology adoption to investigate the impact of alternate policy instruments onthe adoption of productivity-improving technologies. We show that while ad-valoremtaxes have a neutral impact on technology adoption, specific taxes tend to decrease thespeed of technology diffusion. As an application of this finding we demonstrate how,in an open-economy setting, tariffication (i.e., the conversion of quotas to ad-valoremtariffs) can lead to faster technology adoption world-wide.KEYWORDS: Technology Adoption, Specific tax, Ad valorem tax, tariffs, quotasJEL Classification∗Ederington: Department of Economics, University of Kentucky, 335 Gatton Building, Lexington, KY40506 (email: [email protected]); McCalman: Department of Economics, University of California, 1156High Street, Santa Cruz, CA 95064 (email: [email protected]). We would like to thank Stephen Yeapleand seminar participants at the 2005 AEA meetings. Any remaining errors are our responsibility.11 IntroductionInnovation and technology adoption are widely viewed as the primary sources of produc-tivity growth. In this respect the adoption of new technologies is especially important asa superior technology confers no benefits until that technology is employed by potentialusers. Thus, evaluation of government policy should consider, not only standard issues ofstatic efficiency, but also it’s dynamic impact on the incentives to adopt new technologies.Indeed, Goolsbee (2006) finds that the dynamic costs of taxes, by reducing the incentive forfirms to enter new markets, can even outweigh the conventional efficiency costs of taxation.Similarly, we argue that taxes and other government policies can impact the incentives forfirms to adopt new cost-saving innovation. In this paper, we integrate government policyinto game-theoretic models of endogenous technology adoption to investigate how suchpolicy impacts the diffusion of new technologies.There exists an extensive literature in the field of public economics on the relativeefficiency of different forms of taxation, focusing on the differences between unit (specific)and ad valorem (percentage) taxes. This literature dates back to the seminal paper by Suitsand Musgrave (1953) with more recent c ontributions including Delipalla and Keen (1992),Skeath and Trandel (1994) and Anderson, de Palma, and Kreider (2001). These papersdemonstrate that, while per-unit and ad-valorem taxes are equivalent under conditions ofperfect competition, they can have differing impacts when the market is characterized byimperfect competition.There exists a parallel literature in the field of international trade on the relativeefficiency of different forms of trade barriers. A key question in this literature is the relativeefficiency of tariffs versus quota protection.1As in the public economics literature, whiletariff and quotas are perfect substitutes under perfect competition they c an have differingimpacts under imperfect competition (see Bhagwati (1965) and Bhagwati (1968)). Morerecent contributions to this literature analyze trade policy instruments under differentforms of competition (e.g., see Jorgensen and Schr¨oder (2005)) as well as various marketfrictions (e.g., see Matschke (2003) and Herander (2005)).The common thread in both the international trade literature and the public economicsliterature is that the relative efficiency of different governmental policies is almost uni-formly analyzed in static models. However, it s ee ms that another way different tax/tradepolicy instruments could differ is in their respective dynamic effects on firm productiv-ity. Specifically, a key question addressed in this paper is how different governmentalpolicies can impact firm productivity through affecting that rate at which firms adoptnew technologies. In this sense, this paper is closely related to that of Miyagiwa and1An analogous question concerns the relative efficiency of specific vs. ad valorem tariffs (for a reviewsee Helpman and Krugman (1989).2Ohno (1995) which investigates the effect of different trade barriers on technology adop-tion.2However, Miyagiwa and Ohno (1995) investigates the adoption decisions of a singleimport-competing domestic firm engaged in Cournot competition with foreign exporters.In their model, the non-equivalence be tween tariffs and quotas rested on the lack of astrategic effect to technology adoption in the presence of a quota. Intuitively, under atariff regime one of the benefits of adopting a cost-saving technology is that it reduces theexports of the foreign firm and thus increase s home firm profits. This strategic benefit toadoption is absent under a quota regime and thus Miyagiwa and Ohno (1995) concludesthat home firms will adopt new technology earlier under tariff protection than under anequivalent quota.In contrast this paper models the technology adoption decisions of a large numbe r ofsmall firms engaging in monopolistic competition and thus the strategic effects of Miyagiwaand Ohno (1995) do not arise.3Rather, our paper focuses on governmental policies thathave a specific (per-unit) impact on marginal costs versus policies that have an ad val-orem (percentage) impact. Specifically, we show that in a dynamic model of technologydiffusion a specific tax acts as an impediment to the adoption of cost-saving technologyimprovements since it has a disproportionately negative impact on high-productivity firmsas their relative price advantage is reduced.The intuition for this result parallels that of the classic Alchian-Allen conjecture thatspecific transportation costs will lead firms to export high-quality goods abroad sinceper-unit transportation costs lower the relative price of high quality goods. This logic iscentral to the literature on how quotas can lead to an increase in the average quality ofimports by impacting the endogneous quality choices of both consumers and firms (e.g, seeFalvey (1979), Krishna (1987) and Krishna (1990)). As we show in this paper, it also hasimportant implications for the decision of whether to adopt a cost-saving technologicalimprovement.4In our model, per-unit taxes raise the relative price of high-tech, lowmarginal cost firms in foreign markets, thus impeding the desire of