Empirical Evidence on Market Structures

The influence of demand factors on dynamic competitive pricing strategy: An empirical study by A. Roy and J. Raju (2010, Marketing Letters).

This empirical paper investigates the relationship between the dynamic strategic interactions among competitors in a component market and demand factors in the market for the end product. The structure of competition in the US microprocessor (MPU) industry is analyzed using data on prices and sales in both the MPU market as well as the market for personal computers. The pattern of dynamic strategic interaction between competing firms in this market on a key decision variable, price is studied. Non-nested model comparison tests based on equilibrium solutions derived for specific differential games are applied to identify the mode of competitive strategy between pairs of competing brands. The empirical fit to the longitudinal and cross-sectional data, of alternative models of competition, independent (Bertrand–Nash), Stackelberg leader–follower, and Collusion, is used to determine which dynamic model best describes actual competitive behavior over the life of each MPU. Demand for the product market which is downstream from microprocessors, that for personal computers, is estimated using a generalized diffusion model with price effects. Data from the markets for desktop and laptop computers are analyzed at the level of computer vendor and internal microprocessor. Patterns are uncovered, linking downstream demand parameters with upstream competitive strategy. There is evidence to suggest that when there are strong diffusion effects driving sales of both the competing computer brands, there is a higher likelihood of Bertrand–Nash competition among MPU firms. However, when there are higher cross-price effects (substitutability) among personal computer brands there is a greater chance of Stackelberg leader–follower price competition. When self-price effects are relatively high, the likelihood of Bertrand–Nash competition among MPU firms increases. Furthermore, when the potential demand for the computer product category is high, there is a higher likelihood of Bertrand–Nash pricing in the MPU market.

 

Endogenous Market Structures and Innovation by Leaders: an Empirical Test by D. Czarnitzki, F. Etro and K. Kraft (2010).

There is a lot of debate on the role of market leaders in investing in R&D and promoting technological progress. A commonly held view is that firms invest more in a more competitive market where the entry pressure is stronger, and incumbents tend to be less innovative than their followers, so that the persistence of their dominance is typically the signal of market power and of the lack of entry pressure. This view is often associated with Arrow (1962), who has shown that incumbents have lower incentives to invest in R&D than the outsiders, and that in case of free entry in the competition for the market they do not invest at all, leaving the innovative activity to the outsiders. In this paper we adopt a Schumpeterian perspective and we challenge this view both from a theoretical and empirical perspective, showing under which conditions incumbent leaders do invest more than the other firms and providing empirical evidence in support of our thesis.

There are few competing explanations for innovation by incumbents in Schumpeterian growth models. The simplest one, due to Segerstrom (2007) relies on the fact that incumbents may have a technological advantage in the R&D activity. This assumption may be realistic in certain sectors and allows one to study monopoly persistence, but it is basically equivalent to assume the solution of the Arrow paradox rather than solving it. Moreover, taking this view literally, we should conclude that whenever we observe monopoly persistence it is because the incumbent firm is more efficient than the other firms both at producing and innovating. There are many sectors in which incumbents do not appear to have any cost advantage in the development of innovations compared to the outsiders, and still both the incumbents and the entrants keep investing.
Acemoglu (2008; 2009, Ch. 14) has proposed a different rationale for innovation by leaders. This may be due to the fact that only the incumbents can invest in incremental innovations (because outsiders would infringe their patents through small improvements), while entrants can invest in more radi cal innovations (because of the Arrow effect). In such a way, both the incumbents and the outsiders invest, and the growth rate depends on their rates of investment weighted by the respective productivity increases. This is a plausible mechanism, but it explains why incumbents may invest in small improvements of their own technologies, which is a trivial activity, and not why they may directly compete against outsiders to obtain radical innovations,
which is the key issue.

Here, we propose an alternative explanation for innovation by incumbents which does not rely on technological advantages or exogenous market structures, but is based on a pure strategic advantage of the incumbents in patent races with endogenous entry of outsiders. We develop a simple contest for a drastic innovation with strategic interactions in the tradition of the recent works on endogenous market structures and market leadership, and show the crucial role of entry pressure on the different behavior of leaders and followers. Entry reduces profitability and therefore it reduces also the investment of each firm (although the aggregate investment increases). Therefore the endogenous entry threat tends to reduce R&D intensity of each firm. Moreover, in such a context, an incumbent would not invest at all because of the Arrow effect. However, things change under the assumption that the incumbent is also the leader in the contest for the innovation, as reasonable given its strategic advantage in the market. The incumbent that is also leader exploits its first mover advantage to invest more than the other firms. The intuition has to do with the impact of its investment on entry: a small investment attracts large entry and makes it likely that another firm will replace the incumbent, while a commitment to a large investment has the double
advantage of reducing entry and increasing the chances of an innovation.

We bring to the data the two basic predictions of our theoretical investigations: R&D intensity of the average firm is lower when there is an endogenous entry threat, and the R&D intensity of the incumbent leader is larger than the one of the average firm when there is an endogenous entry threat. We test these hypothesis through a Tobit model for R&D intensity. Our empirical investigation is based on a unique dataset on the German manufacturing sector, the Mannheim Innovation Survey from 2005 conducted by the Centre for European Economic Research (ZEW), that includes a wide number of firm level data with a special focus on innovation. A novel aspect of our empirical approach is given by the fact that the same firms provide a subjective view on our key determinants of R&D intensity, the entry pressure and the leadership. Rather than determining arbitrarily the size and composition of a market, assigning a degree of entry intensity in a discretionary way, and assigning a status of leadership on the basis of possibly arbitrary assumptions, using the survey results we allow the firms to identify the size of their main market, the existence of an endogenous threat of entry in the market and the identity of the leader in the market. We also perform robustness tests concerning the potential reverse causality between R&D and entry threats using IV regressions and a number of exogeneity tests. Our main predictions are strongly supported by the empirical evidence: entry pressure reduces the average investment per firm, but incumbent leaders invest more than other firms when there is the pressure of a strong threat of entry.

The Impact of Competition on Bank Orientation by H. Degryse and S. Ongena (2009, Journal of Financial Intermediation).

How do banks react to increased competition? Recent banking theory significantly disagrees regarding the impact of competition on bank orientation—i.e., the choice of relationship-based versus transactional banking. We empirically investigate the impact of interbank competition on bank branch orientation. We employ a unique data set containing detailed information on bank–firm relationships. We find that bank branches facing stiff local competition engage considerably more in relationship-based lending. Our results illustrate that competition and relationships are not necessarily inimical.

The Online Advertising Industry: Economics, Evolution, and Privacy by D. Evans (2009)

Online advertising accounts for almost 9 percent of all advertising in the United States. This share is expected to increase as more media is consumed over the internet and as more advertisers shift spending to online technologies. The expansion of internet-based advertising is transforming the advertising business by providing more efficient methods of matching advertisers and consumers and is transforming the media business by providing a source of revenue for online media firms that compete with traditional media firms. The precipitous decline of the newspaper industry is one manifestation of the symbiotic relationship between online content and online advertising. Online-advertising is provided by a series of interlocking multi-sided platforms (also known as two-sided markets) that facilitate the matching of advertisers and consumers. These intermediaries increasingly make use of detailed individual data, predictive methods, and matching algorithms to create more efficient matches between consumers and advertisers. Some of their methods raise public policy issues that require balancing providing consumers more valuable advertising against the possible loss of valuable privacy.

Price Effects of Market Leadership Volatility in Japan by N. Doi (2008).

An attempt has been made to test the effects of market leadership volatility on industry price change in Japanese manufacturing industries. The results suggest that market leadership volatility has a definite influence on domestic price change, providing empirical support for its effect of a downward movement of prices. But, it is necessary to note that its effect is not strong enough to offset the concentration’s effect, since concentration still is positively related to price change. Also, import intensity has a negative association with price change. Distribution involvement has a positive effect. Thus, market leadership volatility is important as a determinant of market behavior from the viewpoint of competition policy as well as business strategy. Public policy makers should be concerned about market leadership volatility.

The Economics of the Online Advertising Industry by D. Evans (2008).

Online advertising has grown rapidly in the last decade. It now accounts for almost a seventh of all advertising spending and contributes to the preponderance of revenues for most websites. It is projected to increase sharply as more consumers spend time online on their personal computers and as additional devices such as mobile phones and televisions are connected to the web. This article describes the market structure of the online advertising industry and several complex economic aspects of it. Using the lens of the new economics of multi-sided platforms it examines search-based advertising platforms, as well as platforms that facilitate the buying and selling of advertising space on websites. The unique features of online advertising include the use of Internet-based technologies and data collection mechanisms to target and track specific individuals, and to automate the buying and selling of advertising inventory. Like modern finance, online advertising relies heavily on advanced economic and statistical methods.

Invention and Discovery In Science-Based Firms by J. D. Adams and J. R. Clemmons (2008).

To what extent does new invention depend on knowledge from past inventions? How important for invention is knowledge outside the firm compared to inside knowledge? Do knowledge flows from science play an additional role in invention? How do production functions for invention compare with those for scientific discoveries? This paper seeks to provide tentative answers to these questions. Its approach builds on two literatures. The first is the search approach to technological change pioneered by Evenson and Kislev (1975, 1976), and applied to R&D by Nelson (1982) and growth by Kortum (1997). The second is the quality ladders approach to growth, in which higher quality products supersede lower quality ones. Grossman and Helpman (1991) and Barro and Sala-i-Martin (2004) contain expositions. Etro (2004) implicitly builds on this approach in his discussion of patent racing under different assumptions about industry organization.

Using a sample of large R&D-performing firms in the U.S., which are observed during 1981-1999, the empirical work quantifies inventions and scientific discoveries as well as knowledge flows to both. The evidence builds on patents and citations in the NBER Patent Citations Database and papers and citations to papers from the NBER Scientific Papers Database. In addition the authors draw on Compustat for firm R&D stocks and on the CASPAR database of the National Science Foundation for university R&D1.

Main findings are as follows. Simple patent counts depend on knowledge flows from inventions but much less on science. When we replace patent counts with patents weighted by citations received, a measure of invention weighted by impact, the authors find that knowledge flows from patents continue to dominate, but that knowledge flows from science now have a much greater impact. When they compare flows from inside and outside the firm, they find that outside flows have a larger effect on patents produced. When they compare industry-based with university-based they find that the effects of knowledge flows from industry outweigh those from universities by a wide margin.

For comparison with patents the paper estimates production functions for scientific papers. It finds that firm’s papers depend even more that its patents on knowledge flows from outside the firm. In addition, unlike patents, flows from universities have a larger effect on scientific papers than industry flows. These comparative findings on patents and papers suggest that innovation is a sequential process in which the scope of the firm’s research progressively narrows as it proceeds from early to late stages. In the early going the firm’s scientific effort depends mostly on outside knowledge, most of that coming from universities. Later on, the firm’s inventions depend mostly on the firms’ past inventions and on inventions elsewhere in industry, with science strictly secondary. Nevertheless, the authors observe that science plays an important role in patents, especially highly cited patents.

The Economics of the Online Advertising Industry by D. Evans (2008, Review of Network Economics).

Online advertising has grown rapidly and accounts for about 7% of US advertising spending. It is projected to increase sharply as more consumers spend time online on their personal computers and as additional devices such as mobile phones and televisions are connected to the web. This article describes how the online advertising industry works. The industry is populated by a number of multisided platforms that singly or in combination facilitate connecting advertisers to viewers. Search-based advertising platforms, the most well developed of these, has several interesting economic features that result from the combination of keyword bidding by advertisers and single-homing.

The Persistence of Market Leadership: Evidence from Japan by M. Kato and Y. Honjo (2007).

This paper explores the persistence of market leadership in Japanese manufacturing industries over the period 1975–2004. By applying survival data techniques, we examine how long market leadership persists, and describe the differences in the duration of market leadership between industries. According to our results, about half of market leaders are found to maintain their leadership positions for ten years. Market leadership tends to persist in concentrated and cartelized industries, whereas it is less likely to persist in growing, R&D-intensive, and importintensive industries.

The Long-Run Impact of ICT by F. Venturini (2007).

This paper examines the growth impact of digital capital for the US and the EU-15 countries from a long-run perspective. It estimates the output elasticity to ICT within an aggregate production function framework by means of panel cointegration analysis. We provide robust evidence that Europe is lagging behind the US in terms of ICT utilisation, especially due to the dismal performance of the largest countries. On the other side, though, there is a core of some small dynamic economies in Europe whose growth pattern has been highly positively influenced by ICT over the last quarter of century.

Testing for Persistence of Profits’ Differences Across Firms by J. Cable and D. C. Mueller (2006).

A fundamental premise underlying normative arguments for market systems is that competition drives profits to zero, and thus that the price system maximizes social welfare by equating the marginal social benefits and costs of each economic activity. There are essentially two underlying logics for this – one based on a static view of competition, the other on a dynamic view. Under the static view, competition among an existing set of firms suffices to produce zero profits at each point in time. Under the dynamic view, the entry and exit of firms drives profits to zero in the long run, and is thus consistent with there being non-zero economic profits at different points in time.

Testing whether the static view of competition is consistent with observed patterns of profits is fairly straightforward. Not surprisingly such tests were the first to be performed and the preponderance of studies to date still fall into this category. Beginning with Mueller (1977) a by now rather large literature has also developed, which tries to test the dynamic view of competition. Since the dynamic view allows for the possibility of differences in profits across firms at any given point in time, the literature testing it has concentrated on determining whether these persist. The key methodological issue in this literature is, therefore, how to define and test for the persistence of profit differences.

This methodological issue is the main concern of this paper. Cable and Mueller begin by reviewing the premises and inferences of the static view of competition. Much of the dynamic literature has built assumptions from the static literature into its modeling of the dynamic process. Thus to understand the logic underlying the dynamic models it is useful to review the logic underlying static models of the competitive process. Accordingly Section I reviews the logic and implications underlying both types of model.

In Section II the authors discuss the methodological issues that arise when trying to test the various models outlined in Section I. The thrust of this section is to point out that models tested in much of the existing literature may be open to more than one interpretation, and that the results reported may not always support the inferences drawn by the authors.

Section III extends the discussion by considering complications that can arise due to innovation, mergers and cyclical factors. The points raised in Sections II and III are then illustrated in Section IV with estimates of various dynamic models of firm profitability using data from both the United States and the United Kingdom. This work differs from that of previous studies in that Cable and Mueller test a larger set of possible models than have been used in previous work, and we conduct these tests using longer time series on company profits, thereby increasing their power.

The main conclusions drawn this research are that the patterns of firm profits observed over time in both countries are consistent with a much larger and more complicated set of models of the competitive process than has been assumed in the literature until now.

The Odyssey of the Mobile Internet. The emergence of a networking attribute in a multidisciplinary study by V. Saarikoski (2006)

This research focuses on the mobile internet. Japan was successful with its mobile internet launch already in 1999. The West has struggled and has been unable to follow Japan’s success. This research asks why and builds a new understanding to the mobile internet. This is a multidisciplinary research project. The research applies Hargadon’s theory of how to create breakthroughs and shares ideas within disparate social networks. By combin ing Christensen’s theory of disruptive ideas (the innovators dilemma) and Hargadon’s theory a total of 18 potential disruptive opportunities in the mobile internet field are found. The synthesis of the disruptive ideas leads to look at the new science of networks. The research concludes with several novel findings and arguments.

The Japanese chose email as the key messaging media on their mobile phones. However when trying to enter the European market they did not adopt email as the key messaging media. Correlation exists between adopting email and success (as well as with not adopting email and failure). This work suggests that there is also a deeper reason to the correlation, embedded in the structure of email. Email creates scale free networks in a far more efficient manner than SMS.
In the west the extensive use of the SMS is therefore possibly a barrier and an inhibitor, preventing the west from being successful with the mobile internet, because the SMS performs poorly in creating scale free networks.

The ability to create scale free networks is not embedded solely in the technical features of the product, but very much a result of the techno-economic combination (e.g. packaged prices and flat rate pricing encourage the emergence of the scale free structure). This leads to the following hypothesis: success with the mobile internet can be achieved by adding elements, which are efficient in a scale free sense. This work generalises further and also suggests a need for a paradigm change in our understanding of telecommunications. Understanding the principles of creating a scale free structure imply that we should abandon our old business models built on the belief of a transaction based pricing model.

The Market Structure of Internet Service Provision by A. Chaudhuri and K. Flamm (2005).

Telecommunications services in the US are not federally micro-managed. Internet service provision has enjoyed a high degree of freedom at the federal, state, and local levels until now, and this is likely to persist in the near future. Confusion about the existence, extent, and direction of a digital divide is increasing rather than abating with time, and calls for a universal service mechanism paralleling that for basic telephone service are being heatedly debated in policy circles. This paper tries to shed light on three key questions in the debate: (i) the existence and magnitude of intra- and inter-modal price competition in the ISP market at present; (ii) the possibility of market concentration in the future; and (iii) the possible effects of price subsidies in increasing penetration. Using data from multiple sources, the authors find that (i) already high levels of inter- and intra-modal competition effectively impose price-discipline, (iii) the existence of economies of scale may lead to consolidation, and (iii) price-subsidies would probably promote Internet-penetration at the household level, but would most likely be both redundant and extravagant. They argue for a continuation of the current “hands-off” approach of the Federal Communications Commissions, with a recommendation for promoting parity in broadband regulation.

Research and Development in Transition. The Effects of Market Liberalisation by K. Johansen and N. Damm (2005).

Previous research has shown that Former State Owned Enterprises (SOEs) decrease their relative R&D spending following privatisation and deregulation. In this thesis the authors analyse and try to explain what happens to R&D and innovation during these times. First, they examine a large number of privatised SOEs (usually operating on recently deregulated markets), and find that R&D intensity has decreased. Secondly, they pick four former SOEs operating in the fields of telecom and energy, on which they perform a case study and analysis, to identify what happens to the R&D activities in greater detail. The authors link this to related theories and other research on the area. The following similarities are found among the studied companies:
· the former incumbents becomes more service oriented and cost conscious;
· in-house R&D is decreased relative to turnover;
· basic Research is outsourced on suppliers and independent actors such as universities and small technology firms (development is conducted within close proximity of the company and more closely aligned with marketing activities);
· positive externalities in terms of efficiency of investments into R&D and enhanced product offerings.
The Royal Swedish Academy of Engineering Sciences commissioned this thesis as part of the project “Collaboration for growth” which was initiated in order to identify how R&D and innovation capacities change following market liberalisation of telecom, electricity, transportation and the defence sectors in Sweden, and how these insights could help propel Swedish technology and thus promote future national growth. The purpose of this thesis is to add an international dimension to the project.

The Impact of Competition on Bank Orientation and Specialization by H. Degryse and S. Ongena (2005).

How do banks react to increased competition? Recent banking theory offers conflicting predictions about the impact of competition on bank orientation, that is the choice of relationship based versus transactional banking, and bank industry specialization. Arguments about the fundamental incompatibility between relationships and competition have acquired prominence in economics (old and recent Schumpeterian theories have argued that market monopolies offer better incentives for innovation, as sunk R&D expenditures are recoupable only through the generation of future rents flowing to a monopolist-innovator). This important paper empirically investigates the impact of interbank competition on bank branch orientation and specialization. The authors employ a unique data set containing detailed information on bank-firm relationships and industry classification and find that bank branches facing stiff local competition engage considerably more in relationship-based lending and specialize somewhat less in a particular industry. These results illustrate that competition and relationships are not necessarily inimical.