Russell Sage Laboratory
110 8th Street
Troy, NY 12180-3590
Phone +1 518 276 3296
Associate Dean for Research and Graduate Studies
1. Shakeouts in industrial competition
2. Cross-country industry comparisons
3. Entry characteristics and survival
4. Technology and competition
5. Theories of shakeouts
6. The internet and industry competition
7. Inventor motivations
8. Creation of new industries
Paid or for-credit positions are often available for RPI PhD students and undergraduates (who have not yet graduated), to help with research. For summer 2017, one undergraduate position is currently available as noted below. For positions in fall or spring, undergraduates please apply at or just before the start of term. Undergraduates may also apply for summer funding at the very beginning of the spring term, or in the late fall, when it is possible to apply for special URP funding. The undergraduate positions do not require specific skills, and students from all majors are encouraged to apply. Some work may benefit from skills with programming in general or LiveCode specifically, so students with these skills are particularly encouraged to apply.
The following forms can be used to apply for all positions. If only certain projects interest you, please say so on the first page. Use the appropriate one of the following forms: (1) undergraduate application form for jobs during summer, (2) undergraduate application form for jobs during fall/spring, or (3) PhD student application form for jobs during summer.
Research opportunity for Summer 2017:
|Shakeouts in Competitive Markets
Does this near-extinction look familiar? Most industries that have been around for at least several decades have experienced some amount of shakeout, in which the number of companies drops off. Shakeouts are no historical artifact; they happen all the time. What's more, they happen despite rapid market growth.
A series of papers joint with co-author Steven Klepper probes the causes of industry shakeouts. The papers explore some of the most fundamental issues of economics: competition in its raw form as it plays out over time. They also help understand one of the most important issues in economics: technological change, which yields the bulk of economic growth.
Reference: Steven Klepper and Kenneth L. Simons, "The Making of an Oligopoly: Firm Survival and Technological Change in the Evolution of the U.S. Tire Industry," Journal of Political Economy, August 2000, vol. 108 no. 4, pp. 728-760.
Experience shakeouts yourself in my industry simulator (try game versions 2 and 3), which gives a first-hand experience in how real shakeouts seem usually to happen. Beta-test version, supports Internet Explorer and Safari (sorry not yet Firefox).
|Industry Dynamics: A Big Deja Vu
Look familiar again? Industry competitive dynamics usually are. The same processes seem to be at work within and across different countries. This recent study compared matched industries between the US and the UK to see if product-specific traits, rather than random events and the national environment, seem to explain competitive dynamics. After years of tedious data collection, I expected and hoped to see some similarity across the two countries, but I was shocked how well the numbers matched up.
Industries with shakeouts had shakeouts in both nations, and the severity of the shakeout -- as measured by the % drop in number of companies -- was quite similar. The timing of the peak in number of firms was often almost identical across the two countries. And in each country the same industries were lacking in shakeouts (not all industries have shakeouts, and there are good reasons why not).
This tells something fundamental about how industries work. There are driving mechanisms, that work the same in different environments, stem from underlying product-related characteristics such as the nature of the product's technology, and determine the overarching dynamics of how competition plays out in the market. Of course there are international differences, particularly in terms of which country's producers win or lose when a market goes global. But the differences are much less than we might have guessed.
|Dominance by Birthright
When you're a company, it helps to be born with muscle. That's one of the findings of a study of US radio producers when they began manufacturing televisions. The firms were almost the only ones with the engineering know-how and other traits to succeed in this new market. They set up new production lines relatively efficiently and began improving their products. Improvement is a big deal in an industry with tight cost margins and with the repairman having to come every two months on average in the early days of television. Some of the radio producers became the big firms, and out-innovated the others. In fact, I counted all the manufacturing innovations I could catalog from industrial journals, and found that the big firms - born of the radio days - dominated innovation. The market went through a big shakeout that wiped out most firms.
Too bad the US firms died their death too - they couldn't keep up in the face of international competition. Japanese and other foreign firms entered US production with just as much brawn already in hand, and they had better products. Transistor radios may have seemed cheap at first, but the transistor was key to lower-cost manufacturing and fewer breakdowns. US companies ever after lagged 1-2 generations behind in their integrated circuits. The last of the US heavyweights, Zenith, was acquired by Goldstar of Korea in 1995.
Reference: Steven Klepper and Kenneth L. Simons, "Dominance by Birthright: Entry of Prior Radio Producers and Competitive Ramifications in the U.S. Television Receiver Industry," Strategic Management Journal, October-November 2000, vol. 21 no. 10-11, pp. 997-1016.
|Technology and Shakeouts
Why do industry shakeouts happen? There are different ways to tell, but one way is to dig deeply into technological evidence. This study does just that. It's hefty, 3-4 times longer than the usual journal article, even after we trimmed it down. It has to be so long because there's a lot to know about technological change in products with shakeouts.
This study uses a sample of four products, automobiles, tires, televisions, and penicillin, in order to look back over the long span of time at products we know had very severe shakeouts. For each product, innovation in the product and manufacturing process are reviewed with tables of the major innovations and a systematic review of the means to quality and cost improvement. For both product and process improvement, for all four products, the large producers dominated the lists of innovations.
Many other products aren't like this, but in these products something -- seemingly, the need for innovation inside firms -- caused some early entrants' market share dominance and the industries' shakeouts. What was going on here is systematic domination of innovation by the leading firms. They're the only ones for whom it made sense to hire huge cadres of engineers and scientists and other smart people who made their companies' success. The innovative process isn't glamorous; in fact it can be greasy and sweaty, but it's what beats the competition and earns profits.
Reference: Steven Klepper and Kenneth L. Simons, "Technological Extinctions of Industrial Firms: An Inquiry into their Nature and Causes," Industrial and Corporate Change, March 1997, vol. 6 no. 2, pp. 379-460.
|Shakeout Theories: Firm Survival
Shakeouts have become a hot topic for research recently, and an increasing number of theories seeks to explain them. This paper digs into the evidence to find out why they happen. A companion piece to the above paper on technology and shakeouts, this paper emphasizes firm survival patterns. Common theories of shakeouts involving technological change turn out to have sharply contrasting implications for which firms should survive and when they should exit.
The notion that a single innovation from outside an industry triggers the extinctions turns out to be an extinct idea. Suppose firms had to adapt their product and manufacturing methods to the new technology, but once they adapted fine then they could be reasonable competitors. Then there should be a temporary mass exodus, with a rise in the exit rate until the surviving firms are all ones that have adapted. But that doesn't happen. In general, the rate of exit doesn't rise in a shakeout. Instead, new companies stop entering the market and steady competition continually drives out the less-profitable firms.
A popular idea of "dominant designs" in the product has also been applied to shakeouts. But while dominant designs are an important concept, they don't seem cause shakeouts (arguments of dominant design proponents notwithstanding). If firms had to get used to a dominant design to survive, eventually the survivors would be used to the design and the exit rate would decline. This doesn't happen. (And there's more evidence in the earlier technology paper.)
What does matter is constant innovation. Using systematic compilations of information about firm innovation, this paper shows that survival is directly and massively related to constant innovation.
|% of Incumbent versus Entrant IT Consultancies
in the UK with Selected Applications
|The Internet: A Disruptive Competitive
There's a lot of talk about the internet changing the nature of competition. From all the talk, you might expect such changes have had an effect already. Yet few studies have checked for evidence of change. I looked at the evidence, focusing on an industry in which disruptive competitive change is likely to have shown up relatively early. The industry is IT consultancies in the United Kingdom, one of the leading countries at internet adoption.
Theories of disruptive technologies describe a real phenomenon, where upstart firms displace previous industry leaders. Such theories imply that a disruptive technology increases opportunities for market entry, enhances the chances of survival and growth for entrants relative to incumbents, and causes a wave of exit of firms unsuccessful at the new technology. I looked at actual entry and exit patterns to see if the predictions for disruptive technology are occurring yet. I also checked which firms set up internet-related areas of business and whether they achieved enhanced survival and growth thereafter.
The findings through the year 2001 do not match the telltale signs expected if the internet has been a radical disruptive technology. Perhaps, as Michael Porter argued in a recent Harvard Business Review article, the internet will not change the dynamics and strategies of business competition, but merely intensify them and make current strategic concerns all the more relevant.
|Inventors Seeking Gains
Economic research has uncovered many oddities about entrepreneurs. Entrepreneurs would appear to be unrealistically optimistic, wishful thinkers, risk seekers, or just plain driven by their own interests. Moreover, human subject experiments suggest ordinary people - and by implication entrepreneurs - overestimate their own chances of success, and so enter markets despite predicting that entrants will typically lose money. Indeed real entrepreneurs do lose money, relative to what similar people make in corporate jobs. With all these findings, it would seem we should be pessimistic about the way entrepreneurs operate. Yet the empirical evidence on entrepreneurs has largely been based on aggregate patterns, and there has been little attempt to find out how much entrepreneurs respond to profit-related incentives.
The independent inventor is the quintessential entrepreneur, and perhaps the most important entrepreneur, since a substantial fraction of our society's new products have come from independent inventors. This study used data on advance estimates of profit-related characteristics of inventions, and matched the data to inventors' actual decisions whether to commercialize the inventions (and later whether to exit), as well as the inventors' stated reasons for their decisions. By using a very general yet simple theoretical model, we predict how inventors should behave if they respond rationally with profit-seeking motives. It turns out that the financial incentives really do matter in inventors' decisions, and in fact the evidence suggests independent inventors are averse to risk, just like most people.
Reference: Kenneth L. Simons and Thomas Astebro, "Entrepreneurs Seeking Gains: Profit Motives and Risk Aversion in Inventors' Commercialization Decisions," Journal of Economics and Management Strategy, vol. 19 no. 4, Winter 2010.
|How the Light Bulb Was Remade
Economic growth would have been awfully little without technological progress. That is, without the vehicles, electrical and electronic goods, improvements in medicine, agricultural techniques, and tools and machines created since 1500, we would still be living in a world like... the 1500s. Yet, what understanding have we of how all those new products are created? The answer is, not much. A lot of research has focused on how industry competition works after an industry is created, and some has focused on the drivers of new inventions, but the complicated and messy middle ground between invention and industry creation has received little attention.
That's why this study of the making of an industry is so important. It maps out what we know about how industries get made, and reflects on and adds to the map of existing knowledge by analyzing an important and surprisingly fascinating case. That case is solid state lighting. Commonly known as "LED light bulbs," solid-state lights use semiconductor techniques to emit a glow that could be white or tailored to a specific color. As road vehicles at their dawn were odd experiments designed like horse carriages driving on dirt roads, so the LED lights of today are merely a bit more efficient than compact fluorescent bulbs; lack features that one day may be taken for granted, such as universal electronic control, sensing and auto-adjustment and auto-on/off, and data communications embedded in the light flow; and are constrained to the outdated shape of the Edison screw base and old household wiring. The LED bulbs you see in stores today may seem a simple invention and a simple product revolution, but they weren't.
This analysis documents how tens of thousands of researchers contributed to the development of solid-state lighting over a half-century. The technology took surprising twists and turns, gradually improving but with major jumps when key inventors found new means to solve old problems. Patent battles and rivalry to find new profit opportunities contributed to the twists and turns. Niche markets provided stepping stones along which the progress of technology made its lurching way, with businesses joining and falling off from the progress at different steps. All these findings are pieces of a prototypical story of how most industries seem to emerge.
In this case, a few of the researchers involved have won the 2014 Nobel Prize in Physics for their efforts.
Reference: Susan Walsh Sanderson and Kenneth L. Simons, "Light Emitting Diodes and the Lighting Revolution: The Emergence of a Solid-State Lighting Industry," Research Policy, vol. 43 no. 10, December 2014, pp. 1730-1746. FREE access through 5 Dec. 2014
Also related: Susan Sanderson, Kenneth L. Simons, Judith Walls, and Yin-Yi Lai, "Lighting Industry," in Innovation in Global Industries - U.S. Firms Competing in a New World, National Academies Press, 2008, pp. 163-205. This volume is a focus report of the National Academies. Covered in Business Week (slide show gives industry-specific coverage).
Patent analysis of solid-state lighting development: Kenneth L. Simons and Susan Walsh Sanderson, "Global Technology Development in Solid State Lighting Technologies," International Journal of High Speed Electronics and Systems, vol. 20 no. 2, June 2011, pp. 359-382.
|National Innovation System of the U.S.
A country's national innovation system consists of policies, practices, infrastructure, and people who make new technologies and products come into use. Because technological innovation propels corporate success, national economic growth, incomes, and quality of life, it is crucial to have a good understanding of the national innovation system of your country. When I was asked to create an overview of the U.S. national innovation system, for a new encyclopedia of technology and innovation management, I said no way -- the subject was vast, too long a project and too big for a few pages in an encyclopedia to survey in the scope of an encyclopedia. The last serious "academic" survey of the U.S. national innovation system had been by David Mowery and Nathan Rosenberg way back in 1993, and clearly had taken lots of work. Somehow I gave in to the nagging, the editors allowed more pages, and I lined up a good (now former) Ph.D. student as a coauthor, and wrote this more modern (2008 published 2010, update in 2012) overview of the U.S. national innovation system. The paper analyzes what makes the country's innovation system successful; key government policies, their effects, and suggested changes; overviews of corporate, government, and university spending flows; patent policies and trends; and culture, politics, education, and immigration in relation to national innovation.
2008 version published 2010: Kenneth L. Simons and Judith Walls, "The U.S. National Innovation System," in V.K. Narayanan and Gina Colarelli O'Connor, Encyclopedia of Technology and Innovation, Wiley, 2010, pp. 445-467.
Also related: Kenneth L. Simons, "The U.S. National Innovation System: Potential Insights for Russia" (in Russian, by translation), in I. Danilin and E. Klochikhin (eds.), Innovative Development: International Experience and Russia's Strategy, Moscow: MGIMO-University Press, 2009, pp. 97-119.
|Plant Productivity and Ownership Change
Plant productivity declines on average 1% or more per year before the average plant is acquired by a new owner, but beginning immediately after acquisition, productivity rises even more rapidly. That assessment of total factor productivity stems from a recent, massive study of nearly all Swedish manufacturing plants. The productivity shifts coincide with reductions in output and especially employment. The workforce of the average acquired plant shifts to involve more old-time employees, and the reductions in workforce tend to leave a slightly smaller percentage of women employees, non-native-born employees, and less-educated employees.
These conclusions provide new evidence on what goes on during the average plant or company acquisition. Detailed breakdowns by types of ownership change, including part- and whole-firm acquisitions and divestitures and related and unrelated diversifications, also provide a focused analysis of how the benefits and processes of ownership change vary according to the circumstances of acquisition. The very recent data from this study provide, apparently, the first plant-level evidence on ownership change and productivity in continental Europe, and one of the first analyses of ownership-change related fluctuations in composition of the workforce.
Reference: Donald S. Siegel, Kenneth L. Simons, and Tomas Lindstrom, "Ownership Change, Productivity, and Human Capital: New Evidence from Matched Employer-Employee Data in Swedish Manufacturing," in Producer Dynamics: New Evidence from Micro Data, University of Chicago Press for the National Bureau of Economic Research, 2009, pp. 397-442.
Also related: Donald S. Siegel and Kenneth L. Simons, "Assessing the Effects of Mergers and Acquisitions on Firm Performance, Plant Productivity, and Workers: New Evidence from Matched Employer-Employee Data," Strategic Management Journal, vol. 31 no. 8, August 2010, pp. 903-916.
More about employee experiences: John Marsh, Donald S. Siegel, and Kenneth L. Simons, "Assessing the Effects of Ownership Change on Women and Minority Employees: Evidence from Matched Employer-Employee Data," International Journal of the Economics of Business, vol. 14 no. 2, July 2007, pp. 161-178.
Do your earnings rise so quickly?
|Workplace Practices: Incentives,
Training, and the Cash Till
Why do some companies pay employees little at first but give them big salary increases in later years? Could they afford to keep honest employees for life, and how? How can they build up employees' skills to turn the good ones into top management? And if they're bank employees, what keeps them from running off with the cash?
In this study I've teamed up with my colleague Andy Seltzer, an economic historian, for one of his studies on Australian banking. Australian banks in the 1800s and early 1900s had to attract the few highly-educated school graduates, and keep them in the bank in the long term. Steeply graded salary scales and big old-age pensions gave an incentive to stay with the bank. Not all employees turned out to be geniuses, so while the best went on to top management, the rest became long-term clerks doing routine work in big city offices. Promising employees were rotated through different jobs, and if they were good enough they wouldn't get stuck forever managing small rural branches. The frequent job changes not only built experience; they also promoted honesty. If someone else would inherit the books soon, cheating was hard, particularly with mandatory vacations when employees weren't allowed to step into the office.
This paper provides a chance to reflect on some fundamental facets of employee policy. These policies won't all make sense for your modern company, but it's wise to reflect on how workplace policies like these and their alternatives affect a company.
Reference: Andrew Seltzer and Kenneth L. Simons, "Salaries and Career Opportunities in the Banking Industry: Evidence from the Personnel Records of the Union Bank of Australia," Explorations in Economic History, April 2001, vol. 38 no. 2, pp. 195-224.
||The Ratchet Effect and Firm Regulation
Studies of environmental regulation have argued that regulators are better off with more power rather than less, for example by being able to tailor regulatory standards to specific types of firms or even to individual firms. These views ignore the "ratchet" effect, whereby what the regulator finds out about firms leads it to increase regulation in future, sometimes causing firms not to use the best possible methods for fear of the dreaded ratchet. This paper by Anthony Heyes and myself proves that ratchet effects can indeed apply to environmental regulation of firms, through a simple model of optimally behaving firms and a regulatory agency that works to optimize social welfare. The ratchet effect means that the regulator would prefer not to have the ability to tailor standards for some industries, in order to achieve higher social welfare. This is because "pooling" or "partial separation" equilibria may result in which some firms use inefficient technology in order to avoid heightened regulation. Whether dread of the ratchet really matters depends on industry and technology characteristics, the relative social costs of pollution and output, and the actual (presumably non-optimal) behavior patterns of firms and regulators. The paper's contribution is that we need to pay attention to the ratchet, not ignore it, when planning environmental regulation of firms.
||Economic Growth in Dictatorships
It is well known that some dictatorships achieve rapid economic growth while others suffer rapid decline. This study is apparently the first attempt to explain this phenomenon in a formal economic framework. My colleague Michael Spagat and I teamed up to explain both rapid growth and decline in terms of optimal survival strategies for dictators in different circumstances.
Our model of economic growth involves a tricky dynamic optimization problem. (Technically, it is not time-separable.) We solved it computationally, using intensive sensitivity analyses around the range of plausible parameter values. Our model dictatorships turn out to have bifurcation points, with a spiral of deterioration below the bifurcation and a too-rapid spiral of growth above the bifuraction. The results fit with the limited available empirical evidence, and provide an explanation for the previously noted but unexplained high variance of growth rates among dictatorships.
Whether the theory truly explains what's going on we have no way of knowing given existing evidence. But right or wrong, the methods taken here pioneer an approach useful for future models.
|Preventing Environmental Disaster
There's been a long discussion about the environment and growth -- do we need to slow growth to avoid famines or economic collapse? Opponents of growth such as the Limits to Growth authors from the 1970s never came to agreement with proponents such as the late Julian Simon. The lack of serious dialogue on the issue -- and the extent to which economists have let it lapse -- is unfortunate. It's hard to think of more important issues than the healthy and enjoyable lives of the six billion (and growing) people on our planet.
My solution to the issue is to focus on environmental technology. Both opponents and proponents of growth agree that appropriate use of environmental technology is crucial to the continuance of human lives and lifestyles as we know them. And growth is continuing its way regardless of either camp. So we need an alternative approach.
This paper shows such an approach theoretically, then examines how technologies have been changing in practice, and finally computes crude estimates of real-world technology benchmarks.
I define technology benchmarks: minimum amounts of specific environmental technologies required to allow a continued path of population and economic growth. In a general model of economic growth and environment, I prove: 1. With no environmental problems, growth could continue indefinitely. 2. With environmental impacts, if we don't know the form of the impacts, economic output might fall below any desired level. 3. Environmental technology can ensure a desired minimum growth path. Robust minimal technology paths are also defined, so that amounts of technology greater than or equal to the paths yield desired growth. This accounts for feedback of environmental quality on growth.
Environmental technology progress is estimated by alternative measures in 1970-2004. Crop yields typically have improved roughly 1-2% per year. Pollutant emissions per unit of industry or agriculture typically have fallen 2% to 6% per year. Metal and mineral extraction worldwide per unit of industry typically has fallen by about 3% per year.
To illustrate how technology benchmarks can be estimated using global change models developed by teams of scientists, the World3 model is used to estimate specific minimum environmental technology improvements.
|Making Global Change Models Accessible
Models of global change used to run on mainframe computers with thick, unreadable tomes (at best) for documentation. The situation has been improving, because of software like my Beyond the Limits program. This program makes the World3 model, developed in the 1970s by a team of scientists at MIT, available in user-friendly form. The program includes extensive documentation of the model including click-able model diagrams, and it lets you make changes to the model and see the consequences of your changes.
See also information about the Earth Systems Project in which the software was developed. The project also includes my program The Gaia Hypothesis and Daisyworld.
Reference: Kenneth L. Simons, "Beyond the Limits." Computer software, 1997 (first version 1992).
For a full list of papers, see my papers and publications page.
|Network of Industrial Economists
The NIE organizes a range of activities, with typically 2-4 conferences per year: the main annual conference plus themed conferences. The last conference that I ran, Challenges of Asian Technological Development, took place at the LSE on 31 May 2003.  For information on the Network's email list and on past conferences, see the NIE web site. I was Chairman of the NIE in 2001-2003.