Output in the Innovation Process

1. Measurement of Output in the Innovation Process

In terms of measuring output in the innovation process, patents have been used despite their shortcomings. First, one of the issues with viewing patents as output is the lag between receiving a patent and introducing a new product into the market incorporating this particular innovation (Alexopolous, 2011). Second, patents do not capture innovation that occurs outside of the patent system (Moser, 2013). Third, not all patented innovations enter the market embodied by a product (Gardner and Joutz, 1996).

Considering these shortcomings of using patents as output in the innovation process, one of the recently proposed measurements of innovation implies new titles, which implies manuals describing a new technology or product. The important point about using new titles as a proxy for output is that they are not released until the product is already in the market. In other words, they precede diffusion. Therefore, new titles circumvent two of the shortcomings of patents, namely the discrepancy between receiving the patent and introducing the product into the market using this particular patent as well as the possibility that the patent may not lead to a new product at all (Alexopoulos, 2011).

2. Production and Determinants of Innovations

This section discusses the country- as well as the firm-level determinants of innovation.

Among the country-level determinants, a policy that is conducive to innovation includes strong property rights, effective patent laws, and a strong legislature. There needs to be an incentive for people to innovate. The patent system seeks to solve this by trying to provide inventors with monopoly status on their products. There are varying degrees of success in the effort. In the semiconductor industry, patents have not guaranteed monopolies, while in chemical companies they have been very successful (Flaherty, 1984). One would argue that when innovation applies to an intermediate good as in the case of semiconductors, it may be difficult to establish a monopoly position. However, when innovation is applied to the final product as in the case of chemical products a monopoly position may be possible.

Patent policies can have some negative effects as well. Our current policy may hinder innovation by preventing the diffusion of ideas (Brander, 2010; Moser, 2013). A solution to this problem is the formation of pools. Firms inside the pools can share information and reduce litigation costs; however, the same cannot be said for outside firms. Results have been mixed on the usefulness of pools. In one case, innovation increased in response to a pool for CDs (Moser, 2013). Compulsory licensing can also increase the efficacy in spurring innovation and the diffusion of ideas (Moser, 2013).

Clearly, the effectiveness of the legal system in protecting intellectual property is a function of the development level of a country. This and the fact that a developing country may not have the resources to put toward innovation generally make such countries imitators instead of innovators (Madsen, Islam, and Ang, 2010). However, developing countries that are distanced from the technological frontier still enjoy output gains through imitation. Even imitation requires some degree of education and learning-by-doing, which may facilitate a country’s convergence to the technological frontier. Thus, a developing country stands to gain more from knowledge seeking expenditures than OECD countries (Madsen et al., 2010).

When it comes to innovations, social and private costs and benefits are not always properly aligned. For instance, an alternative energy that would slow down or reverse the adverse effects of climate change, such as preventing the melting ice caps would have enormous social benefits (Brander, 2010). However, funds may not be channeled to innovations with large social benefits because of small private gains and large private costs. Motivating innovations in areas where social and private benefits diverge requires proper alignment of resource and energy prices so that these prices reflect their full social costs (Brander, 2010).

In terms of firm-level determinants of innovation, R&D is considered as the main input. There are also others that are not as widely recognized. A company’s expenditures on product design, marketing and customer support as well as human capital and organizational development can all be included in innovation-related expenditures (inputs). If these expenditures were included, in 2007, intangible investments would be $1.3 trillion dollars higher than the BEA’s analysis (Corrado and Hulten, 2010). Advertising by itself is as large as R&D spending in the aggregate (Mcgratten and Prescott, 2014). Thankfully, there have been some updates to BEA’s estimation methods. In 2013, the BEA expanded its coverage of intangibles beyond software to include investment in R&D and artistic originals. It also created a new category of fixed investment called intellectual property products (Mcgratten and Prescott, 2014).

The size of a firm is an important determinant of R&D and also the effectiveness of R&D (Scherer, 1965). The marginal product of research increases as size increases, and it decreases as the number of R&D employees increases (Vernon and Gusen, 1974). Size could increase the marginal product as a result of a wide range of operations (Vernon and Gusen, 1974). The wide range of operations creates more possibilities for inventions to be useful. Since the marginal product rises with firm size, it is more efficient to place a given team of R&D researchers in a large firm (Vernon and Gusen, 1974). In Scherer’s study, patents were found to be closely related to the amount of R&D employees with a lag. For instance, the number of R&D employees in 1955 explained 72% of the variation in 1959 patents (Scherer, 1965; Cohen and Klepper, 1996). Although, large firms have many advantages, small firms may be more efficient. They patent more inventions per unit of input (Scherer, 1965).

Although, as previously stated, larger firms generate less innovation per dollar of R&D spending, they have considerable advantages. They have greater outputs over which they can apply their innovations, they may be able to use technological spillovers better, and they are able to spread their costs amongst more products (Cohen and Klepper, 1996). Also, large firms tend to be more diversified relative to small firms, this increases the ability of large firms to keep R&D spillovers in-house. There are different theories on why large firms are less productive per R&D dollar. One is that they pursue more research at the margin, because they are able to apply the results over all their products (Cohen and Klepper, 1996). More marginal research leads to less productive R&D on average.

Another determinant of firm level R&D is the firm’s concentration level in a market. In high tech environments the cost and uncertainty of R&D may be too high to stimulate R&D investment in the absence of high concentration (Angelmar, 1985). More concentration in a market can insulate a firm from some risk. On the other hand, smaller firms might not have the financing available to engage in R&D if the payback period is long. Also, a high speed of imitation decreases the amount of money a firm can appropriate from their inventions. Thus, longer payback times and high speeds of imitation may reduce R&D, unless the market is concentrated (Angelmar, 1985). “A 10% increase in concentration adds .77 percentage points for a business with long product development, it adds only .04 percentage points to research spending where product development time is short.” (Angelmar, 1985).

Companies that are vertically integrated benefit more from innovation compared to the non-integrated companies (Armour and Teece, 1980). Therefore, one would expect that vertically integrated firms may spend more on innovations. Specifically, in petroleum industry, adding another stage of integration to the production process increases basic research spending by $700,000 and applied research by $500,000 (Armour and Teece, 1980).

3. Measuring the Diffusion of Technology

There are different ways to measure diffusion. Usage lags are a very useful method, because measures of technological adoption can be compared across technologies. Additionally, usage lags do not require a long time series dataset, which is not widely available except for the U.S. (Comin, Hobijn, and Rovito, 2008). Countries can also be compared by developmental period rather than time period which is useful. If Argentina lags United States electricity usage in the year 2000 by 15 years, then you can compare other factors using that lag to see their discrepancies. Then in 2010 you can see how the lag has changed, instead of just looking at electricity differences in terms of kilowatt hours used. Research indicates that adoption lags are large, with mean of 45 years, median of 32 years, and standard deviation is 39 years (Comin and Hobijn, 2010). Another interesting finding is that newer technologies have diffused much faster than older technologies. Technologies invented 10 years later are on average adopted 4.3 years faster. For instance, PCs and the internet took less than 15 years for half the sample countries to adopt. Adoption lags have been decreasing steadily for the past 200 years, which is consistent with an increasing rate of diffusion. Not surprisingly, compared to low per capita income countries, high per capita income countries have shorter adoption lags (Comin and Hobijn, 2010).

Diffusion can also be measured on a firm by firm basis. When technology is adopted by a firm, there is a learning cost associated with its adoption (Conley and Udry, 2010). Learning can be relatively less expensive if there is social learning, as opposed to individual learning. In Ghana for instance, a pineapple farmer is more likely to adopt a new technology if his neighbors adopt it as well (Conley and Udry, 2010). This could be an example of social learning in practice, or it could be the result of an unobserved variable (Conley and Udry, 2010). Another interesting takeaway is that veteran pineapple farmers are more likely to be in each other’s information neighborhood than expected by chance (Conley and Udry, 2010). Also, novice farmers react in more measurable ways to good news than veteran farmers (Conley and Udry, 2010).

References

Ang, James B., Md. Rabiul Islam, and Jakob B. Madsen. 2010. “Catching up to the technology frontier: the dichotomy between innovation and imitation.” The Canadian Journal of Economics, 43(4): 1389-1411.

Angelmar, Reinhard. 1985. “Market Structure and Research Intensity in High-Technological- Opportunity Industries.” The Journal of Industrial Economics, 34(1): 69-79.

Alexopoulos, Michelle. 2011. “Read All about it!! What Happens Following a Technology Shock?” The American Economic Review, 101(4): 1144-1179.

Armour, Henry Ogden, and David J. Teece. 1980. “Vertical Integration and Technological Innovation.” The Review of Economics and Statistics, 62(3): 470-474.

Brander, James A. 2010. “Presidential Address: Innovation in retrospect and prospect.” The Canadian Journal of Economics, 43(4): 1087-1121.

Cohen, Wesley M., and Steven Klepper. 1996. “A Reprise of Size and R & D.” The Economic Journal, 106(437): 925-951.

Comin, Diego, Bart Hobijn, and Emilie Rovito. 2008. “Technology Usage Lags.” Journal of Economic Growth, 13(4): 237-256.

Comin, Diego, and Bart Hobijn. 2010. “An Exploration of Technology Diffusion.” The American Economic Review, 100(5): 2031-2059.

Conley, Timothy G., and Christopher R. Udry. 2010. “Learning about a New Technology: Pineapple in Ghana.” The American Economic Review, 100(1): 35-69.

Corrado, Carol A., and Charles R. Hulten. 2010. “How Do You Measure a “Technological Revolution”?” The American Economic Review, 100(2): 99-104.

Flaherty, M. Therese. 1984. “Field Research on the Link Between Technological Innovation and Growth: Evidence from the International Semiconductor Industry.” The American Economic Review 74(2): 67-72.

Gardner, Thomas A., and Frederick L. Joutz. “Economic Growth, Energy Prices and Technological Innovation.” The Southern Economic Journal, 62(3): 653-666.

Gusen, Peter, and John M. Vernon. 1974. “Technical Change and Firm Size: The Pharmaceutical Industry.” The Review of Economic and Statistics, 56(3): 294-302.

McGrattan, Ellen R., and Edward C. Prescott. 2014. “A Reassessment of Real Business Cycle Theory.” The American Economic Review, 104(5): 177-182.

Moser, Petra. 2013. “Patents and Innovation: Evidence from Economic History.” The Journal of Economic Perspectives, 27(1): 23-44.

Scherer, F. M. 1965. “Firm Size, Market Structure, Opportunity, and the Output of Patented Inventions.” The American Economic Review, 55(5): 1097-1125.