Projects

Role and Value of Innovation

The Role and Value of Innovation in the Pulp and Paper Industry

Project Description

Research Strategy and Methodology

The principle objectives of our research are to measure overall firm-level productivity growth, identify specific types of innovations and quantify their contribution to the overall productivity growth, and assess the impact of innovations on firm profitability and growth after controlling for M&As. Our approach will include analysis of publicly available data which we will collect from industry publications, firm-level data available from Compustat and Thompson’s, and develop a case-study from paper mill visits that will provide valuable insights and guide our broader empirical analysis. Our preliminary research and anecdotal evidence from industry sources suggest that incremental product innovation and M&As are key elements of firm strategy. Hence our proposal to study the productivity impact of innovation also controls for M&As.. Our analysis, by providing a detailed quantitative assessment of firms’ innovation strategies, may facilitate change and greater awareness in an industry that is facing increasing competitive pressures. The main elements of our two-year research plan are described in detail below.

Year One

During the first year of the project, we will conduct a detailed review of the relevant literature, develop a theoretical framework that will form the basis of our analysis, collect and assemble the data germane to our analysis, and conduct visits to selected paper mills to develop a case-study which will provide the broader framework for our empirical analysis. (The actual write-up of the case-study will be completed during the second year.)

a. Literature review

The main focus of the literature review will be to overview the existing theoretical and empirical research in the area of innovation, emphasizing their relevance to the pulp and paper industry. We will also review the literature on M&As since this will be an important control variable.

b. Development of a theoretical model of innovation applicable to the paper industry

Before we outline our data collection and econometric strategies, we highlight how we plan to model the structure of the paper industry. As is well known, the industry is fragmented with some globally dominant firms coexisting with a large number of fringe players. Therefore, it seems appropriate to characterize the industry as dominant oligopoly-competitive fringe. The theoretical models of dominant firm(s)-competitive fringe allow for variation in cost-efficiency levels across the firms within the industry (see Shy, 1995, and the references there). Depending on the level of costs and the changes in costs over time, firms’ performances are affected leading to gain or loss of market share. If the dominant firms as a group innovate (improve cost-efficiency) at a higher rate in the longer-run, then the fringe will slowly shrink; either exit or be acquired by the dominant firms. An outline of the theoretical considerations that will help us examine the innovation dynamics are contained in the papers by Gort and Klepper (1982) and Klepper and Simons (2000). In the longer run these changes may lead to better control over productive capacity and balance between demand and supply. The eventual outcomes critically depend on the final equilibrium market structure and the nature of competition among the dominant firms. As noted earlier, while there is anecdotal evidence documenting the importance of the vintage of the equipment, innovation and M&As, there is a palpable gap in the industry’s systematic knowledge and understanding of how all these factors work together to influence firms’ profits, growth and competitive position.

c. Methodology, hypotheses testing and data

Some of the main questions that we will attempt to answer are:

1. What are the differences across firms within the US in terms of their level of technology and productivity growth?
2. What specific areas of productivity enhancements (e.g., waste reduction, process management, upgrading capital, energy savings) do firms focus on? What are the differences across firms in this dimension? Do the differences across firms in this dimension help explain the differences in their overall productivity growth and performance?
3. Is there a difference between the smaller and larger firms in terms of their innovative activities and overall productivity growth?
4. What are the broad differences between the major countries (e.g., US, Canada , Northern European) in terms of overall productivity growth in the pulp and paper industry, adoption of newer technologies and innovative practices in general?
5. For the US firms, what are the relative impacts of innovation and M&As on firms’ productivity growth and profitability?

To answer these questions, we will collect data and conduct our analysis at several levels. The data will be at the firm-level and is either publicly available, or can be acquired from specialized sources. We will be using several data bases in this project. The Paperchem database covers international patents and literature on paper and pulp technology. The PIRA database will provide additional information on technical issues, market statistics, production processes, machinery and equipment, environmental issues and new products. The Thompson's Financial databases on mergers and acquisitions and joint ventures will provide us with information on the kinds of M&As, their value, licensing arrangements, objectives of collaborations etc. We will also use COMPUSTAT and Dun and Bradstreet to get relevant firm level financial and other information. The UNCTAD database will provide information on exports, imports and trade barriers. While mill-level data may be ideal, they are nearly impossible to get without cooperation of the mills. We will not require firms to provide data or participate in surveys for this project. Below we present some additional details regarding our approach:

1. For the US firms, we will use annual firm-level data for the pulp and paper companies and compute firm-level TFP.1 The annual TFP series will be used to show the differences across firms in the year-to-year changes as well as the cumulative productivity differences over time.2 Apart from computing TFP, we will use data on R&D expenditures and patents and make similar comparisons across firms. We will use data on firm size - based on employment levels and capital stock – and examine how differences in firm size correlate with the levels and growth of productivity. Finally, we will need data on, for example, profits, sales, imports and exports, aggregate new investment spending, etc. The above data are all available from Compustat (US and Global editions) Thomsons Financial and Dunn and Bradstreet databases.
2. 2. We will use several publications devoted to the pulp and paper industry to gather information on the availability of new technologies and their adoption. Since these data are not available in readily useable form, we will have to create a database by examining the relevant publications and coding the necessary information. This is a highly labor-intensive task and will entail a significant amount of work during the first year by the co-PIs and the graduate student. For example, we will use: (a) Pulp and Paper; (b) Solutions; and (c) TAPPI Journal (d) PIRA database (e) Paperchem. TAPPI provides extensive information on various process and product innovations. The ‘Industry News’ section of Solutions contains detailed US and international news on sales of new technologies and capital equipment. The ‘Orders and Deliveries’ section of Pulp and Paper contains extensive information on purchase of news machines, adoption of new processes and other information for US firms. Our strategy will be to first identify some of the key processes related to pulping, drying, bleaching, waste reduction, and energy conservation among others. Second, identify existing and emerging technologies in these areas from the TAPPI Journal and related sources and databases listed above. Third, use the Solutions and the Pulp and Paper publications to assemble a database on actual adoptions. Since these sources identify the firms to whom the sales are made, we will end up with a firm-level database on technology adoption. In addition to the above, there are other sources that provide useful data. These include www.paperloop.com, www.PulpandPaper.net, and www.nlkassociates.com. For example, the ‘paperloop’ source provides information on mill maintenance and upgrades, automation upgrades and IT upgrades; the ‘NLK associates’ source contains mill-level capacity data for over 3,000 global producers in 100 countries. Some of these data are available for free, while some others are available for a fee. Overall, these data can be combined to produce a rich firm-level database. Needless to add, given the diverse nature of the sources, our information/data base will not be exhaustive of all the firms in the industry, but a useful representative sample. We anticipate that this database will be complied and verified for accuracy by the end of the first year. This information will be matched up with the broader information described in (1) above and will allow us to examine what types of innovative activities lead to greater overall productivity growth. One way of conceptualizing this is as follows: (i) compute firm level TFP; (ii) assemble a firm level database on key processes P i ; (iii) empirically quantify the contribution of the P i’s to TFP: TFP=f{P 1; P 2; P 3; …}, controlling for M&As. Apart from the sources mentioned earlier, we will consult with industry experts regarding the key elements of firms’ innovation strategy. We will seek guidance from the American Forest & Paper Association (AF&PA) and other industry organizations regarding the types of companies or industry organizations that can be contacted for information.
3. 3. We will visit selected paper mills after consultations with CPBIS. The mills will be chosen after taking into consideration factors related to the vintage of machines, union and non-union status, among others. The primary objective of the mill trips will be to develop a case-study containing details about the specific innovative practices that paper mills engage in and feel are central to maintaining their longer-run competitiveness. The mill trips and case-study will provide us with valuable information that will guide our broader empirical study and help us identify the specific types of information and data we will need to gather as detailed in item #2 above. Our previous mill trips have been useful in identifying such specific information and we are confident that a more in-depth examination will uncover important details. ( Note: We will not ask the mills/companies for any proprietary data, but merely broad information.)
4. 4. For the cross-country comparisons, we will use pulp and paper industry-wide data from the US , Canada and the main European paper producers to construct measures of TFP. As before, the TFP measures will be constructed using annual data and the annual TFP time-series can be used to generate a cumulative productivity series for the sample period. These TFP series will inform us of the extent of year-to-year differences in productivity growth across nations as well as the differences in levels that emerge over time. This analysis will help us visualize the potential extent of technology gap across countries. The data for constructing the TFP measures are readily available from the OECD cross-country database.

At the end of the first year, we will provide CPBIS with an interim report documenting the progress of the project and the initial findings.

Year Two

The second year of the study will involve analyzing the data, hypothesis testing, report writing, writing the case-study and dissemination of the results of the projects. These steps are discussed in detail below:

a. Data Analysis and Hypothesis Testing

Using multivariate regression analysis, some of the key relationships we will examine are:

1. Profits, growth rates and exports are positively related to innovation measured by TFP growth. We will estimate regressions using firm-level data, with the variable of interest as the dependent variable and the measure of innovation as an independent variable. Among the additional control variables, we will include M&As. These regressions will also allow us to assess the relative impacts of innovation and M&As.
2. Firms that combine innovation with M&As are more successful than firms that pursue only one of these strategies. The regression will be similar to those above, but will include an interaction term between the measures of innovation and M&As. The sign on the innovation-M&A interaction term will shed light on this hypothesis.
3. Observed TFP growth is positively related to firm-specific innovative activities in the dimensions of, for example, energy savings, waste reduction, worker training and R&D expenditures. The estimated regression with firm-level data will have TFP as the dependent variable, and the specific measures of innovative activity as the independent variables. This will allow us to assess the relative contributions of the specific innovative activity.
4. Innovation and TFP growth are negatively related to the age of capital. The regressions will include the age of capital as the explanatory variable, with the various measures of innovation as the dependent variable. (This part is critically dependent on us getting data on vintage of capital. If we do not get it for a large sample of firms, we will examine this for a selected set of firms.)
5. Firm size is positively related to innovation. The regressions will include firm size as the explanatory variable, with alternate measures of innovation as the dependent variable. (To the extent data are available, we will also control for vintage when examining this relationship.)

In addition we note the following:

1. We will use firms’ aggregate new investment spending data, available from the Compustat, as another indicator of technological improvements. Since new physical capital typically embeds newer technology or greater efficiency, this is a good summary measure;
2. Several mills we visited mentioned they have incentive schemes where a worker or groups of workers could be rewarded if they proposed, and the company subsequently implemented, cost-saving innovations. What this implies is that we may have to assume that some fraction of the firm’s skilled workforce is actively engaged in R&D-type activities, even thought they may not be formally classified as R&D workers. In other words, to get the full picture of R&D, we will have to use the R&D expenditures data plus some correction for the skilled workforce aspect.

While a detailed description of the process and drivers of innovation as well as the key characteristics of successful innovating firms is in itself invaluable in broadening the industry understanding of these issues, the research methodology also provides for a more rigorous regression based analysis as discussed above.

b. Report writing and presentation of results

The last stage of the research involves report-writing from our empirical analysis, writing the case-study based on our mill-trips, creation of educational and training materials for the industry as well as articles for publication in the pulp and paper industry trade publications. We propose to present the findings at industry and academic conferences. We will also develop journal quality articles from this project that will find a home in high quality economics and management journals. It is important to note that this analysis can be conducted without any confidential information from industry sources. The site visits and consultations with firms will undoubtedly serve to enhance our understanding of the industry and how innovation and technology adoption takes place, but survey data from individual firms is not necessary to conduct this analysis.

Desired Outputs and Contributions to Theory and Research

We anticipate that the proposed project will generate the following deliverables:

• Research reports detailing the objectives and drivers of innovation in the coated and uncoated papers sector, a list of critical firm-specific characteristics and factors that facilitate and impede these strategies and their successful implementation and a quantitative evaluation and comparison of the impact of innovation on firm profitability and growth, including an assessment of technologies that provide the most bang for the buck.
• Educational material that describes the main methodology used and results of the study, which would be suitable for use in the classroom.
• Training of one graduate student who will become knowledgeable about the P&P industry in general and develop the ability to evaluate and compare different innovation strategies pursued by the firms. This highly trained graduate with industry-specific knowledge will be a valuable resource to the industry, and is clearly a way in which academe-industry cooperation and collaboration is fostered.

The proposed research would make several contributions towards advancing our knowledge on firm strategy in the paper industry.

• First, the research seeks answers to several interesting and timely issues: What types of innovations (or cost-efficiencies) do firms focus on in their day-to-day operations? How do they evaluate the impact of such innovations? What factors related to the firm’s internal and external environment hinder or facilitate innovation? Our project, by providing answers to these questions, will contribute to the literature on technological change by enhancing our understanding of firms’ innovation strategies in general and the P&P industry is particular.

• Second, it will provide an understanding of the relative roles of M&As and innovations in their impact on firm profitability and growth in the P&P industry.

• Third, our interdisciplinary approach will aim to bring together the economics and strategic management literatures on firm strategies related to innovation, M&As and adaptability in an era of rapidly increasing global competition.

• Fourth, by using the dominant oligopoly-competitive fringe model for the pulp and paper industry and the quantitative evaluation of the different firm strategies, we will help enhance the industry’s understanding of the drivers and impact of different innovation strategies on industry and firm level profits and growth.

• Fifth, the methodology developed could be used as a teaching tool in economics and management classes in understanding the impact of different firm strategies on firm profitability and growth.

• Finally, the proposal will include hands-on involvement by a graduate student in research projects and training of the graduate student.

Desired Outputs and Contributions to Industry/Other Groups

The outputs and contributions to industry and other groups are similar to the academic contributions listed above.

• Our proposal aims to provide a detailed analysis of the drivers of innovation (or cost-efficiency) in the P&P industry. Given that the objective of the US paper industry is to improve its bottom line, our study will provide a comparative analysis of the main types of innovations and their relative contributions to the bottom line: Which types of innovations are likely to provide a bigger bang for the buck? Because we have to control for M&As while studying innovation, we will be able to compare innovation versus consolidation and quantify their relative impacts: The quantitative evaluation of alternative firm strategies will provide managers with useful insights that can aid in their choice of firm strategy and reduce the uncertainty associated with this process. This would be a useful contribution given that the industry generally has been slow to embrace change.
• By providing an assessment of the costs and benefits of alternate strategies, we will aim to raise the consciousness in the industry regarding the value of innovations. While innovations and cost savings may often contribute a relatively small amount to a firm’s bottom line in a given year, the cumulative impact over several years can be large. Those mills/companies that underestimate the impact of such incremental innovations are likely to lose their competitive advantage in the longer-run.
• The increasingly global nature of the industry suggests that the firms have to develop appropriate strategic and tactical plans, both to be effective against foreign competition and to exploit new opportunities that arise in the global market place. Our analysis will provide useful information about the role of global factors as drivers of innovation, and offer insights on the effectiveness of these strategies in moderating the impact of foreign competition on the domestic industry structure and profitability.
• Educational materials that describe the main methodology used and results of the study will be presented in a manner that is suitable for training of industry personnel and that can facilitate the use of the methodology at the firm level.
• Another important outcome of the project will be a highly trained graduate student knowledgeable about innovations, M&As, global economic and trade issues, and with particular knowledge about P&P industry and an ability to evaluate different firm strategies and quantify their impact on profitability and growth. The student will be a valuable resource for corporations in the P&P industry, and members of CPBIS will have access to this student.

It is very important to clarify how the results of the research will be disseminated to the industry. We propose to present the general results of this project to industry groups at conferences and other forums. These results, working papers and the related power point presentations will also be made available online at CPBIS’ website.

1 The literature on TFP computations is both standard and voluminous. Ghosal (2003a), for example, reviews the key aspects of this literature. The theoretical basis for TFP is as follows. Consider the following production function for a firm (omitting the ‘time’ subscripts): q=Af(K,L,E,M), where q measures output or production, K, L, E and M are the broad inputs related to capital, labor, energy and materials, and A is an index of technology which is unobserved. From such an equation, we can subtract the input-share weighted growth of inputs from the growth of output. The residual is the growth of output that cannot be explained by the growth of inputs. This residual is called the technology or productivity residual: i.e., output improved in part due to productivity or technological enhancements. Regarding implementation, i ndustry-level TFP will be calculated from annual data readily available from the Census of Manufactures and the Annual Survey of Manufactures. Firm-level TFP can be computed from analogous data available from the Compustat. Fin a lly, cross-country TFP co m parisons can be made using data that are available from the OECD. Both the researchers in this project have familiarity with these datasets.

2 Ghosal (2003b) presents the annual and cumulative TFP series for the US P&P industry.