(joint with Laure de Preux and Ulrich Wagner)
We estimate the impacts of an energy tax – the Climate Change Levy (CCL) – on the manufacturing sector using panel data from the UK production census. Our identification strategy builds on the comparison of trends in outcomes between plants subject to the CCL and plants that were granted an 80% discount on the levy after joining a so-called Climate Change Agreement (CCA). Since the CCAs stipulate specific targets for energy usage or carbon emissions, this comparison yields a lower bound on the impact of the discount. To address a likely selection endogeneity in CCA participation, we adopt an IV approach that exploits exogenous variation in pollution discharges that determined eligibility for CCA participation. We find robust evidence that CCA participation had a strong positive impact on growth in both energy intensity and energy expenditures. An analysis of fuel choices at the plant level reveals that this effect is mainly driven by stronger growth in electricity use and translates into a positive impact on CO2 emissions. We do not find any statistically significant impacts of the tax on employment, gross output or total factor productivity. We conclude that, had the CCL been implemented at full rate for all businesses, further cuts in energy use of substantial magnitude could have been achieved without jeopardizing economic performance.
Available as CEP discussion paper.
Tuesday, 7 April 2009
Absolution through Innovation? - The Willingness to pay for climate change-related R&D
(joint with Susana Mourato and George MacKerron)
Abstract: With increasing concern about climate change there have been many calls recently to increase direct government subsidies for R&D spending on climate change-related research. While governments in many industrialised countries have put a number of different schemes in place, spending still falls short of the doubling in spending levels many experts are calling for. Concern for climate change has equally made so-called voluntary offsetting schemes popular, ranging from tree-planting projects to subsidies for the purchase of energy-saving light bulbs for schools in developing countries. In this paper we use a choice modelling approach to investigate the feasibility of channelling some of these ‘charitable’ contributions for climate change projects into R&D spending.
Looking at aggregate figures, it seems plausible that this could have significant effects. For example, in the UK, aggregate government spending on energy-related R&D and private spending on voluntary offsetting are roughly commensurate. However, a key difference between typical offsetting schemes and spending on R&D is the certainty of the impact on CO pollution. At least on paper, a tree-planting project offers a clear account of how much CO is avoided: simply use the amount of CO a tree ‘breathes in’ over its lifetime. With spending on R&D any CO reductions are uncertain and occur, if at all, at an unspecified future date---although they might potentially be much larger. The feasibility of an ‘innovation offsetting’ scheme consequently depends on the perception of its future carbon-reduction potential, and of the risks involved, among typical participants in the carbon offset market.
In this exploratory study we have used a web-based choice experiment survey to investigate preferences for innovation offsetting. We presented respondents with hypothetical scenarios for offsetting the carbon emissions from their leisure and business flights. In each scenario, respondents could: i) buy a standard, verified offset, which promised to save a fixed amount of carbon; ii) buy an 'innovation offset', which contributed to a fund providing grants for carbon-reducing R&D; or iii) not buy any offset. From scenario to scenario we varied the proportion of carbon emissions saved by the conventional offset option, the area of research funded by the innovation offset (out of eight possible research areas), and the prices of the offsetting options.
Our results suggest that there is potential demand for innovation offsetting. Our respondents show a marked preference for investing in a portfolio comprising all possible research areas, rather than choosing individual research areas. In fact, some individual research areas have a negative impact on utility under certain model specifications (these areas include hydrogen, nuclear power, bio-fuels, and planetary engineering). Apart from the expected lower sensitivity to the price of offsets for business flight emissions---which are hypothetically to be paid for by the respondent's organisation---we find no significant discrepancy between offsetting behaviour in relation to leisure and business flight emissions. Our ‘carbon innovation exchange rate’ ranges from £0.30 to £0.40 per tonne of CO. To our knowledge, this is the first study to investigate demand for offsetting schemes that invest in carbon-reducing research and innovation.
Abstract: With increasing concern about climate change there have been many calls recently to increase direct government subsidies for R&D spending on climate change-related research. While governments in many industrialised countries have put a number of different schemes in place, spending still falls short of the doubling in spending levels many experts are calling for. Concern for climate change has equally made so-called voluntary offsetting schemes popular, ranging from tree-planting projects to subsidies for the purchase of energy-saving light bulbs for schools in developing countries. In this paper we use a choice modelling approach to investigate the feasibility of channelling some of these ‘charitable’ contributions for climate change projects into R&D spending.
Looking at aggregate figures, it seems plausible that this could have significant effects. For example, in the UK, aggregate government spending on energy-related R&D and private spending on voluntary offsetting are roughly commensurate. However, a key difference between typical offsetting schemes and spending on R&D is the certainty of the impact on CO pollution. At least on paper, a tree-planting project offers a clear account of how much CO is avoided: simply use the amount of CO a tree ‘breathes in’ over its lifetime. With spending on R&D any CO reductions are uncertain and occur, if at all, at an unspecified future date---although they might potentially be much larger. The feasibility of an ‘innovation offsetting’ scheme consequently depends on the perception of its future carbon-reduction potential, and of the risks involved, among typical participants in the carbon offset market.
In this exploratory study we have used a web-based choice experiment survey to investigate preferences for innovation offsetting. We presented respondents with hypothetical scenarios for offsetting the carbon emissions from their leisure and business flights. In each scenario, respondents could: i) buy a standard, verified offset, which promised to save a fixed amount of carbon; ii) buy an 'innovation offset', which contributed to a fund providing grants for carbon-reducing R&D; or iii) not buy any offset. From scenario to scenario we varied the proportion of carbon emissions saved by the conventional offset option, the area of research funded by the innovation offset (out of eight possible research areas), and the prices of the offsetting options.
Our results suggest that there is potential demand for innovation offsetting. Our respondents show a marked preference for investing in a portfolio comprising all possible research areas, rather than choosing individual research areas. In fact, some individual research areas have a negative impact on utility under certain model specifications (these areas include hydrogen, nuclear power, bio-fuels, and planetary engineering). Apart from the expected lower sensitivity to the price of offsets for business flight emissions---which are hypothetically to be paid for by the respondent's organisation---we find no significant discrepancy between offsetting behaviour in relation to leisure and business flight emissions. Our ‘carbon innovation exchange rate’ ranges from £0.30 to £0.40 per tonne of CO. To our knowledge, this is the first study to investigate demand for offsetting schemes that invest in carbon-reducing research and innovation.
Modern Management: Good for the Environment or just Hot Air?
(joint with Nick Bloom, Christos Genakos and Raffaella Sadun)
We use an innovative methodology to measure management practices in over 300 manufacturing firms in the UK. We then match this management data to production and energy usage information for establishments owned by these firms. We find that establishments in better managed firms are significantly less energy intensive. They use less energy per unit of output, and also in relation to other factor inputs. This is quantitatively substantial: going from the 25th to the 75th percentile of management practices is associated with a 17.4% reduction in energy intensity. This negative relationship is robust to a variety of controls for industry, location, technology and other factor inputs. Better managed firms are also significantly more productive. One interpretation of these results is that well managed firms are adopting modern lean manufacturing practices, which allows them to increase productivity by using energy more efficiently. This suggests that improving the management practices of manufacturing firms may help to reduce greenhouse gas emissions.
Available as NBER working paper.
We use an innovative methodology to measure management practices in over 300 manufacturing firms in the UK. We then match this management data to production and energy usage information for establishments owned by these firms. We find that establishments in better managed firms are significantly less energy intensive. They use less energy per unit of output, and also in relation to other factor inputs. This is quantitatively substantial: going from the 25th to the 75th percentile of management practices is associated with a 17.4% reduction in energy intensity. This negative relationship is robust to a variety of controls for industry, location, technology and other factor inputs. Better managed firms are also significantly more productive. One interpretation of these results is that well managed firms are adopting modern lean manufacturing practices, which allows them to increase productivity by using energy more efficiently. This suggests that improving the management practices of manufacturing firms may help to reduce greenhouse gas emissions.
Available as NBER working paper.
The effect of industrial policy on corporate performance: Evidence from panel data
(joint with Chiara Criscuolo, Henry Overman and John van Reenen)
Abstract: Industrial or business support policies designed to raise productivity and
employment are a common feature of the policy landscape. Despite this ubiquity, rigorous micro-econometric evaluation of their causal effect is rare primarily because of the difficulty of achieving credible identification. In this paper we tackle this problem by exploiting multiple changes in the area-specific eligibility criteria for a major business support scheme in the UK (“Regional Selective Assistance”). These changes arose because of the need to comply with the European Commission revisions of the eligibility criteria and coverage rules. We match over twenty years of administrative panel data on program participation and firm performance from the Census Bureau to investigate the causal impact of the policy on employment, investment, productivity and entry/exit. Using an instrumental variable approach we find that the program has had a positive effect on both employment and investment, which naïve estimators underestimate. There is no statistically significant effect on total factor productivity, however. There is also some evidence that the program, by supporting less efficient enterprises, may slow down reallocation from less efficient plants, negatively affecting aggregate productivity growth.
Abstract: Industrial or business support policies designed to raise productivity and
employment are a common feature of the policy landscape. Despite this ubiquity, rigorous micro-econometric evaluation of their causal effect is rare primarily because of the difficulty of achieving credible identification. In this paper we tackle this problem by exploiting multiple changes in the area-specific eligibility criteria for a major business support scheme in the UK (“Regional Selective Assistance”). These changes arose because of the need to comply with the European Commission revisions of the eligibility criteria and coverage rules. We match over twenty years of administrative panel data on program participation and firm performance from the Census Bureau to investigate the causal impact of the policy on employment, investment, productivity and entry/exit. Using an instrumental variable approach we find that the program has had a positive effect on both employment and investment, which naïve estimators underestimate. There is no statistically significant effect on total factor productivity, however. There is also some evidence that the program, by supporting less efficient enterprises, may slow down reallocation from less efficient plants, negatively affecting aggregate productivity growth.
Productivity Dispersion, Competition and Productivity Measurement
Abstract: A startling fact of firm level productivity analysis is the large and persistent differences in both labour productivity and total factor productivity (TFP) between firms in narrowly defined sectoral classes. The competitiveness of an industry is potentially an important factor explaining this productivity dispersion. The degree of competition has also implications for the measurement of TFP at the firm level. This paper firstly develops a novel control function approach to production function and TFP estimation explicitly taking imperfect competition into account. This addresses a number of issues with the control function approach to productivity estimation. Secondly, applying this new approach to UK data it shows that productivity dispersion on average is about 50 percent higher than with standard TFP measures. It also shows that accounting for imperfect competition matters for estimates of the persistence of TFP. Thirdly, the paper finds a negative relationship between competition and productivity dispersion.
Available as CEP discussion paper.
Available as CEP discussion paper.
Distance to Which Frontier? Evidence on Productivity Convergence from International Firm-level Data
(joint with Eric Bartelsman and Jonathan Haskel)
Abstract: An extensive literature on the convergence of productivity between countries examines whether productivity is pulled towards the global frontier country, perhaps due to learning and knowledge spillovers. More recently, studies within countries use the wide dispersion of productivity across firms to explore convergence to the national frontier. Given this within-country dispersion however between country-dispersion is hard to interpret, for it is quite possible that the best firms in a laggard average country are above at least some firms in a leading average country. This paper therefore uses micro data sets across many countries to build better measures of global and national frontiers and firms’ distance from them. Using UK data, we then find that (a) the national frontier exerts a stronger pull on domestic firms than does the global frontier and (b) the pull from the global frontier falls with technological distance, while the pull from the national frontier does not. This result suggests that firms might lag so far technologically that they cannot learn from the global frontier, while they still are able to benefit from domestic knowledge.
Available as CEPR discussion paper.
Abstract: An extensive literature on the convergence of productivity between countries examines whether productivity is pulled towards the global frontier country, perhaps due to learning and knowledge spillovers. More recently, studies within countries use the wide dispersion of productivity across firms to explore convergence to the national frontier. Given this within-country dispersion however between country-dispersion is hard to interpret, for it is quite possible that the best firms in a laggard average country are above at least some firms in a leading average country. This paper therefore uses micro data sets across many countries to build better measures of global and national frontiers and firms’ distance from them. Using UK data, we then find that (a) the national frontier exerts a stronger pull on domestic firms than does the global frontier and (b) the pull from the global frontier falls with technological distance, while the pull from the national frontier does not. This result suggests that firms might lag so far technologically that they cannot learn from the global frontier, while they still are able to benefit from domestic knowledge.
Available as CEPR discussion paper.
Subscribe to:
Posts (Atom)