Abstract: We study learning by doing (LBD) by New York City yellow taxi drivers. We use data on each individual fare during 2009 to construct measures of driver productivity overall, across neighborhoods in which production is easy versus hard, and with measures of neighborhood-specific experience. The breadth of data allows us to introduce a collection of controls for unobserved demand conditions and driver selection that are often impossible in the LBD literature. We find that driver learning is significant but may be overstated in the absence of suitable controls. In addition, new drivers perform worse in quantifiably more difficult situations, although learning erases these performance differences fairly quickly. Finally, we find that driver experience in specific neighborhoods has a larger impact than overall experience on productivity in those neighborhoods. 1

Learning how to utilize new technologies is a key step in innovation, yet little is known about how firms actually learn. This paper examines firms ’ learning behavior using data on their operational choices, profits, and information sets. I study companies using hydraulic fracturing in North Dakota’s Bakken Shale formation, where firms must learn the relationship between fracking input use and oil production. Using a new dataset that covers every well since the introduction of fracking to this forma-tion, I find that firms made more profitable input choices over time, but did so slowly and incompletely, only capturing 67 % of possible profits from fracking at the end of 2011. To understand what factors may have limited learning, I estimate a model of fracking input use in the presence of technology un-certainty. Firms are more likely to make fracking input choices with higher expected profits and lower standard deviation of profits, consistent with passive learning but not active experimentation. Most firms over-weight their own information relative to observable information generated by others. These results suggest the existence of economically important frictions in the learning process. 1

T he concept of a learning curve for individuals has been in widespread use in the psychology literature since the beginning of the twentieth century. Ebbinghaus (1885) famously demonstrated the fi rst learning curve by memorizing ever-longer strings of nonsense syllables; Bryan and Harter (1899) studied learning curves exhibited by telegraph operators sending and receiving Morse code; and Book (1908) studied a learning curve in typing skills. The idea that a phenomenon analogous to the learning curve might also apply at the level of the organization took longer to emerge, but it had begun to fi gure prominently in military procurement and scheduling at least a decade before Wright’s (1936) classic paper providing evidence that the cost of producing an airframe declined as cumulative output increased. Wright (1936), who was careful not to describe his empirical results as a learning curve, proposed three explanations for the relationships between cost and cumulative quantity produced that he observed. The fi rst was the “improvement in profi ciency of a workman with practice ” (p. 124), characterized by the individual learning curve. The others were “the ability to use less skilled labor as more and more tooling and standardization of procedure is introduced, ” and “the greater spread of machinery and fi xture set up time in large quantity production ” (p. 124). Only the fi rst of these is what we would unambiguously identify as a source of orga-nizational learning; the others are consistent with organizational learning but also with standard static economies of scale.

We document the evolution of productivity in a steel mini mill with fixed capital, producing an unchanged product with Leontief technology. Despite the fact that production conditions did not change dramatically, production doubles within the sample period (almost 12 years). We decompose the gains into: downtime reductions, more rounds of production per time, and more output per run. After attributing productivity gains to investment and an incentive plan, we are left with a large unexplained component. Learning by experimentation, or tweaking, seems to be behind the continual and gradual process of productivity growth. The findings suggest that capacity is not as well defined, even in batch-oriented manufacturing. JEL classification numbers: We thank Dan Aks for excellent research assistantship, and Liran Einav (the editor), three annonymous referees, Shane Geenstein, and Allan Collard-Wexler for useful comments. The financial assistance of the BSF (Grant No. 2008159) is gratefully acknowledges.

the harvard environmental economics program The Harvard Environmental Economics Program (HEEP) develops innovative answers to today’s complex environmental issues, by providing a venue to bring together faculty and graduate students from across Harvard University engaged in research, teaching, and outreach in environmental, energy, and natural resource economics and related public policy. The program sponsors research projects, convenes workshops, and supports graduate education to further understanding of critical issues in environmental, natural resource, and energy economics and policy around the world. acknowledgements

I study transactions between aircraft manufacturers and airlines as well as airlines ’ uti-lization of their fleet. Aircraft production is characterized by economies of scale via learning-by-doing, which creates a trade-off between current profit and future competitive advantage in the aircraft market. The latter consideration makes large buyers more attractive than small buyers and induces quantity discounts. The resulting nonlinear pricing strategy may distort both production and allocation in favor of large buyers. There is a negative correlation between the size of aircraft orders and the per-unit price. There is also a positive correlation between the price paid and the utilization rate of the aircraft model, which suggests that the manufacturers ’ price discrimination leads to the misallocation of aircraft. To assess whether there is an inefficient allocation, I model the market and show that price discrimination by upstream firms may lead to an inefficient outcome compared with uniform pricing. Then, I construct and estimate a dynamic model of the aircraft market that includes a model of utilization. Finally, I conduct counterfactual simulations using the estimated parameters. I find that uniform pricing increases aircraft production by 10 % and total welfare by 1.6%. ∗I am extremely thankful to Igal Hendel, David Besanko, Aviv Nevo, Rob Porter and seminar participants at Northwestern University for their valuable comments and suggestions.

Canada’s recent productivity growth has been low by historical and international standards. Canadian and international studies have suggested this may be partly due to firm-level determinants. A chance to study this hypothesis has arisen from the improved availability of firm-level data through the Canadian Centre for Data Development and Economic Research. We argue that this creates an important opportunity for researchers and describe one attempt to capitalize on it by developing a research network. DRUMMOND (2011) HAS POINTED OUT that output per hour worked in the business sector in Canada has only grown at 0.7 per cent per year since 2000, compared to about double that from 1973 to 2000 and more than five times

Computing Platform provided an excellent setting for our main numerical analysis. All opinions, findings, interpretations, and conclusions represent the views of the authors and not those of the affiliated institutions or the National Bureau of Economic Research. At least one co-author has disclosed a financial relationship of potential relevance for this research. Further information is available online at

In the literature, the study of price subsidies of new technologies has focused on the tactical problem of achieving maximum penetration under a budget constraint over a short horizon. In practice, such tactical, short-term penetration objectives seem both to adequately reect the structure of many price subsidy programs (e.g., solar energy), and account for their high failure rate. The present paper goes beyond determining optimal price and subsidy paths under a budget constraint and focuses on strategic decision making. A generic strategic objective of price subsidies is to make a product or technology competitive at an earlier point in the future than would otherwise be possible. This begs the natural question of feasibility. While determining feasibility, we derive detailed price and subsidy paths and the government budget required. Then, we can iteratively determine the program that maximizes welfare.

written by Yibei Liu