In the US and abroad, firms operating in the same industries and same narrow markets, managing the same number of workers and the same amount of capital, can vary wildly in their output. Economists have long wondered about different factors that account for these large disparities in efficiency and productivity.
Chad Syverson, J. Baum Harris Professor of Economics at University of Chicago Booth School of Business spoke about the source of these disparities in an April 28, 2017 Friedman Forum lecture for undergraduates. Syverson reviewed what the existing research tells economists about the reasons for these differences and outlined the major questions that still remain toward understanding the source in the variation in efficiency across firms.
Over the past 25 years, economists, aided by advances in information technology, gained access to a massive infusion of detailed data comparing firms’ production activities. Much of this data has been gathered by the U.S. Economic Census; third-party research endeavors, or firms’ own data shared via agreement. This data has allowed comparisons of firms’ performance and productivity, defined as how much output a producer obtains from each unit of inputs.
Research shows that if you rank all firms in an industry on their productivity and toss out the best and worst 10 percent as outliers, you find that typical 90-10 percentile total factor productivity ratio within 4-digit industries in U.S. manufacturing is 2-to-1 or higher. That is, the 90th percentile producer obtains twice as much output from the same measured inputs (capital, labor, energy and materials) as the 10th percentile producer.
“Nothing is special about this large dispersion of productivity and performance in U.S. manufacturing,” Syverson explained, noting that it’s even greater elsewhere. “If you look at China, the ratio is 3-to-1 and in India, it is 5-to-1.”
This productivity gap is persistent; the high-productivity businesses in a given year are likely to also be more efficient the following year. Conversely, low-productivity businesses are likely to stay that way unless they fail and shut down instead, which is frequent, according to Syverson.
The Impact of Management
Syverson grouped factors that determine productivity into two broad sets – levers, or things within businesses’ control, and external factors, or aspects of the operating environment within which a business functions.
Levers can include managerial practices and talent acquisition; better information management systems and research and development; product innovation; higher-quality labor and capital; learning-by-doing; and firm structure decisions. Comparing management decisions within firms has been difficult until recently. Today the World Management Survey collects broad and consistent data on management practices within and across firms through detailed surveys and discussions with plant managers around the world.”
“You can think of a management team like a conductor of an orchestra,” Syverson explained. “Even though a good conductor can make an orchestra sound better, a bad conductor can make it sound poorly regardless of having the same set of instruments.”
The findings of such studies indicated that multinational firms are better managed everywhere, and that management is closely correlated with productivity.
Syverson illustrated the link between management and productivity with an example of an experiment run at 28 cotton fabric plants in Mumbai. Treatment group plans received 5 months of management consulting intervention on specific practices tied to factory operations, quality, and inventory management. Control firms, who received only computer support necessary to collect comparative data. Firms receiving management help increased productivity by 17 percent, worth an extra $300,000 in profits.
However, Syverson noted, “It turns out that companies have no idea how well or poorly managed they are. If you plot survey responses of companies scoring themselves on their management practices versus the actual productivity levels, there is essentially no relationship.”
Learning, Competition, and Regulation
Syverson has also studied how on-the-job experience, or learning-by-doing (LBD) improves production efficiency. By analyzing detailed data on the hundreds of intricate processes in an auto assembly plant and tracking defects in the first production year of a new model, economists showed that defects per car started at a high level, then dropped off rapidly as line workers learned the new processes. When a second shift started producing the new model, their initial defect level was lower than the first shift, because workers learned from their colleagues’ experience.
Competition can drive an inefficient firm out of business, so it’s a logical contender as a driver of productivity. Syverson has studied its impact on productivity among concrete producers. In this industry the product is identical, but heavy, with a short life. The differentiating factor is how quickly and cheaply firms can deliver it. Syverson has studied how density of concrete providers and greater competition in a region affects productivity.
“In denser markets, where transportation costs associated with concrete are low, systematically fewer inefficient businesses exist,” he explained. “In less denser markets, there are systematically more inefficient companies.”
Similarly, Syverson illustrated how new foreign competition boosted efficiency in the US iron ore industry. “While the U.S. iron ore production was stuck at two tons per worker for nearly 30 years, it almost doubled after Brazil entered the market,” he said.
Regulation is another external factor that can affect productivity. The U.S. Sugar Act, implemented from 1934 until 1974 as part of the New Deal, was an effort to enforce quotas for both domestic and foreign sugar refining. Syverson illustrated how this government regulation adversely affected productivity and how such regulations can become a burden on consumers as companies collude in response. During the Sugar Act’s implementation, a significant drop in productivity can be observed as a direct consequence of the regulation, Syverson explained. The trend reversed when the act was repealed.
While a lot of recent research has been carried out on such productivity disparities, much still remains unknown, according to Syverson. Some unanswered questions include how economists can predict innovation, how supply and demand affect productivity, how badly resources are allocated, what types of markets make management more important and what role policies can play in encouraging productivity growth.
—by Diana Petrova