Productivity dispersion is so large across manufacturers that productivity, the amount of output a firm obtains from a given number of inputs, in the 90th percentile of production within a particular US industry is twice as high as it is in the 10th percentile in the same industry, according to professor of economics Chad Syverson.
“The 90th percentile producer is going to get twice as much output, twice as much revenue, [and] twice as much value in production from the same inputs as the 10th percentile producer of the same product during the same time period," said Syverson during a presentation entitled “What Determines Firm Performance?". Syverson spoke at Gleacher Center on January 27, 2012, as part of the Steingraber/A. T. Kearney Speaker Series.
In China, the disparity in productivity is about 5-to-1 and in India, about 7-to-1, he said. “Furthermore, productivity measurements are persistent," Syverson said. “In other words, a high-productivity business this year is likely to be a high-productivity business next year. For example, in the US, a third of the top 20 percent [is] still in the top 20 percent five years later. If you’re bad, you’re likely to be bad until you exit.
"Two broad sets of factors—external issues associated with operating environments and those that businesses, in concept at least, can control—determine productivity, Syverson said. The six sets of levers for internal factors he cited include:
- Managerial practice and talent
- Higher quality of labor and capital
- Information technology and research and development
- Learning by doing
- Product innovation
- Firm structure decisions
A recent experimental study of management practices in 20 textile plants in India showed that firms that received consulting on 38 management practices for five months were able to reduce manufacturing problems by 40 percent within a few months, he said. In contrast, firms in the study that received only observation for one month showed no change in manufacturing problems, Syverson said.
Altogether, the average plant that received management training was able to save between $200,000 and $300,000 a year, he said. “This effect really does seem to matter," Syverson said of the study’s results. “The gold standard of a randomized trial does seem to indicate that management practice directly affects performance."
Outside of management, external environmental factors that determine productivity include productivity spillovers, competition, regulation, and input market flexibility, Syverson said. He cited the US Sugar Act, which lasted from 1934 to 1974, as an example of how regulation can provide an incentive that leads to a decrease in productivity.
The Sugar Act rewarded farmers who grew sugar beets by the amount of sugar each beet contained, and paid the subsidies with a tax on sugar company production, Syverson said. In exchange, the government let sugar companies “collude on prices."
Encouraged by the subsidy, farmers grew the largest possible sugar beets, which were then, in fact, very difficult to refine into sugar, Syverson said. The number of pounds of sugar extracted per ton of beets had grown from 220 pounds in 1902, to 315 pounds in 1934. It then steadily dropped until 1974, when the act expired and it began to climb again, he said.
“It’s a really amazing picture of how the incentives of a regulatory structure can create incentives to not perform very well," Syverson said. “There are good reasons for certain regulations, but you have to think hard about the incentive effects those regulations might have."
Syverson’s data could be used by firms to make and justify strategic decisions, Evening MBA Program student Gustavo Millan said. “Things other than good management account for higher productivity," Millan said. “Production gains sourced from intangible capital [are] oftentimes mismeasured or even unobserved."
Photo by Beth Rooney