Stanford economist Nicholas Bloom’s study of 30,000 firms identifies the practices common to well-managed operations
Digital tools allow business leaders to oversee their organizations with unparalleled speed and efficiency. And, increasingly, new technologies are influencing how the tasks of management are themselves performed. To better understand what makes management most effective in this environment, Nicholas Bloom, William Eberle Professor of Economics at Stanford University, conducted an extensive study of 30,000 U.S. factories, and found that two practices, underpinned by innovative software and IT systems, stand out in highly effectively managed operations: monitoring and incentives.
Although monitoring employees and offering incentives correlate with higher productivity in factories, their effectiveness has implications for many other industries. Bloom’s other research — exploring the challenges and impacts of employees working from home at a large Chinese company — found that the same practices played a central managerial role and led to a significant rise in productivity.
MIT Sloan Management Review spoke with Bloom about how sophisticated new technologies are reshaping managers’ ability to monitor workers’ performance and to better align incentives to performance, thus improving firms’ profitability, sustainability and growth.
Your research surveys management practices in 30,000 firms across the United States. What were the main findings?
The survey was called the Management and Organizational Practices Survey (MOPS). It’s the first large-scale survey of management in America and was run by the U.S. Census, looking at 30,000 factories across the country. Manufacturing tends to be at the forefront of management technology because it’s super competitive, so it’s often the seed bed of new ideas. What’s new about our work is the application of big data to management: We’re looking at tens of thousands of firms and trying to find reliable, stylized facts.
The survey provided data on organization’s management practices, which we define as an organization’s use of monitoring, targets, and incentives. This definition builds on the concept of Lean manufacturing that developed in the ‘80s and the associated principles of continuous oversight, evaluation, and improvement. We worked closely with the Census Bureau; the survey questions drew on research I had previously conducted with a colleague, and we carried out the first analysis of the findings, with one wave of responses in 2010, and another wave just completing for 2015.
We found that firms that scored highly on management per our scale also scored well on productivity. In fact, our research shows that the more structured management practices — using performance monitoring, targets, and incentives for employees — correlate strongly with better performance across all sorts of important areas, such as productivity, profitability, innovation and growth. Structured management depends in part on effective use of technology.
What role is technology playing?
There are really two core elements of what we define as good management. One is monitoring, the idea that you measure everything that happens in the factory, and when you find a defect, you act on that. It’s very data intensive. And the other part is about incentives. This is more classic, the idea that you promote and reward good employees and you deal with underperformers: You retrain them, move them, or eventually kick them out.
Technology allows you to measure more, which increases the value of focusing on monitoring, and also makes it easier to introduce strong performance incentives. The old-school way was to manage by walking around. It was hard to know what was going on in your shop or hospital or factory or even school without getting out there physically, because paper-based systems took too long.
Today not only can you oversee a workplace from your desktop, you can do so far more effectively than you can in person, particularly with respect to spotting deviations. When you have data, you can act on it, and you can develop feedback systems. We’ve done work in retail, schools, and hospitals, and IT is playing a bigger and bigger role in management across all these sectors. Big data enables managers to assess averages, highlight outliers, and use that information to improve performance and prevent defects or mistakes. As an analogous benefit, better data also helps remove divisive issues from the management equation: Factors like age or seniority, workplace politics, or family connections should no longer matter as much to an employee’s promotion if this is based on hard data.
These digital tools are relevant not just in factories, but also in office work and, increasingly, knowledge work settings. People talk a lot about the gig economy; for example, platforms that capture regular images of a worker’s computer, allowing managers to ensure they are getting value from freelancers being paid by the hour. Companies are innovating in how they conduct their employee reviews, combining quantitative with qualitative assessments and starting to use software and data analytics to measure and assess performance or deploying digital tools that continuously monitor employees’ progress towards their goals. Google is famous for measuring everything and is very meritocratic, offering rapid promotions for high flyers.
And in spite of the micro-level of employee monitoring we can now engage in, we find that people generally like working for firms that recognize effort and performance. This is something I’ve looked at in other research on work-life balance and management practices. Most people prefer to be somewhere where the boss seems to appreciate hard work and promotes based on performance and ability rather than office politics. So often employees are typically happier in these well-managed environments. There are concerns over it, but, on net it seems to be a better way to operate.
You have also done research on how digital monitoring and incentive systems make it easier for people to work from home.
The practice of working from home is rising quite rapidly — from a low base but at a high rate. In the United States, the proportion of employees who work from home has tripled in 30 years, from 0.75% in 1980 to 2.4% in 2010, and it spans a wide range both of jobs and incomes. It’s enabled by technology because people who work from home generally use a computer and use broadband, but it’s also about monitoring. Employers can effectively follow what remote workers are up to.
My research looked at Ctrip, a Chinese company that allowed employees to work remotely, as an experiment over a nine-month period. Ctrip is similar to Expedia in China. It’s Nasdaq-quoted, a massive company with about 15,000 employees. The workers in the experiment were answering telephone calls — a boring, tedious job, and frankly, if no one were overseeing what the employees were doing they wouldn’t be very likely to work hard. Workers involved in the experiment engaged in four types of activities: taking orders from customers, placing orders with airlines and hotels, correcting orders, and a night-shift that both placed orders and corrected any mistakes that had been made. About half of the employees’ pay was based on the number of calls that they completed; it was easy to measure and the pay system did it automatically. There were managers in the office, and if they were worried about quality, they could look at the quality metrics composed of conversion rates (phone calls with customers that resulted in orders) and random checks — 1% of calls were evaluated by an external team. So they could have confidence that things were working well.
Since managers could monitor and measure remote employees’ performance with precision — how many calls they were answering, and the quality of their activities — the company could allow people to operate at home, even in their pajamas if they liked. And we found huge improvements in performance, profitability, and productivity. During the experiment, the performance of employees working from home improved by 13%, and in the long run, after Ctrip offered the option to work from home to the entire firm, worker performance rose by 22%. Letting employees work from home was very profitable for Ctrip, and of course, the employees loved the flexibility.
Going back to your research on management practices at American factories, you found that structured management practices were dispersed across different parts of the United States. Can you tell me more about the link between location and management?
We compared all the states that had more than 250 firms from the survey, and when we weighted the scores to come up with a figure for each state, a positive correlation emerged between the level of structure in management and location. Kansas-based organizations came in on top, scoring 0.631 out of a possible max of 1.0, followed by those in Tennessee, South Carolina, Iowa, and North Carolina. In fact, the Midwest and South tended to do pretty well in general.
Our high-performing management states are the ones where manufacturing is booming, where foreign plants are going in and growth is taking place.
Were you surprised by these results?
Initially, yes. Kansas, Alabama, and Mississippi are not states you think of as being at the forefront of modern industrial activity. You typically think of Pittsburgh, parts of California (which in 2012 had the most factories of any U.S. state) or maybe Detroit as leaders in American manufacturing; but Pennsylvania, California, and Michigan placed 32, 30, and 31 in the survey respectively. In hindsight, though, the explanation becomes clearer. The South is dominated by so-called right-to-work states. They tend to be pro-business. They’re light on regulation. They’re anti-union. A state’s policies tend to correlate with higher-scoring management practices. You basically leave firms to operate the way they want to without any interference, and that tends to improve their management practices. They run the firms the way they want to that maximizes profits rather than the way a government or union official tells them to.
But right-to-work legislation — which Michigan itself introduced in 2012 — has been highly controversial. It may enable more structured management practices, but don’t you think there are implications for employees, who end up with weaker collecting bargaining rights and less job security?
It depends. If efficiency improves, then job prospects get better. I would rather be an automotive worker in the South, where firms are expanding, than in the Northeast and Midwest, where they are laying off workers. So in the long run, having a well-managed firm protects jobs — and that is good for employees. But of course, in the short run, employment legislation can often protect jobs by stalling layoffs. So it’s a classic short-run versus long-run trade-off. Short run it’s more painful, but in the long run it’s a better outcome.
Should we expect to see more and more tech-enabled monitoring and incentivizing taking hold in management practice?
Yes, we already see that very clearly. Evolution works very well in the corporate world, and organizations are continuously evolving and improving — those that aren’t tend to shrink and die off. Our research finds that structured management is a strong predictor for growth and survival. We continue to collect data and management practices are steadily getting better. Between 2005 and 2010 in our survey, data-driven performance monitoring rose significantly.
So these technologies, and the management practices they facilitate, are spreading out across industries. Twenty years ago, Lean manufacturing was just starting to move from manufacturing to retail and healthcare. Twenty years from now, I assume it will be completely standard — organizations will routinely have performance displays, meritocratic promotions, and motivating targets, no matter what sector or business a company is in.