(istockphoto) (istockphoto)

In most companies, managers keep a close eye on employees to prevent them from doing something wrong. But what if monitoring employees actually keeps them from doing something right?

That's what Harvard Business School professor Ethan Bernstein found in recent research that won the 2013 Best Published Paper Award from two Academy of Management divisions and that was recently highlighted on the business school's knowledge site. The paper, which Bernstein titled "The Transparency Paradox," found that productivity actually increased when a group of Chinese manufacturing employees were not being closely watched by their managers.

Bernstein found that when managers were looking, employees in the factory, a global contract manufacturer that produced mobile devices, did everything by the book so as not to call attention to themselves. But when managers weren't watching, employees used a variety of easier and even safer tricks of their own to keep production humming at an even faster pace. In one study, simply hanging a curtain so that managers couldn't see workers increased productivity by 10 to 15 percent.

Bernstein conducted the study by embedding Chinese-born Harvard undergraduates, who were trained in qualitative research, on the factory floor. For 11 days, they lived in the factory dorms and worked alongside Chinese workers on production lines, where their new colleagues shared the innovative tricks they had developed to work faster and, in many ways, smarter than the prescribed methods. As a worker told one of the embedded students: "Even if we had the time to explain, and they had the time to listen, it wouldn't be as efficient as just solving the problem now and then discussing it later. Because there is so much variation, we need to fix first, explain later." It quickly became apparent to Bernstein and his undergraduate researchers that the work slowed when the company-mandated tactics were used.

Then, in a follow-up quantitative study done over a period five months, Bernstein hung curtains surrounding several of the production lines, leaving some workers thinking their work was going unobserved. He carefully measured production rates, examining the "defect-free units per hour" created by both the exposed and the hidden group. After the first week, performance increased by 10 to 15 percent for those working behind the curtain and remained higher than the other workers' over the rest of the five-month period. Performance increased not because the workers were cutting corners out of sight of their bosses, Bernstein writes, but because they were sharing ideas, increasing experimentation and working to improve defects.

There are obvious limits to the research--to start, other companies could have organizational cultures that better reward or create incentives for new ideas from the front line. (It should be noted, however, that the factory's managers "believed they supported an organizational culture that encouraged speaking up," Bernstein noted.) It's also worth taking into account that the work was done by unskilled workers in China, where similarity is culturally valued over individuality. And  it's unclear whether Bernstein's findings would translate to a professional workplace, where workers may not always be visually observed by their managers but are monitored digitally in almost every form. Still, he writes, there are "reasons to believe that these results are more universally applicable to organizations in other contexts."

At a time when companies like Yahoo are bringing once-remote workers back into the office, and when "big data" is providing H.R. departments with reams of information about everything employees do, Bernstein's research at least begs the question of how much visibility is too much. When workers have enough autonomy to experiment, fail and share ideas outside the watchful eye of their managers, they could very well develop and perfect tools that make them more productive, not less.

"This paper does not challenge the value of transparency," Bernstein writes. Instead, he says, it challenges what--and how much--managers should see. "Management by walking around may sometimes be inferior to management by standing still."

Jena McGregor is a columnist for On Leadership.

Read also:

Yahoo's perplexing work-from-home ban

The red sneaker effect

Like On Leadership? Follow us on Facebook and Twitter.