Why Productivity Suffers When Employees Are Allowed to Schedule Their Own Tasks

Deviating from an organization’s prescribed task schedule tends to erode productivity, even among the most experienced workers, according to new research from María R. Ibáñez, Jonathan R. Clark, Robert S. Huckman, and Bradley R. Staats.
by Carmen Nobel
Source: rawpixel

Many jobs involve completing a series of sequential, independent, prearranged tasks. Physicians see patients; teachers grade papers; insurance agents process stacks of claims.

In the interest of productivity, some organizations have a predetermined scheduling policy, requiring that tasks be completed in a particular order. But in many instances, workers have more freedom over their workday: They can follow the prescribed schedule or else choose to deviate, completing tasks in a different order at their own discretion.

It’s easier than ever for managers to grant that freedom, thanks to technological advances like internet connectivity, mobile devices, and software. Autonomy is a much-touted benefit at many organizations, and it’s nice to believe that we each know the best path to our own optimal productivity. But new research shows that deviating from an organization’s recommended schedule tends to erode productivity, even among the most experienced workers.

“We wanted to find out what happens when people deviate from the sequencing structure that their organization has set for them, and how do they make the decision to do that”

The findings are detailed in the working paper Discretionary Task Ordering: Queue Management in Radiological Services by María R. Ibáñez, a doctoral candidate at Harvard Business School; Jonathan R. Clark , an assistant professor at the University of Texas at San Antonio; Robert S. Huckman, a professor at HBS; and Bradley R. Staats, an associate professor at UNC Kenan-Flagler Business School. The study will appear in a forthcoming issue of Management Science.

The researchers set out to answer two questions: One, what drives workers to deviate from an employer’s task scheduling policy? Two, what are the performance implications of deviating from that policy?

“There is a lot of research about how the sequencing of tasks affects productivity, but there is not much that we know about how individual decision makers organize their own work,” Ibáñez explains. “We wanted to find out what happens when people deviate from the sequencing structure that their organization has set for them, and how do they make the decision to do that?”

To find out, they analyzed data from a large outsourced radiological services firm, full of doctors whose jobs involved sequentially reading and diagnosing X-rays, CT scans, MRIs, ultrasounds, and other images randomly assigned to each of them by the firm’s centralized queuing system. The analysis covered all 2,766,209 cases that the firm processed between July 2005 and December 2007. Because the radiologists did their work at computer workstations, the researchers could track when and how the images were received and processed.

Going rogue, schedule-wise

At any given time, each radiologist had an average of 5.6 images in his or her processing queue. The firm had an implicit “first-in-first-out” scheduling policy, meaning that the doctors were expected to read the images in the order they arrived. However, they had the option of choosing to read them in a different order.

The researchers sought to find out whether deviating from the prescribed schedule would slow down or speed up the time it took the radiologists to read each image. During the two-and-a-half-year sample period, the radiologists strayed from the prescribed scheduling order 42 percent of the time.

Certain factors increased the likelihood of this behavior. Experience was one: For every year of working at the firm, the likelihood of deviating from the scheduling policy increased by 18.4 percent. The human tendency to save the worst for last was another factor: When a queue included both easy and difficult cases, the radiologists were more likely to avoid the first-in-first-out policy, presumably so they could get the shorter cases out of the way before tackling the more complicated ones. “Batching” was yet another: Some radiologists chose to complete the tasks in the queue by category—chest X-rays first, brain scans next, and so on.

Importantly, there is no evidence that attempts to maximize speed affected the quality of the work. The firm kept track of instances in which customers questioned any given diagnosis; the researchers found that only 0.3 percent of the images in the sample (about 80 cases) received even a minor customer complaint. “These people are highly trained, and they think deeply about their work,” Ibáñez says. “They understand that both the quality and speed of their work are very important to their patients.”

Here’s what happened when the doctors chose not to follow the standard scheduling policy: In total, the time it took to read an image increased by 13 percent. That lost time cost the firm a lot of potential money. “Overall, our calculations suggest that foregoing deviations would have led to faster reading times that could have saved 2,494 hours per year, which, when translated to the bottom line, would have increased annual profits by 3%,” the researchers write in the paper.

When the doctors chose not to follow the standard scheduling policy, it took them a longer time to complete the task of processing an image. This was the case regardless of how experienced they were. (Source: “Discretionary Task Ordering: Queue Management in Radiological Services”)

“In the company we studied, exercising discretion to select tasks outside their prescribed sequence tends to erode [the] radiologists’ productivity,” Ibáñez says. “This productivity decline lessens as doctors learn from experience, but is large enough to suppress the learning effect from two years of experience.”

The mistake we make when we deviate

So why did productivity drop when the doctors deviated from the firm’s scheduling policy? And why did they do it anyway?

It’s not that a first-in-first-out scheduling policy is necessarily more effective than, say, a batching policy. In fact, multiple studies have demonstrated the benefits of batching tasks by category rather than taking them on in the order they arrive. In general, people are not necessarily wrong to assume that their personal process is more effective than the firm’s policy, which sometimes is not the most practical one.

Rather, it’s likely that people fail to consider the time it takes just to choose one task over another, in what they believe to be the best interest of personal productivity. Those who always follow a recommended schedule don’t have to spend time deciding when or how to stray from it.

“Searching through your queue and deciding which task to choose next may not seem like it, but it’s actually taking you a long time,” Ibáñez says.

That’s not to imply that discretion is a bad thing when scheduling tasks, she adds, but the researchers highlight the importance of weighing the harm (time) against the advantages (freedom to innovate and make improvements).

“Managers should pay attention to the effects of deviations on productivity in their settings,” the researchers write. “Although an initial task sequence assignment might not be optimal, allowing front-line workers to take an active role in scheduling might not be advisable in settings where the time required to exercise discretion exceeds the benefits of doing so.”

While the study focuses on task scheduling, it highlights a broader idea—namely, that managers must consider the time costs associated with decentralizing any type of decision-making process in their organizations.

For example, a large retail chain may empower local store managers to make decisions about pricing, product displays, or brand-related social media content. On one hand, that makes good sense because they are more informed about local conditions and customers than anyone at headquarters. On the other hand, the time it takes to make those decisions means less time for other vital tasks. “Chains should weight these costs of exercising discretion against the benefits of local customization and potential increase in customer engagement,” Ibáñez says.

Question for readers: In your experience, what are the scenarios in which it does or doesn’t make sense to decentralize decision-making? Please share your thoughts in the comment section below!

Related Reading:

Web Surfers Have a Schedule and Stick to It
How Electronic Patient Records Can Slow Doctor Productivity
Hiding from Managers Can Increase Productivity

About the Author

Carmen Nobel is the senior editor of Harvard Business School Working Knowledge.

Post A Comment

In order to be published, comments must be on-topic and civil in tone, with no name calling or personal attacks. Your comment may be edited for clarity and length.