Tag: research



Dennis Zhang, Olin associate professor of supply chain and technology, has been named a co-winner of the 2022 Production and Operations Management Society (POMS) Early Career Research Accomplishments.

Dennis Zhang
Zhang

The award is one of the most prestigious honors in the operations management field. Zhang received the award in April during the POMS conference, which was virtual because of COVID concerns.

“I am honored to receive this recognition,” Zhang said. “This is very motivating, and I hope to continue contributing to the field through research and service.”

POMS chose Zhang based on his contribution to platform operations, especially retail platform operations, as well as his contribution to data-driven methodologies in operations, such as field experiments and applied machine learning.

For example, Zhang’s paper, “Reducing Discrimination with Reviews in the Sharing Economy: Evidence from Field Experiments on Airbnb,” published in Management Science, is the first study on how to use review information to fight against statistical discrimination on sharing platforms.

And “Customer Choice Models vs. Machine Learning: Finding Optimal Product Displays on Alibaba” is the first to implement a choice-model-based assortment optimization algorithm in a large-scale ecommerce setting. Operations Research published the paper.

Zhang joined the Olin Business School in 2016. His research focuses on operations in innovative marketplaces and in the public sector. He has built theoretical models to extract reliable insights from data and uses data to improve existing models. Before he joined the Olin faculty, he finished his PhD at Northwestern University and worked at Google as a machine learning software engineer.

The award co-winners are Ruomeng Cui of Emory University and Hummy Song of the University of Pennsylvania.




Small private firms that provide health insurance for their employees have better worker productivity and retention—as well as overall profitability—when compared with small firms that don’t offer health insurance, according to research by Ulya Tsolmon, assistant professor of strategy for Olin Business School.

Ulya Tsolmon
Tsolmon

The results suggest that investments in employee health and well-being provide a competitive edge to firms, especially when labor market competition for workers is high.

Firms have been shifting the costs of health care to employees, but they “might be wise to view employee health benefits as an investment that can yield significant returns,” Tsolmon and coauthor Dan Ariely, of Duke University, write in “Health Insurance Benefits as a Labor Market Friction: Evidence from a Quasi-Experiment,” in Strategic Management Journal.

“The results tell me that firms are gaining financial advantage even with their expenses toward health insurance benefits,” Tsolmon said. “The productivity results suggest that workers are ‘giving back’ to the firms by being more productive, which translates into higher profits.”

“Healthy and happy employees are innovative and productive employees.”

Ulya Tsolmon, assistant professor of strategy

The research also explored the link between high unemployment insurance benefits at the state level and more small firms providing health insurance in that state. High unemployment benefits ease employee mobility between companies, and firms respond by increasing “internal market frictions,” like offering health insurance, to keep their employees, the researchers found. That correlation didn’t apply to bonuses, pensions or training—making health insurance a unique lever among employee benefits.

The paper is the first to explain health insurance provision in small firms from the perspective of human capital management and to use empirical evidence to test its impact on firm performance, the authors say.

Data from 15,000 small firms

“Health insurance is a significant investment for small firms, so the interesting question to me was not why firms don’t offer health insurance, but rather looking at firms that do offer health insurance, asking why they do that and whether it’s a smart strategy and under what conditions,” Tsolmon said.

The research used data from the financial records of 15,000 small firms (with no more than 500 employees) in the US. The data set included accounting details on all expenses and revenues, as well as employee records, for five years. The authors looked at twelve different variables, including training costs for an employee.

Tsolmon supplemented the financial records with 761 Glassdoor reviews and 11 open-ended interviews with randomly selected small business owners, representing different industries and firm sizes. Just like with the numbers’ data, employee satisfaction was reported to be higher in firms that offered health insurance, and business owners spoke about more easily attracting and retaining employees after they began offering health insurance.

Implications for large firms

“By investing in worker well-being,” Tsolmon said, “firms can tap into their latent productivity and innovation that’s difficult to incentivize with monetary rewards alone. Healthy and happy employees are innovative and productive employees.”

The research also has implications for large firms, most of which provide health insurance but whose benefits differ in generosity.

“Given our finding that policies intended to increase employee wellness can affect turnover, productivity, and firm performance, large firms should consider increasing the employee uptake rate of health benefits by bearing a greater share of the insurance costs themselves,” the authors write.

Jill Young Miller contributed to this report.




Professor Andrew Knight teaches a hybrid course in Emerson auditorium. In front of him, socially distanced students site in the auditorium, while behind, students participating remotely appear on screen.

For nearly 20 years, Andrew Knight has been interested in unobtrusive research methods. A professor of organizational behavior, he’s passionate about learning how people can best work with one another, and his current focus is on improving people’s virtual collaborations.

Knight used the onset of the COVID-19 pandemic in 2020 as a spark for packaging something he’d been experimenting with to analyze photos and video recordings.

The result: a new, free software named zoomGroupStats.

“With teaching shifting to a virtual realm and a pressing need to understand virtual collaboration, I was motivated to accelerate the development of this software,” he said.

The package is to enable researchers to use the virtual meetings platform Zoom to collect data that illuminates how people interact with one another, Knight said. With it, users can quickly turn files downloaded from Zoom into datasets, analyze the dynamics of spoken and text conversations in virtual meetings, and extract information from the video feeds of virtual meetings.

Who do you imagine using this software? And for what?

Knight

The first category is researchers. The software is currently designed especially for researchers who study teamwork, negotiations, interpersonal relationships and group dynamics. However, the basic functionality of the library would be useful to anyone who wants to extract insights into conversation dynamics and emotion during virtual meetings.

The second category is teachers running virtual classes. The software can provide insights into who is engaged in the conversation during a class (i.e., class participation) and the ways in which people are making contributions (e.g., vocal contributions, text-based chat contributions).

The third category is leaders and managers who conduct virtual meetings. When paired with a web application that I created (http://meetingmeasures.com), the software can give leaders feedback on how effectively they facilitate virtual meetings. 

How easy are the tools to use?

When paired with the step-by-step tutorial that I created (http://zoomgroupstats.org/), I hope the library is accessible for anyone with a basic level of proficiency with the open source statistics software R. Some elements of the functionality—such as the capacity to read the emotional expressions of people’s faces through their cameras—requires an additional level of proficiency in setting up and configuring Amazon Web Services. 

What are the privacy concerns with this software?

Like any use of recorded human behavior, users must take into account privacy considerations. In a way, the privacy concerns for research are equivalent to more traditional methods (e.g., having human research assistants rate and classify people’s behavior from a video recording). However, people have variant perspectives on software-based “automatic” coding compared to human-based coding of their behavior. As a general rule, anytime a meeting is being recorded, a meeting leader should explicitly request permission to record from all meeting participants.

How have you used your software?

I’ve used this for research and teaching purposes so far. On the research side, I have primarily been working to validate a set of metrics that can be automatically derived from a virtual meeting vis-à-vis traditional, survey-based metrics. This is important to situate the automatic metrics within the current landscape of research on interpersonal relations and group dynamics.

On the teaching side, I have used this software in combination with my Meeting Measures web application to give students feedback on their virtual meetings. This is helpful for showing students when, for example, they dominate the conversation or make inadequate contributions to their team meetings. 

So the software is free?

Yes. The R package is free and open source. It is available through the web-based repository (Comprehensive R Archive Network, or CRAN) that is used to distribute packages for R. 

Top photo: Professor Andrew Knight teaches a hybrid course in September 2020 in Emerson Auditorium. In front of him, socially distanced students site in the auditorium, while behind, students participating remotely appear on screen.




The economy and coronavirus pandemic were two of the top issues for voters in the 2020 election, according to exit poll surveys. Notably, 52% of voters said controlling the pandemic was more important, even if it hurts the economy. But what if we didn’t have to choose?

In communities where masks were mandated, consumer spending increased by 5% on average, showing that a safety rule can stimulate economic growth as well, according to a new study from the Olin Business School.

Researchers found the effect was greatest among non-essential businesses, including those in the retail and entertainment industries—such as restaurants and bars—that were hit hard by the pandemic.

​Thomadsen

“The findings exceeded our expectations and show that we can have a strong economy with strong, commonsense public-health measures. Mask mandates are a win-win,” said Raphael Thomadsen, professor of marketing and study co-author.

Thomadsen, along with Olin’s Song YaoNan Zhao and Chong Bo Wang, analyzed the impact of social distancing and mask mandates on both the spread of COVID-19 and consumer spending. They used cellphone location data to track the degree of social distancing in nearly every county in the U.S. and compared that with community voting patterns, coronavirus infection rates and consumer spending rates.

The researchers found social distancing has a large impact on reducing COVID-19 spread, while the evidence on mask mandates is mixed. But while social distancing reduces consumer spending, mask mandates has the opposite effect. They also found that social distancing decreased in communities with mask mandates, magnifying the positive effect on spending.

Feeling safer to spend

Yao

“Preventive measures such as social distancing and facial masks should be considered as pro-business,” said Yao, associate professor of marketing. “When people feel safer to spend, or more importantly, when the pandemic is kept at bay, the economy is more likely to have a quick recovery. Not to mention the lives that will be saved.”

Perhaps not surprising given the political lines drawn over masks, they also observed that political affiliation had a significant impact on social distancing. Even after controlling for local characteristics such as the population density, income and other demographics, counties that voted for President Donald Trump in 2016 engaged in significantly less social distancing than counties that voted for Hillary Clinton.

“If the entire country had followed low levels of social distancing seen in Trump-supporting areas, we estimate there would have been 83,000 more American deaths from COVID to date, which represents a 36% increase over the current death count of 225,000 Americans,” Thomadsen said.

They estimate the tradeoff would have been a relatively small boost in the economy. Consumer spending dropped $605.5 billion from April to the end of July, compared with the same time last year. The country would have recovered $55.4 billion, or approximately 9%, had all counties remained as open as the most pro-Trump areas.To put it in more dramatic terms, Thomadsen said this means that opening up is only a reasonable policy if one values lost lives at roughly $670,000 each or less. This value was determined by dividing the hypothetical $55.4 billion boost to the economy by the 83,000 lives lost in this scenario.

“The calls to open up the economy come with huge costs of COVID spread and only modest benefits of increased economic activity,” Thomadsen said. “Opening the economy before getting the virus under control only makes sense if you put a very low value on life.”


A central puzzle of corporate strategy is whether headquarters can add value to their business units beyond the burden of their own overhead. The record is bleak: On average, corporations trade at a 20% discount relative to their breakup value.

“This is the problem that we want to try fix,” said Anne Marie Knott, Olin’s Robert and Barbara Frick Professor of Business.

Anne Marie Knott

She proposed and tested a theory of how corporations could overcome that record. On November 10, she presented the findings as part of the Olin Business Research Series. More than 60 people tuned in for the virtual event.

The 20% discount could mean that multibusiness firms fundamentally destroy value or that they are poorly managed. Regardless, a whopping $5 trillion economic gain could be had from a better understanding of how headquarters add value in multibusiness firms, Knott says.

Bank One and its return on assets

Bank One, a bank holding company, motivated the theory. Knott and co-author Scott Turner, of the University of South Carolina, explain how in “An Innovation Theory of Headquarters Value in Multibusiness Firms” in Organization Science.

Bank One increased the return on assets of its target banks by 40-70%.

“This would be really easy if they were purchasing underperforming banks,” Knott said. But they weren’t. They were buying well-managed banks.

The theory relies upon dynamics between business units where laggard units improve their performance by imitating leaders. In turn, this “competition from below” stimulates leaders to innovate more.

Knott polls audience members during her Business Research Series presentation.

Beyond demonstrating that headquarters can add value through innovation and growth, the theory offers prescriptions on how to do that. For instance, they can establish systems that create norms for sharing, which eases innovation. They also can offer high-powered incentives to fuel innovation.

In general, Knott’s research examines the optimal environment and policies for innovation, which she summarizes in her book, “How Innovation Really Works” (March 2017). This interest stems from issues arising during an earlier career in defense electronics at Hughes Aircraft Company.

KEY TAKEAWAYS:

  • A $5 trillion economic gain could be had from a better understanding of how headquarters add value in multibusiness firms.
  • Bank One increased the return on assets of its target banks by 40-70%.
  • The theory relies upon dynamics between business units where laggard units improve their performance by imitating leaders.
  • In turn, this “competition from below” stimulates leaders to innovate more.