A classroom full of students sat behind desktop computers in a School of Management computer lab recently making business decisions and wondering whether they could trust one another.

As students role-played the parts of manufacturers and suppliers in an online simulation,  they calculated each decision as if they were playing chess.

Taking learning beyond the classroom, Professor Özalp Özer offered these students an intriguing blend of teaching and research by simulating real-life business challenges with a game of trust, the subject of his National Science Foundation-funded research.

The School of Management operations management professor uses the classroom experiment to demonstrate to his students that not all business moves are made for corporate gain. In fact, his research suggests that human emotions play a large role in how managers make decisions internally and share private information externally, decisions that help products get to market at the right time at the lowest cost.

“Lack of credible information sharing adversely affects companies as well as consumers who suffer from high prices and lack of product availability.”

Dr. Özalp Özer,
associate professor
of operations management

Ozer, who has been working on the study for three years, recently received a$280,000 grant to continue and expand the work because the National Science Foundation believes it has potential to provide transformational solutions to major business challenges.  This is the first study, Ozer says, to incorporate trust into the economic modeling of contracts and forecast information-sharing in supply chains, a technique that could be expanded into other areas of business that require information-critical decisions.

Ozer and co-authors Karen Zheng, a PhD student at Stanford University working under Ozer’s supervision, and Dr. Kay-Yut Chen, an experimental economist at Hewlett Packard Laboratories, have created a new theory that they believe better explains what drives decisions and how people share private information.

“Lack of credible information sharing adversely affects companies as well as consumers who suffer from high prices and lack of product availability,” said Ozer. “Yet, some companies seem to avoid such failures and manage to share information credibly even when the communication is non-binding.”

Traditional economic models assume that people make decisions based solely on monetary motives. Ozer incorporates behavioral research to better explain why some companies effectively share information with other members of their supply chain and why other companies don’t.

The study, “Trust in Forecast Information Sharing,” has recently been accepted for publication in Management Science, the flagship academic journal in the field of operations management. The paper explores business relationships between manufacturers such as Hewlett Packard, which solicits forecast information from retailers like Best Buy to plan production numbers and capacity. Lack of credible information sharing can lead to such disastrous outcomes as the $2.1 billion in excess inventory Cisco faced in 2001 when its customers inflated their forecasts.

“Manufacturers like Cisco collect forecast information from their customers who many times provide overoptimistic forecasts.  But it’s Cisco who is going to end up building the product, and if there’s excess capacity due to exaggerated forecasts, it’s Cisco who is going to end up hurt,” Ozer said.

“The customers didn’t procure as much as they forecast, and Cisco had to write off billions of dollars worth of excess inventories, which turned out to be worthless.”

Dr. Özalp Özer

“In fact, in 2001, when Cisco collected forecast information from its customers, such as telecom operators, the forecasts didn’t make sense, so Cisco turned to its customers and said that there are not that many consumers on the planet to work with this stuff. Yet, Cisco’s customers insisted that Cisco secure capacity according to their forecast information. And what happened?  The customers didn’t procure as much as they forecast, and Cisco had to write off billions of dollars worth of excess inventories, which turned out to be worthless.”

To create a new theory that accurately predicts human decisions, the researchers used economic modeling and behavioral experiments to explain and predict outcomes that were not possible to predict before, Ozer says.

Ozer’s team conducted laboratory experiments, placing participants in the roles of “manufacturers” and “suppliers.” The game helped the team build a quantitative mathematical model that measures trust and trustworthiness and predicts decisions and actions.

“What we’ve observed is that people and decision-makers have a range of trust and trustworthiness. There are two extremes,” Ozer said.

“You’re either never trustworthy and never trusting – and in this case you should really sit down and write down every single detail in a contract – or you’re fully trusting and you probably don’t need any legal context to do it,” Ozer said. “What we’re finding out is that there’s a continuum between these two extremes that is influenced by variations in the supply chain environment that affects related operational decisions.”

Results from this study, says Ozer, can help change a firm’s information management and contracting strategy to prevent overbuilding, but also ensure that enough product is built to satisfy demand. In fact, some firms, such as Hewlett Packard, are already using key results from this research to develop effective forecast sharing programs with their distributors.

“Doing some of these things is very new and quite exciting to me, and at the end of the day, I will be able to say how we can quantify what people believed to be non-quantifiable. But at least for this specific context – forecast information sharing – I can look at our data and say, ‘Your trust factor is X and your trustworthiness factor is Y,’ and I can predict whether their relationship will be more cooperative or less cooperative,” Ozer said.