In “The Ministry of Fear,” a novel set in World War II England, author Graham Greene wrote, “It is impossible to go through life without trust: That is to be imprisoned in the worst cell of all, oneself.” If that perspective is true in times of war, it is generally true at most times in the world of business. Should people who have done little or no business together trust one another? That is a question that often comes up in concrete ways, especially as business increasingly becomes global and migrates to the Internet.

In e-commerce, clearly, building trust is a critical factor on business-to-business (B2B) websites. Consider, for example, Covisint, the online marketplace that five giant automakers—General Motors, Ford Motor, DaimlerChrysler, Nissan and Renault—have agreed to create to streamline the purchase of $300 billion in parts from suppliers. The Big Three U.S. automakers have been aggressive competitors for much longer than they have been partners. How far should they trust each other’s intentions?

Or consider another scenario: If an auto company gets a low quote from a supplier it has never dealt with before, should the car company’s executives depend upon that supplier to deliver on its obligations? Or would it be more prudent to withhold trust?

Similar issues can come up even on consumer-to-consumer (C2C) websites such as eBay, which conducts online auctions. Such auctions are becoming very popular and are completely changing the ways buyers and sellers interact. Traditionally, buyers and sellers either meet face-to-face or engage in phone conservations. Now, the Internet has eliminated the need for these meetings. This decrease in personal contacts increases the need for alternative trust-building mechanisms. For example, if you are selling a product on the site, should you trust someone to pay, if you have never done business before with him? Or if you happen to be buying a product, should you mail your check before you get a chance to see what your money has bought? What if it turns out to be a shoddy product?

 Teck-Hua Ho of Wharton’s marketing department and Keith Weigelt of the school’s management department tackle these issues in a paper titled, “Trust Building Among Strangers: An Experimental Investigation of Mutual Trust and Trustbuilding.” Arguing that the trust building process is fundamental to social science, Ho and Weigelt investigated the process with an experiment based on a novel, multi-stage trust game.

The four-stage game involved 386 subjects from universities in the U.S., Singapore and the People’s Republic of China. It was set up to explore if strangers involved in a one-time anonymous interaction could generate trust and trustworthy behavior. To determine whether they could, the game was set up in a way that the players could achieve social gains, called “property rights,” if they trusted one another.

At each stage in the game, players had an opportunity to appropriate these gains, or to be trustworthy by sharing them. The players remained strangers because they did not know each other’s identities. In other words, the game attempted to study trust-building in situations where there was little scope for social relations and networks.

Ho and Weigelt found that the experimental game yielded interesting results. They showed that a significant percentage of students initiated trust building by giving up property rights. About 70% of the subjects gave up property rights in stage one in all three countries. The results also showed a considerable absence of trustworthiness in the initial stages, though subjects were found to be more trustworthy in the fourth stage.

Unlike other studies, the work of Ho and Weigelt correctly distinguishes between two components of the trust-process: trusting behavior and trustworthiness. As the experiment showed, there is divergence in behavior. In the initial stages of the game there was significantly more trusting behavior than trustworthiness.

Ho and Weigelt’s findings have important implications. Despite the complete absence of social relations among participants in the experiment, trusting and some degree of trustworthiness were observed in the experiment. The authors attribute this to the promise of social gain in the future. Subjects at stage four exhibited a greater level of trustworthiness because they were certain about the trusting intention of others. This suggests that subjects develop a generalized morality that compels them to be nice to those who they are sure were nice to them. Moral sentiments may make emergence of such cooperative behavior among strangers possible and be the driving force behind developing trusting behavior.

These conclusions, according to Ho and Weigelt, have important implications for the business world. “Our big thesis is that when you demonstrate good intentions, it increases the level of trust,” says Ho. Adds Weigelt: “As long as you demonstrate your intentions, people are willing to trust you. This is particularly important in the case of people you have not dealt with before.”

How would these conclusions apply to, say, transactions on a B2B website? Simply put, a new dealer trying to get business from an auto company would increase its trustworthiness if it clearly demonstrated its intentions by, for instance, being willing to supply some parts on spec. In the case of C2C websites such as eBay, a potential buyer could demonstrate his or her intention to purchase by sending a payment check even before the purchased good is received. “The key takeaway for managers is that you have to make your intentions clear when dealing with strangers,” says Ho. “The more you do that, the more you will generate trust.”