Video by Amelia Kundhardt
They keep on coming—corporate scandals involving revelations of deplorable working conditions at overseas factories. If it’s not the Foxconn factories that Apple employs in China, then it’s Gap’s garment workers in Bangladesh.
When industrial accidents or poor working conditions occur in the United States or Europe, usually there is a cry for more regulatory enforcement—along with an ensuing debate about how much government is enough (or too much). When such incidents happen overseas, however, the blame—and responsibility for change—usually falls to the brand-name multinationals that contract out the work.
“In a sense, global supply chains are serving a regulatory function, with companies imposing an additional layer of rules and investing resources to enforce them,” says Harvard Business School Professor Michael W. Toffel, who researches health and environmental standards in the United States and overseas. “I don’t think Nike or Levi’s relishes this role, but in the context of weak government regulatory regimes, it’s these companies’ brand reputations that are on the line when problems are found in their suppliers’ factories.”
Grudgingly or not, multinational brands have taken up the mantle of responsibility, establishing codes of conduct that their suppliers are—theoretically at least—obligated to adhere to. Often, the brands also monitor their suppliers’ performance with periodic inspections by third-party auditors. Beyond that function, however, it’s unclear what the brands should be doing with the information. “There’s an ongoing debate as to whether these audits are simply snapshots to figure out the facts on the ground so multinationals can vote with their feet and choose factories with better performance or whether the audit findings are meant to stimulate improvement,” Toffel says.
in the context of weak government regulatory regimes, it’s these companies’ brand reputations that are on the line when problems are found in their suppliers’ factories
Making matters more difficult, it is oftentimes unclear to companies how to gauge the quality of an audit, or how to find factories with better performance records before spending millions of dollars in a country. On the other hand, if companies are committed to staying with their suppliers and helping them improve, how can they know whether these factories will follow through?
In a series of studies, Toffel and several colleagues have strived to answer these questions—which carry enormous weight for companies doing business overseas. Get them wrong and brands risk doing business with suppliers that treat their workers terribly—and in the process, risk damaging their own reputations. At the same time, by helping shine a light in these areas, Toffel says he hopes this research might help overseas labor conditions improve, either by rewarding the factories that treat workers better, or by helping those with problems learn how to clean up their act.
Uncovering Bad Factories
All three studies were made possible by a social auditing company providing Toffel with data on more than 40,000 inspections in 66 countries that its auditors had conducted over several years. The first study, titled Codes in Context: How States, Markets, and Civil Society Shape Adherence to Global Labor Standards and published in the journal Regulation & Governance, was authored by Toffel, UC Hastings Law Professor Jodi L. Short, and former HBS research associate Melissa Ouellet. The study began by asking what factors in a country were associated with better working conditions in its factories. Many countries, for example, sign international labor treaties promising to uphold certain workplace standards. Labor observers, however, have long wondered whether those treaties represent better working conditions on the ground, an aspiration to improve conditions, or an attempt to merely deflect international criticism about the countries’ human rights records.
In fact, the researchers found that the treaties corresponded to better factory conditions. For each international labor treaty a country signed, its factories had 12 percent fewer violations, on average. In other words, the more labor treaties a country signs, the more likely it will live up to its promises.
Another factor that had a significant effect on factory performance was the degree of freedom of the press in the country. To measure this, the researchers used a scale of 0 to 100 as scored by the nonpartisan press rights organization Reporters Without Borders. They found that for every 10-point increase in the press rights score, countries averaged nearly 10 percent fewer violations in their factories. The team attributes this relationship to factories being more concerned about having their poor conditions exposed. In other words, the more concerned they are, the more likely they have more incentives to maintain better conditions.
“Press freedom empowers journalists and labor activists to investigate and publicize poor working conditions at factories,” says Professor Short. “This type of monitoring by private civil-society actors can be particularly valuable in countries where government labor inspection agencies are weak.”
That finding in particular may be helpful not only to a company looking to choose a country for its overseas production, Toffel says, but also for NGOs and governments interested in improving conditions. “It suggests that development agencies looking to improve workplace conditions should be thinking not only about training workers at particular factories, but also [about how they] might invest in increasing press freedom as an unexplored opportunity.”
While these general factors might help a company decide in which country to invest its production, of course different factories will have different track records—so it’s important to look at individual audits. But how can a multinational brand be sure that it is getting accurate information, and that the auditors aren’t themselves biased, corrupt, or incompetent, leading them to miss or sweep violations under the rug during inspections?
“We assume that most clients want the auditors to tell them the unvarnished truth,” says Toffel. “Obtaining accurate information from these auditors is critical to enable brands to manage this risk.”
Toffel’s second study, Monitoring Global Supply Chains with Short and Georgetown University McDonough School of Business Professor (and former HBS doctoral student) Andrea R. Hugill, looks at how audit teams may be best composed to get at that truth. Published in this month’s issue of Strategic Management Journal, the researchers examined data on how many violations different teams exposed at the same factory—which, all things being equal, was an indication of a more effective audit. They found three factors that seemed to cause more violations to be found and reported.
First, there was the question of whether an auditor on the team had previously audited the same factory. Teams consisting of all new auditors unfamiliar with a factory tended to find more violations than teams that included returnee auditors. While this might be due to corruption between returnee auditors and factory managers or a desire by the returnee auditor to show improvement over time, the coauthors contend that the result is more likely due to there being more “fresh eyes” on the scene.
In other words, since individual auditors tend to be fairly consistent in the items they investigate, they’ll most likely look at the same issues upon returning to audit the same factory. Since factories, meanwhile, tend to focus improvement efforts on problems identified in the last audit, many of those items will be fixed. New auditors, meanwhile, would be more likely to look at different issues, uncovering new violations. “If there are a thousand things to look for at an inspection, one auditor might have time to look at only 200,” Toffel says. “A new auditor might look at some of the same things, but will also focus on different things.”
Second, perhaps predictably, was the experience of the auditors on the job—with those who had a longer tenure tending to find and report more violations. Surprisingly, however, level of education had no effect on the findings, implying that learning in school is no substitute for on-the-job experience.
Even more surprising was the third factor that had bearing on the results: gender. The researchers found that auditing teams with at least one woman reported, on average, 6.6 percent more violations than male-only teams.
“None of the auditors we spoke with, including the auditor whose data we used, said they considered gender composition in forming teams,” says Toffel. “In fact, some said they expressly didn’t consider gender, believing it inappropriate to do so.”
Not considering gender when forming audit teams makes sense in terms of fairness to employees, but as the data show, this policy might be hurting the overall accuracy of audits. The authors note several reasons why this might be the case: Sociology studies have found that women tend to be better at both perceiving and integrating facts they observe, lending them a greater ability to notice violations. Women also tend to be rule-followers more than men do.
“That might make them more likely to cite violations they perceive by the book, and less likely to give the benefit of the doubt,” says Short.
Another reason, however, might result from the fact that workers in many overseas factories are mainly women, who might be more willing to divulge factory violations to auditors of the same gender—thinking they may be more likely to be believed and less likely to be punished. Significantly, mixed-gender teams did no better than all-female teams in uncovering violations, meaning perhaps the male employees were not especially forthcoming to male auditors.
Building Capacity for Change
Both studies—considering factors in countries where factories are located and at the makeup of the teams that audit them—may be helpful in arriving at an accurate depiction of factory performance with the most efficient use of time and resources. But what if a multinational brand actually wants to help its supplier factories improve? Toffel’s third study, also written with Short and Hugill, looks at exactly this question, examining what factors make an audit team more effective at changing factories for the better over time.
In this study, entitled Beyond Symbolic Responses to Private Politics: Examining Labor Standards Improvement in Global Supply Chains, the researchers compared the results of two consecutive audits, evaluating what factors in the first audit might lead to fewer violations at the next audit. They looked not only at the composition of the audit team, but also at how the audit was conducted, as well as external factors in the country that might influence company response. They found a number of factors that led to improvements—and moreover, that some of their effects were multiplied when several factors operated simultaneously.
The first set had to do with the audit team’s training—as measured by the number of audit skill courses they’d taken—and whether the audit was announced or a surprise. They found that the biggest improvements came when a highly-trained team performed an announced audit. While an unannounced audit might find more violations initially because the factory would have less time to sweep things under the rug, Toffel cautions that “an unannounced audit might also create a more contentious environment where factory management might not be so receptive to hearing any best practices the auditors might share.”
In other words, while an audit team is there to police violations, there’s often another no less important task they are performing: informally counseling the factory on how to improve conditions based on their own experience in the audit field. “It seems to be a knowledge transfer story,” says Toffel, “and for knowledge to be transferred effectively, you have to have a knowledgeable auditor, but you also have to have a receptive factory manager.”
The other factors that could impact change, however, had less to do with receptivity and more to do with fear, based on and the level of press freedom in the country where the factory was located. Factories in countries with greater press freedom were substantially likelier to improve. “Factories in countries with more press freedom might perceive more pressure to act on audit reports that reveal poor performance, given the risk of that news leaking outside the factory,” says Toffel.
This all indicates that improving working conditions through social monitoring is a learning process that depends on both the quality and the quantity of the knowledge transmitted and the factories’ willingness to learn. Well-trained auditors who spend significant time on-site can maximize the value of the information conferred. But, as Short says, “no matter how good the information is, it doesn’t help if factories don’t want to learn.” Factories may have different motivations for learning. Structuring audits to transfer knowledge in less confrontational ways may make factories more receptive to learning because, according to Short, “they feel trusted, like they’ve been brought into the process.” Other factories may be more motivated to learn when they fear that their poor practices will be exposed. “It’s like when you are studying for an exam you might fail—that fear helps you study harder,” Toffel says.
However that learning occurs, Toffel and his colleagues see hope in the potential for foreign factories to change and get better over time. If multinational brands want to play an active role in that learning process, then they have their own learning process to undergo first. Looking at the factors behind what makes for a responsible factory, an effective auditor, and a healthy chance for improvements over time may better help them effect that change—and avoid future negative headlines.