Protesters demand the impeachment of Dilma Rousseff, then the president of Brazil, during a March 2016 rally in Brasilia. A large inflatable doll depicts former president Luiz Inácio Lula da Silva, her mentor, in prison garb. Rousseff was ousted later that year amid a bribery scandal. (Eraldo Peres/AP)

In January 2018, an appeals court upheld a 12-year sentence for former Brazilian president Luiz Inácio Lula da Silva on corruption charges. The sentence arose from the “Operation Car Wash” investigation, which may be the world’s biggest corruption case. Executives from Brazil’s state oil company accepted more than $5 billion in bribes in return for construction contracts, funneling some of the illicit gains back to politicians and political parties.

This investigation involved multinational corporations (MNCs) from 11 countries, including Mexico’s Pemex and Singapore’s Keppel O&M — and the construction contracts of 12 stadiums built for the World Cup and Olympics. The case illustrates how multinational firms around the world take advantage of opportunities to bribe their way to procurement contracts, investment licenses and other lucrative benefits.

Could an international convention help cut back on bribery? Such corruption raises the costs of business for all multinationals and ultimately diminishes the actual services delivered, as business opportunities don’t go to highest-quality providers. Clearly, it would be better if all multinational corporations could credibly commit to stop bribing.

Here’s the problem: While such a commitment might improve welfare the world over, it raises the incentives for a few bad apples to reap the rewards while their competitors are sidelined. Keppel O&M, for instance, allegedly made $360 million in profits in contracts it won with $9.8 million in bribes.

The OECD-ABC holds perpetrators of corruption accountable

Signed in 1997 and in force since 1999, the Organization for Economic Cooperation and Development’s Anti-Bribery Convention (OECD-ABC) is an ambitious international effort to motivate the major investor countries to police the behavior of their firms abroad. Constructing such an organization is challenging when countries are wary of surrendering their sovereignty over law enforcement. Yet OECD-ABC coordinates the activities of individual countries in passing laws against bribery and “names and shames” countries for not enforcing these laws.

Using the concept of extraterritoriality, signatory countries passed laws banning bribery by their companies both at home and aboard, and the OECD-ABC monitored the enforcement of these laws. We argue that the agreement gained its real teeth in 2009 when it entered Phase 3 of its rollout by instituting a peer review mechanism that named and shamed countries for failing to enforce their anti-bribery laws. The first two phases covered the passage of anti-corruption laws and checks to make sure the laws were correctly implemented.

But can a bad report card really rein in powerful multinational firms? Our published research analyzes how this convention reduced bribery behavior in a single country, Vietnam, which has considerable investments from both signatory and non-signatory countries. Vietnam’s largest sources of private foreign investment come from countries that signed the anti-bribery convention (Japan and South Korea) and non-signatory countries (China, Singapore and Taiwan).

Using a survey of firm managers at the time their companies entered Vietnam, we can calculate the bribery behavior of firms over time and harness this data to examine the effectiveness of the OECD-ABC. Our strategy also allows us to pinpoint when firms changed their bribery behavior.

How do you measure bribery if managers have incentives to hide it?

It is impossible to know whether firms answer survey questions accurately, but one piece of evidence suggests that bribery is a sensitive topic. When we asked managers a direct question about their perception of bribery in Vietnam, one-third of managers refused to answer the question.

In our study, we avoided this sensitivity by using a technique known as a list question. To measure bribery, we surveyed firm managers from OECD-ABC signatories and non-signatories in Vietnam and asked how many of the following activities they engaged in when registering their business. Managers noted the total number of activities but did not identify individual activities.

We gave the “control group” a list of only nonsensitive items and asked managers to indicate the total number of activities engaged in during registration and procurement, but not to indicate their participation in any particular activity. We gave the same list to the other half of our sample, our “treatment group,” but included one additional item related to bribery, asking whether the firm paid an informal charge to expedite procedures. In the list-question methodology, the difference in mean activities between the treatment and control group provides the share of firms admitting to paying bribes.

We worked with the U.S. Agency for International Development and the Vietnamese Chamber of Commerce and Industry to include the questions in the Provincial Competitiveness Index Survey, Vietnam’s largest annual survey of domestic and foreign investors. This allowed us to collect thousands of responses from firms, without directly asking firms to admit to bribery.

Did the OECD-ABC reduce bribery? 

We compared the responses of those given the control question to those given the treatment question. We found that between 2005 and 2014, 25 percent of foreign firms engaged in bribery to obtain their investment license.

What does this say about the OECD-ABC? Before the enforcement phase of the OECD-ABC started at the end of 2008, firms from signatory countries were actually more likely to bribe than their non-signatory peers (30 percent vs. 23 percent). However, after Phase 3 began, the trend completely reverses. Firms from OECD-ABC signatory countries reduced their bribery by 8 percentage points. The bad news is that firms from non-signatories increased their corruption activity to nearly 27 percent over the same period.

We didn’t expect to see this spike among non-signatories and think it could reflect a number of factors, such as a general increase in bribery in Vietnam or individual bureaucrats in Vietnam learning that they should squeeze non-OECD-ABC firms more to make up for lost income. There is also a disappointing possibility that OECD-ABC is increasing bribery by driving out investors from signatory countries and, thus, making some sectors less competitive. 

What did we learn here? The simple answer is that OECD-ABC was effective in its goal of reducing bribery by convention signatories. But the overall effect on global bribery remains an open question.

Nathan M. Jensen is a professor at University of Texas at Austin, and Edmund Malesky is a professor at Duke University. They are co-authors of the forthcoming book Incentives to Pander: How Politicians Use Corporate Welfare for Political Gain.