TRAC Critique of Dartmouth Report on Nike
March 8, 1998
[With much fanfare, on October 16, 1997 Nike released its own summary of astudy done by a group of MBA students from the Tuck School of Business at Dartmouth College. The Dartmouth study was described as an investigation into whether wages paid by Nike contractors in Vietnam and Indonesia matched the workers' cost of living. The study purported to find that Nike wages not only paid for basic needs but even provided for significant discretionary income. It was no accident that Nike released its summary just two days prior to a massive international protest of the company's sweatshop practices. Nor was it any surprise, given how flawed the study proved to be, that Nike refused to release the actual report until some months later. As the following critique points out, the study was deeply flawed, in both methodology and analysis of its own data.
The following critique was written by Dara O'Rourke. An earlier report by O'Rourke, which exposed dangerously unhealthy conditions in a Nike factory in Vietnam and which included data from a secret Nike audit leaked by a disgruntled employee, made the front page of the New York Times in November.]
NOTE: Campaign for Labor Rights edited the following analysis for the sake of brevity. In the process, some paragraph order was changed and somesection headings were added. For the complete analysis, see the CorporateWatch web site at www.corpwatch.orgComments on the Vietnam Section of the Tuck School Report:
"Nike, Inc.: Survey of Vietnamese and Indonesian Domestic Expenditure Levels"by Dara O'Rourke February 17, 1998Transnational Resource and Action Center (TRAC), (415) 561-6567 (8) [email protected]
METHODOLOGY:
The report purports to examine "income and spending levels required tosustain individual needs" (p.3) and yet the researchers developed a studydesign that explicitly avoided interviewing individual workers at Nikefactories.
As the Tuck team explains without even the slightest sense of irony, "Afterinitial discussions with NIKE Vietnam management, it became evident to usthat the factory workers would not be an 'objective' population to sample"(p.10). Fearing the potential biases of workers, the Tuck team conducted a"random" survey of income and expenditures of people living near two Nikeplants, complemented with an outdated living standards survey, and thencompared this to management reported wages for the factories.
The Tuck team assumed no bias on the part of the factory management, whileassuming excessive bias on the part of workers. So while they cautionagainst the motivations of workers towards biased responses, ProfessorJoseph Massey (who supervised the project) explains "we have no reason... apriori, to assume that they [ factory managers ] would give to an academicresearch team fraudulent data. So we accepted both the information that wereceived from the factories about what the factory minimum wage, the factoryaverage wages were."
Although the report makes claims about Nike workers, it does notintentionally study the actual wages or expenditures of Nike workers. Themost relevant information the Tuck team collected was when they interviewedworkers who happened to live in the houses they randomly selected.
The calculations of discretionary income were not based on actual wages. AsProfessor Masseyexplained, the Tuck team "did not look at wage stubs. We took average wagesas reported to us by the factories." It should seem obvious that usingmanagement provided wage data without cross-checking information againstactual pay-stubs is highly problematic.
No verifiable data on discretionary income were collected. A close readingof the report shows that the most the Tuck team can actually claim is that"workers appear to be able to generate discretionary income" (p.29).
The selection of factories was biased. The Tuck team selected the factoryarea with the lowest minimum wage and highest standard of living. VS and VJare more rural than either VP (Pou Chen) or VT (Tae Kwang Vina). This meansthat average income and expenditures around the two factories selectedcompare more favorably relative to Nike wages than would the two largestNike factories in Vietnam.
Comparison of wages to annual average per capita GDP for all of Vietnam($230) is inappropriate. VS and VJ, while less industrialized than the areasaround other Nike plants, are nonetheless, hardly agricultural areas. As theTuck team notes, only 21% of adults in the VS area, and only 10% of adultsin the VJ area are farmers. These areas are in the most rapidlyindustrializing and urbanizing region of Vietnam. They have little in commonwith the 80 percent of Vietnamese who remain in rural agricultural areas,and who drive the national GDP figure. A better comparison is to the averagewage in Dong Nai province. The Dong Nai Department of Labour, Invalids andSocial Affairs (DoLISA) reported that average income per capita for Dong Naiprovince was $449 in 1995. GDP has grown by 16% per year since 1995. A roughestimate would thus put the average per capita income to be around $600 peryear in 1997.
The report fails to compare Nike wages to wages in other factories in theregion. DoLISA reported average wages to be: $90 per month for State-ownedenterprises; $60 per month for foreign firms; and $50 per month for smallprivate Vietnamese firms. Nike workers at the VJ factory reported earning anaverage of $41 per month.
The "typical expenditure" profile is flawed in a number of ways. Expendituredata is presented forhouseholds and then divided by the number of members. This is inappropriatefor analyzing expenses of single individuals living alone. As Tuck'sIndonesia team asserts, there are "five general groups of factory workers,each with differing spending and saving habits" (p.7). Workers living awayfrom home - which is the majority of workers in Nike's largest factories inVietnam - have higher expenditures than those living at home (p.8).
Major expenditures are inexplicably omitted from the calculation of workers'average expenses. For instance, the Tuck team assumed that Nike workers havezero transportation, education, or entertainment expenses (p.27). However,their own survey data show that "transportation and communicationexpenditures have risen to represent 18.6% and 16.6% of household spendingfor the VS and VJ areas respectively" (p.26) and education expenditures arearound 14% of expenses (p.24). The chart on page 13 shows thattransportation and education are the second and third highest expensecategories for households, and yet the Tuck team asserts that Nike workersspend no money on these basic needs.
A MORE APPROPRIATE STUDY METHOD
A research project designed to seriously analyze the question of workers'income and expenditures would have been conducted very differently. Putsimply, a study of income and expenditures of Nike workers should takeactual Nike workers as the primary unit of analysis.
This would involve:
- Collecting actual wage data from a large sample of Nike workers. Thisshould be based on management records, interviews with workers, and reviewof pay stubs.
- Analyzing expenditures for different types of Nike workers. As theIndonesian study notes there is a range of types of workers in these factories.
- Analyzing average wages in the region to give some perspective on how Nike jobs compare to other manufacturing employment.
- Analyzing normal living expenditures for the region. A "purchasing powerparity" methodology, while problematic in a number of regards, could be usedto compare the costs of a basic "basket" of goods and services across regions.
- Evaluating savings and discretionary spending patterns of Nike workers.This could involve surveys of different types of Nike workers.
DATA VS. CONCLUSIONS:
Nike has used the Tuck school report to support the claim that it isproviding "highly desired, good-paying jobs" in Vietnam (Nike Press Release10/16/97). However, the actual data collected by the Tuck team paints a verydifferent picture. Many workers in these Nike factories are being paid belowthe legal minimum wage. Not one worker at Nike's Sam Yang factory reportedany savings. Discretionary income is theoretical at best.
What the Tuck Team's Data Actually Shows
The most interesting finding in the Tuck team's report --- and most relevantto the question of wages and expenditures - is buried deep within theAppendices. Income data from the Tuck team's surveys show that many of theVS (Sam Yang) and VJ (Chang Shin) workers interviewed as part of their"random" surveys are being paid far below the minimum wage. Page 1 of theAppendix titled "Income" shows twelve workers at VJ reported pay below thelegal minimum wage of $35 (406,000 Vietnamese dong) per month. Eight workersreported being paid 300,000 dong per month, which is equal to $25.86 permonth. Seven of the 11 VS workers interviewed also reported being paid belowtheir local legal minimum of $45 per month. The Tuck team apparentlydid not find this violation of Vietnamese law relevant or interesting enoughto merit comment in the main report. This information directly contradictsthe statement in the report summary that "factory workers in both Indonesiaand Vietnam consistently earn wages at or above government mandated minimumwage levels."
The Tuck team uses management provided wage amounts in their calculations.However, a close analysis of their own data contradicts managementestimates. The report fails to note that the average wage of VJ workersinterviewed (for which wage data was provided) was 475,135 dong per month($40.95). This average wage is far below the management reported "mean wage"of 653,000 dong ($56.29), or the management reported "base wage" of 548,000dong ($47.24), which the Tuck team use for their hypothetical income andexpenditures analysis. The Tuck team's own research shows that managementdata on wages is simply incorrect.
Similarly, in another part of their report the Tuck team notes that "theaverage annual income of a Nike worker [ in Vietnam ] is $384" (p.9). Thisworks out to an average monthly wage of $32, which is well below the legalminimum, not to mention the factory managers' claims. It is not clear whythe Tuck team ignored this information in their calculations of income andexpenses.
The most disingenuous finding in the report is the conclusion that "factoryworkers...can generate a significant amount of discretionary income" (p.6).At the same time, interviews with workers showed that "they very rarelyaccumulate personal savings on a recurring basis either for emergencypurposes or for anticipated future expenditures such as education" (p.6).Faced with this seemingly contradictory finding, one might think tointerview workers to find out if they really do have discretionary income,and if so what it is being used for. As I noted however, the Tuck team feltit too perilous a task to face the potential biases of workers to broachthese questions.
The data the Tuck team did collect on savings tells a very interestingstory. In the Appendix titled "Savings" the Tuck team documents that only 18of the 51 Nike workers interviewed reported any savings whatsoever. In thecase of VS, not one of the workers interviewed reported savings. It is thusunclear how the Tuck team can claim in their report, which Nike repeated inits press release of October 16, 1997, that workers "can more than make endsmeet," if their own data shows how few Nike workers can generate savings.
CONCLUSIONS
My overall analysis of the Tuck team's report is that they earned a B+ forobfuscation, and an F for reasoned analysis.
Given the very politicized debate surrounding labor issues in Nike factoriesin Asia, the Tuck team should have taken greater care to design and conducta rigorous and unbiased assessment of wages and expenditures.
Instead, using recommendations from Nike, the Tuck team carried out amethodologically flawed and biased study, and then failed to properlyanalyze the data they did collect.
A careful analysis of the Tuck team's data clearly call into question Nike'sclaim that it is providing "highly desired, good-paying jobs" in Vietnam.The data also highlight that both Nike's Code of Conduct, and its statedcommitment to the legal minimum wage, are being violated.
If Nike is interested in analyzing the adequacy of the wages it pays itsworkers in Asia, it will need to do much better than the Tuck report.Locally based, independent NGOs (such as human rights, labor, or religiousorganizations) that have experience performing wage and expenditure studies,would provide a much more accurate picture of the current situation of Nikeworkers in Asia.
Nike in the News