Funding terror.

• Author: Baradaran, Shima
• Position: II. Experimental Results through III. The Future of the War on Financial Terror, with appendix and footnotes, p. 508-536

2. Experimental Results

   While the government has done much to disrupt and dismantle terrorist networks worldwide, (164) security officials and commentators are concerned that the United States has not invested sufficient resources to cut off the true terrorist lifeline: illicit financing, (165) As the United States' military and intelligence efforts prove increasingly effective at dismantling terrorist networks, terrorists must seek ever more clandestine approaches to finance their activities. As described above, this has led to terrorist organizations laundering money through trusts, charities, and shell corporations. (166)

   The international community has responded by developing policies to combat money laundering and terrorism financing. (167) For example, intergovernmental organizations (IGOs) and nations such as the United States have enacted extensive regulations that track FATF recommendations. (168) These efforts produced strict regulations governing the formation of shell companies. At least on paper, the "anonymous" shell corporation was completely prohibited. (169)

   While these international and domestic laws have been in place for years, few empirical studies have investigated their effectiveness. Previous articles attempt to show how easy it is to form a shell company, but they have only provided anecdotal evidence or a summary of several examples. (170) By contrast, our study used 7462 approaches to 3773 providers in 181 countries. We randomly assigned a variety of approaches to determine what factors cause providers to be more or less likely to comply with domestic and international regulations affecting the formation of shell companies. (171)

   In our study, we attempt to answer the question of how effective the post-September 11 regulations have been at curbing the incorporation of anonymous shell corporations. We first seek to discover which countries are the most compliant and what factors might contribute to noncompliance. We then explore whether compliance in the United States differs from that elsewhere in the international community. Finally, we compare the results from both the international and domestic tests to analyze what factors make some countries more or less compliant than others.

   1. Design Study: Finding Providers, Composing Treatments

      To determine the effectiveness of post-9/11 financial regulations, we analyzed countries' informal compliance with FATF recommendations and IRS regulations by way of private actors, including firms and incorporation services providers. (172) Mindful that the field experiment is occasionally criticized for contributing only a pragmatic, "what works" analysis, (173) we focused on larger theoretical questions. Accordingly, this field experiment presents more than statistics; we utilize the major international law and relations theories to get to the heart of what actually causes compliance. (174)

      To complete this experiment, we compiled a list of incorporation services providers and created a set of emails to be sent from a number of aliases through which we posed as international consultants seeking anonymous shell corporations. To find and compile the list of providers, because no definitive list exists of incorporation services providers, we performed

      Internet searches for terms such as "company formation" and "business law." We successfully collected a pool of 3773 corporations and law firms drawn from nearly every nation of the world--181 to be precise. Of these, 1785 were from the United States, 444 from other Organisation for Economic Cooperation and Development (OECD) countries, 1039 from developing nations, and 505 from countries with reputations as tax havens. This does not, of course, represent every incorporation services provider, but the sample size is sufficiently large for the purposes of this Article.

      We conducted two experiments using this pool of providers. First, we sought to test the effectiveness of international transparency law--particularly FATF regulations--using a pool of 2051 firms that included 63 U.S. firms and all of the non-U.S. firms. Second, we subjected the remaining 1722 U.S. firms to the same FATF conditions but also presented additional conditions, including a treatment to test the effectiveness of IRS regulations on provider behavior. In both experiments we also explicitly tested the ease with which customers who match the profile of terrorists could incorporate anonymously. All of these conditions are explained below.

      Next, to complete each of these experiments, we randomly assigned and sent emails that were embedded with different experimental conditions. Before discussing how these treatments differed, we first note that each email shared several common features: (1) each was sent from a fictitious customer seeking a consultant; (2) each provided a rationale for wanting a shell company (including reduced liability and confidentiality); and (3) each asked about cost and identity document requirements. Beyond these commonalities, each email was specifically crafted to test compliance with either an international or domestic regulation. Furthermore, the emails allegedly originated from various areas of the world ranging from low-corruption OECD nations (175) to nations that are often associated with terrorism. The recipient firm was thus able to decide to either comply or refuse to comply with transparency standards.

      1. Placebo

         The first email was our "placebo" or baseline condition, (176) which was sent from one of eight smaller, wealthier countries. Several factors made the placebo emails appear the least suspicious. First, the placebo emails hailed from relatively low-corruption OECD countries with conceivably less risk of terrorist influence. These placebo countries were Australia, Austria, Denmark, Finland, the Netherlands, New Zealand, Norway, and Swedeneach listed as among the least corrupt countries on Transparency International's Corruption Perceptions Index (CPI). (177) For convenience, we refer to them collectively as "Norstralia." The variety of placebo countries ensures that a negative means event--for instance, a lurid transnational financial crime story or a government scandal--might bias results. Therefore, other than asking for anonymous incorporation, the placebo email does not contain anything especially suspicious. Where the email hailed from a non-English speaking nation, we injected spelling, syntax, or grammar errors to enhance authenticity. This placebo email served as a benchmark with which to compare response rates and requests for identity documentation under the remaining conditions.

         Including the placebo, we examined twelve different conditions, four of which we report in this Article and summarize in Table 1.

      2. Terrorism Treatment

         In the second treatment, we posed as terrorist risks. (178) The aliases purported to consult for Islamic charities, work in Saudi Arabia, and originate in countries recognized as sites of suicide terrorism--Lebanon, Pakistan, Palestine, and Yemen. (179) In this treatment, we tested the effectiveness of two of the FATF's recommendations: the warning against "[c]ountries identified by credible sources as providing funding or support for terrorist activities that have designated terrorist organisations operating within them" (180) and the requirement that companies screen "[c]harities and other 'not for profit' organisations which are not subject to monitoring or supervision." (181) The combination of individuals (1) coming from a country perceived as a host to terrorists, (2) working for an Islamic charity, and (3) seeking financial secrecy should present a very obvious terrorist-financing risk.

      3. FATF and IRS Treatments

         Our FATF treatment adds to the basic control template a straightforward reference to the FATF. (182) As with the control template, the email purportedly originates from one of the eight "Nostralia" countries, but here, the fictitious consultant directly references FATF provisions that require the production of identification to create a shell corporation. (183) However, after referencing the provisions, the consultant reaffirms a desire for anonymity and asks what documents are actually needed.

         We developed this treatment to test the international law and relations theories of managerialism (184) and legalization. (185) These theories imply that noncompliance results from either ignorance of the law or ignorance of the conditions under which the law applies. Accordingly, we would expect to see high rates of compliance in response to this condition because the email both invokes the relevant law and suggests that this specific context is one in which it applies.

         The next treatment--the IRS treatment--builds upon the FATF treatment by additionally mentioning the possibility of IRS sanctions in the case of noncompliance. (186) Although the IRS does not enforce identity reporting requirements, its website lists the agency as an FATF partner, (187) and we included this information to see if it would act as an additional deterrent. Our expectation was that the additional information would raise the proportion of providers insistent on compliance and lower the number of those willing to do business, noncompliance notwithstanding.

   2. Coding the Responses

      1. Compliance Coding

         After sending out emails according to the aforementioned protocol, we coded five types of responses to measure treatment effects: (1) no response, (2) refusal, (3) compliant, (4) partially compliant, and (5) noncompliant.

         Some typical responses identified the email as a possible scam, (188) while others arguably sought more information and requested a higher premium. (189) A complete lack of response, which could have occurred for a number of reasons, was coded as "no response." When corporation service providers (CSPs) simply refused service, irrespective of the stated reason, we coded the responses as "refusal."

         To be "compliant," a CSP must have asked for specific...