Notes - Global Shell Games

January 18, 2019

Chapter 2: Explaining the global shell game

What are shell companies and their importance?

Shell companies are legal fictions – entities brought into existence by law, possessing only their basic legal essence without physical operations like workers, headquarters, or products. They are powerful and important fictions, capable of holding bank accounts, owning assets, and engaging in transactions. While a company may start as a shell and later acquire substantial business operations, it can also maintain legal existence after its substantive business ceases.

Shell companies are cheap to establish, typically costing a few hundred to a few thousand dollars, and can be formed very quickly online, sometimes in as little as 10 minutes in certain jurisdictions. Rich OECD countries, particularly English-speaking nations, have intentionally made incorporation easy as part of deregulation efforts and reforms to aid small businesses. This philosophy is mirrored in the World Bank's Doing Business Survey, which commends countries facilitating fast incorporation. However, the book notes that the World Bank's Ease of Doing Business rankings are not significantly correlated with the compliance rates found in the experiments, suggesting that non-compliance involves more than just ease of doing business. In contrast, forming a company in developing countries is often more expensive, slower, and involves more bureaucracy.

Shell companies are often confused with companies formed in tax havens or offshore financial centers (OFCs). Tax havens are jurisdictions designed to attract foreign customers through their financial, tax, and corporate laws. While most companies formed offshore are shell companies, the majority of shell companies globally are likely formed in onshore jurisdictions. For instance, leading offshore jurisdictions like the British Virgin Islands and Panama form 40,000 to 70,000 shell companies annually, compared to roughly 300,000 in Britain and 2 million in the United States.

A subtype, shelf corporations, are pre-formed and kept "on the shelf" for later sale. They can reduce waiting times for clients and provide an appearance of longevity or a corporate track record, which can be useful for dealing with banks or partners. For criminals, shelf companies can complicate investigations due to the transfer of control from original formers to new owners.

While the book primarily focuses on illicit uses, most shell companies are used for legitimate business purposes.

  • Walt Disney Corporation used an array of shell companies (e.g., Tomahawk Properties, Reedy Creek Ranch) to hide its identity when purchasing land for Disney World in Florida, preventing price gouging by local landholders.
  • Chinese nationals commonly use foreign shell companies to list and raise capital on US, British, and Hong Kong stock markets because forming a company in China is slow and Chinese firms are generally barred from listing on most foreign exchanges. The foreign shell company lists in the US and owns the original Chinese company as a subsidiary, funneling capital back to China.
  • In joint ventures between partners from different countries (e.g., US and Canada operating in a corruption-prone developing country), shell companies can be set up in a neutral jurisdiction (e.g., England or Channel Islands) to jointly own the operating company, ensuring any future disputes are resolved in a familiar, reliable court system.
  • Shell companies are also used to hold assets such as real estate, capital equipment, share portfolios, or intellectual property.
  • A "borderline example" involves a Russian NGO monitoring elections that receives foreign funding. To avoid registering as a "foreign agent" under new Russian law, the NGO could establish a Russian shell company to receive foreign donations, thereby transforming foreign money into Russian money.

Despite legitimate uses, shell companies are a crucial means for malefactors to obscure illicit financial affairs. The book provides examples:

  • Tax evasion: The US Department of Justice's indictment of Swiss banks UBS and Wegelin illustrates how foreign shell companies were used to hide US clients' undeclared accounts and interest income from the IRS. Even when funds were transferred through unorthodox means like cash-filled envelopes, shell companies provided a more reliable way to hide client identities. For example, a couple (J and K) established "White Tower Holdings LLC" in Nevis to receive money from UBS and open an account with Wegelin, even though Wegelin knew the company's only function was to conceal that J and K controlled the funds.
  • Corruption: The case of Sergei Magnitsky, a Russian lawyer who uncovered a $230 million tax fraud by corrupt Russian officials, tragically illustrates this. Olga Stepanova, a tax official, approved a fraudulent refund, and the money was wired through international banks to shell companies like Bristoll Exports (New Zealand), Arivust Holdings (Cyprus), and Aikate Properties Inc. (British Virgin Islands). Bristoll Exports, formed by the GT Group, had not collected information on the real owner, thus the trail ended with the untraceable shell company, protecting the identities of those behind the fraud.
  • Sanctions-busting: The Islamic Republic of Iran Shipping Lines (IRISL), blacklisted by the US government for its role in Iran's nuclear program, used shell companies to evade international sanctions. IRISL's fleet was renamed, re-flagged, and "sold" to ostensibly independent third-party shell companies based in Germany, Malta, Cyprus, Hong Kong, and the Isle of Man, allowing the ships to continue operations without being on the sanctions list. Despite the transfers, IRISL maintained control through these shell companies.

Corporate Service Providers (CSPs) and the Shell Company Industry

Corporate service providers (CSPs) are professional intermediaries that establish, sell, and maintain shell companies. They are the crucial locus of enforcement and compliance for rules mandating corporate transparency; if CSPs do not collect information on the real owners, no one else will. CSPs can range from large firms with international offices to single individuals operating via a website, or they can be part of law or accounting firms.

The primary function of CSPs is to lodge documentation and pay government fees for their clients, serving as the official contact address for the company. This often means many shell companies share the same official residence (e.g., 138,000 companies at one London address, or a Wyoming house). CSPs also offer a wide range of auxiliary services, including phone/mail forwarding, secretarial assistance, and legal/tax advice. Crucially, they can act as "nominee" directors or shareholders, meaning the corporate registry records their name, but the actual owner retains control through a legal agreement with the provider. This allows the true owners to remain anonymous, as exemplified by Lu Zhang acting as a nominee for the GT Group.

The industry has different types of players:

  • Wholesalers (e.g., Offshore Incorporation Limited, Mossack Fonseca, CT Corp) form and sell companies in bulk to retailers. These large providers may establish thousands or tens of thousands of shell companies and can be worth hundreds of millions of dollars.
  • Retailers (e.g., banks, law firms, sole traders) buy companies from wholesalers and sell them to end-users. For example, Euro-American in New York formed 2,000 Delaware shell companies, selling many to Russian brokers who then sold them to unknown end-users.

The industry also exhibits regulatory competition, where jurisdictions compete for incorporation numbers, leading to a "race to the bottom" in terms of ease of incorporation. Some large CSPs can even influence governments to pass legislation favorable to shell company businesses, as seen with Mossack Fonseca in Niue and Samoa. The "brand" or reputation of a country (e.g., solid US/British vs. exotic/disreputable tax havens) can also affect the desirability of a shell company.

The global corporate transparency standards

Due to the pervasive misuse of shell companies for criminal schemes, international standards have been formulated to regulate their use, with the key aim being the ability to "look through" the corporate veil to find the controlling person(s). Various reports by governments, international organizations, and NGOs consistently highlight the centrality of shell companies in combating serious crimes.

The Financial Action Task Force (FATF) is the most important body in regulating shell companies. Established in 1989 to counter money laundering from the drug trade, its mandate expanded to include terrorism financing.

  • The FATF issues "soft" international law through its 40 Recommendations (later augmented by 9 Special Recommendations to combat terrorism financing, and redrafted in 2012).
  • It has been remarkably successful in getting over 180 states to adopt its measures, partly through public blacklisting campaigns against non-compliant countries.
  • Mutual Evaluation Reports systematically monitor adherence to rules, grading members on compliance.
  • The essence of the Recommendations is to collect more information on those using the financial system.
  • A "risk-based approach" delegates much of the implementation to private firms, especially banks and CSPs, requiring them to evaluate customers and transactions for money laundering risk.
  • Recommendation 24 (formerly 33) is most relevant to shell companies: it mandates that countries ensure "adequate, accurate and timely information on the beneficial ownership and control of legal persons that can be obtained or accessed in a timely fashion by competent authorities". Beneficial ownership refers to the real person in control, distinct from legal ownership by a nominee.
  • Recommendation 22 imposes "extensive customer due diligence and record-keeping responsibilities" on CSPs. This includes "identifying the customer and verifying that customer’s identity using reliable, independent source documents, data or information" (Recommendation 10), such as passports, national identity cards, or drivers’ licenses, and maintaining these records.
  • Other international standards (e.g., UN Conventions Against Corruption, Transnational Organized Crime, Financing of Terrorism) echo this priority. The G20 has also prioritized strengthening standards on beneficial ownership.

The book argues that corporate service providers offer the only reliable route to tracing the real owners of shell companies, as other approaches have limitations.

  • Relying on strong law enforcement powers (e.g., US sting operations) has limitations, especially when information was never collected in the first place. For example, some US providers explicitly state, "Regarding confidentiality, no information is taken so none can be given – it is that simple!".
  • Relying on the company registry is also often insufficient. Most registries function as archives that receive documents but do not verify them, and the information is generally sparse and often incomplete or out of date. Only Jersey requires beneficial ownership information to be supplied for each registered company.
  • The third option, requiring CSPs to collect and verify identity documentation, is key. This involves collecting certified or notarized copies of official documents (e.g., passport, ID card) and proof of address (e.g., utility bills), which must be kept on file and produced upon request by authorities. Jurisdictions like the British Virgin Islands, Cayman Islands, and Jersey have such requirements.
  • However, even this approach has limitations: identity documents are not public, requests from foreign authorities can be delayed, and in countries like the US and Britain, individuals can register companies directly online without an intermediary. CSPs themselves can be negligent, willfully blind, or actively collude with criminals, as seen in the case of Teodorin Obiang and his lawyer Michael Jay Berger, who set up California shell companies for Obiang to obscure his controlling role, as intermediaries in the US were unregulated.

In general, international standards are clear: authorities must be able to "see through" shell companies to find the beneficial owner, which practically means CSPs demanding and holding certified identity documents. The study aims to provide data on how often providers fulfill these obligations.

Experimental design

The study's core is a global field experiment where researchers posed as different customers to solicit untraceable shell companies.

Subject Pool Compilation

  • The first challenge was compiling a comprehensive list of CSPs, which is difficult as many countries (like the US) have no official licensing requirements.
  • Researchers used internet searches (e.g., "company formation," "shell company," "offshore company" with country names) for every country.
  • This yielded a list of more than 3,700 firms in 181 countries, including 1,700 in the US.
  • This list is considered the most comprehensive ever compiled, likely capturing the majority of shell companies worldwide and those involved in cross-border business. However, smaller concerns without a web presence might have been missed, potentially understating non-compliance.

Email Approaches

  • The experiment centered on email solicitations.
  • The Placebo email served as the benchmark, sent from aliases in "Norstralia" (Australia, Austria, Denmark, Finland, Netherlands, New Zealand, Norway, Sweden) – developed, minor powers with low corruption and terrorism risk. The emails purported to be from international consultants seeking to limit taxes and reduce liability, asking about required identifying documents and prices, and desiring confidentiality.
  • Ten different treatment emails included crucial variations to signal different risks or inducements.
  • Each CSP was randomly assigned at least two different emails (Placebo and a treatment, or two different treatments), spaced months apart to minimize detection.
  • 33 different base emails were drafted, varying style, diction, and syntax, with minor spelling/grammar errors for non-English-speaking aliases to enhance authenticity.
  • Summary of treatments mentioned:
    • FATF (Managerial): Explicitly referenced FATF requirements for identity documents.
    • Penalties: Included FATF standards plus a statement about possible legal penalties for violations.
    • Norms: Referenced FATF standards, stated they were widely accepted, and expressed a desire to "do the right thing" as "reputable businessmen".
    • ACAMS: Attributed identification standards to the private Association of Certified Anti-Money Laundering Specialists.
    • ACAMS + FATF: Combined reference to both ACAMS and FATF standards.
    • Premium: Offered extra payment if the rule to collect identity documents was waived.
    • IRS: For US providers, attributed identification standards to the US Internal Revenue Service.
    • US Origin: Placebo frame but customer was a US citizen.
    • Corruption: Customer from a corruption-prone "Guineastan" country (Guinea, Equatorial Guinea, Guinea-Bissau, Kyrgyzstan, Papua New Guinea, Tajikistan, Turkmenistan, Uzbekistan) and working in government procurement.
    • Terrorism: Customer from a terrorism-prone country (Lebanon, Pakistan, Palestine, Yemen), living in Saudi Arabia, and consulting for a Muslim charity.
  • The general cover story of "consultancy business" provided a plausible yet vague backstory, as it is a common alibi for illicit pursuits and money laundering.
  • Measures were taken to conceal the true source and purpose of emails, including using internet email accounts registered to Ugandan mobile numbers, setting up accounts via proxy servers in Kampala, and using proxy servers for all communications to obscure IP addresses.
  • Standardized responses were used for common queries, and direct phone/Skype conversations were avoided to maintain anonymity and a record of communication.

Coding Responses

  • Responses were classified into five outcomes: no response, refusal, compliance, partial compliance, and non-compliance.
  • No response: Firms made no reply after being prompted twice. This outcome is ambiguous; it could mean "soft compliance" (firm judged client too risky and ignored), or simply scam/spam deletion, or low business return. Follow-up checks suggested most non-responses were not soft compliance.
  • Refusal: Providers explicitly stated they could not/would not help. Very few corporate service providers responded with outright refusals.
  • Compliance: The provider required certified or notarized copies of official government photo identification documents (e.g., passport picture page, national ID card) and often secondary proof like utility bills. This information is kept on file and produced to authorities upon request to "pierce the corporate veil". An example from St. Kitts and Nevis is provided, requesting certified copies of passport/driver's license, two certified utility bills, and two reference letters.
  • Partial compliance: The provider asked for some proof of identity but not certified copies of official photo ID. This would yield some information but is easier to falsify.
  • Non-compliance: The provider was willing to supply a shell company without any photo identification at all. This means the resulting company is untraceable, and authorities would have few options to connect it to the real individual in control. Examples of non-compliant responses include: "If we are just setting up a Corporation for you and that’s it we don’t require any documents from you at all," and "All that you need to do is to provide the name you want for you [sic] new company, that’s it".
  • The study did not actually buy shell companies due to expense and the illegality of signing legal documents under false names. Previous audit studies confirmed that providers' stated requirements for identity documentation remained consistent throughout the process.

Chapter 3: Overall compliance, tax havens, OECD and developing countries

Overall results

The overall descriptive findings from the study indicate a significant compliance problem with global rules on corporate transparency.

  • Almost half (48.2 percent) of all replies received did not require certified identity documents.
  • More than a fifth (22.1 percent) of replies required no photo identification whatsoever.
  • Untraceable shell companies are remarkably easy to obtain. An individual could likely find one within half a day of searching.
  • International CSPs were less likely to offer anonymous shells (16.4 percent non-compliance) compared to US CSPs, which had an astounding 41.5 percent non-compliance rate among responses.
  • A further 26.1 percent of replies asked for some identity documentation but not certified/notarized copies (partial compliance).
  • Only 24.9 percent of replies were fully compliant, and 26.9 percent refused service.

A complication in interpreting results is the large number of non-responses: 49.0 percent in the international sample and 77.3 percent in the US.

  • If these represent "soft compliance" (firms avoiding risky clients by not responding), then the pessimistic conclusions about effectiveness might be misleading.
  • The "Dodgy Shopping Count" (DSC) is a simplified measure of non-compliance, gauging the average number of providers a customer would have to approach to receive a non-compliant response (no photo ID required). For example, a 5 percent non-compliance rate equals a DSC of 20. A lower DSC means it's easier to get an anonymous shell corporation.
  • The book notes that even high effectiveness rates (e.g., 95 percent for terrorism financing) might not be enough to create real policy impact, as search costs are still low.
  • To test if non-responses were soft compliance, a follow-up email was sent to non-responsive providers, stripping away all risk factors and simply asking if the firm was still in business. A large majority (88 percent) of these firms still failed to respond, indicating that non-response is weak evidence of soft compliance.

Comparing aggregate results with expectations

The FATF's own Mutual Evaluation Reports indicate that Recommendation 24 (beneficial ownership) has long had one of the lowest overall levels of assessed compliance. Only 7 percent of 159 countries were graded "Compliant" and 15 percent "Largely Compliant" for this recommendation. This suggests that the study's findings of significant non-compliance might not be surprising, given existing FATF reports based on statutory compliance.

However, the study found a weak correlation between FATF-reported statutory compliance and actual compliance rates in their field study. For example, statutory compliance with Recommendations 10 and 20 was not significantly related to actual compliance, and Recommendation 24 showed only a modest, weak statistical relationship. This "especially raises questions about standard measures of compliance relying on statutory faithfulness to international law".

The book posits that if few countries legally require CSPs to identify customers, why do so many still require some form of ID?

  • Some providers may feel a moral duty to take precautions with shady customers.
  • Others might fear reputational damage if their company was involved in malfeasance.
  • Even without criminal or regulatory penalties, civil law liabilities might exist for reckless providers, especially if they acted as nominee directors.

Compliance by country type: The rich, the poor, and the havens

The study's descriptive results on compliance rates among rich, poor, and tax haven countries starkly contradict conventional wisdom.

  • Conventional wisdom: Poor countries are unable to comply (lack of capacity), tax havens are unwilling (attract business through lax regulation), and rich countries are both able and willing to comply.
  • Study findings: Tax havens demonstrated by far the highest levels of compliance. Rich and poorer countries showed similar levels, but differences tended to favor developing nations. This suggests the conventional wisdom is "wrong on each count".

Defining country types:

  • Developing countries are based on the World Bank's categorization of middle- and low-income countries. They are assumed to have difficulty meeting financial standards due to lack of money, skilled professionals, and bureaucratic hurdles.
  • Rich countries (OECD members) are assumed to regulate their financial sectors well due to economic dependence on them and greater resources.
  • Tax havens (or offshore financial centers) are defined pragmatically using the OECD's classification. They are conventionally condemned as primary culprits for financial secrecy, attracting foreign capital through strict secrecy, low taxes, and minimal regulation, often via anonymous shell companies. The economic logic suggests tax havens have a stronger incentive to defy transparency regulations due to a "cut-throat market of extremely mobile potential clients".

The results: Confounding the conventional wisdom

The findings directly contradict prevailing expectations: tax havens exhibit superior compliance rates, and developing countries are at least as compliant as developed ones.

  • Response rates: Tax havens had a much higher response rate (64.9 percent) than rich (49.4 percent) or poorer countries (44.6 percent). This is likely because haven providers are more accustomed to international inquiries. The high response and compliance rates of tax havens demonstrate that it's possible to be both strictly regulated and internationally competitive.
  • Compliance levels: Tax havens outperformed OECD countries on nearly every count: lower non-compliance, higher partial compliance, and higher full compliance, all at statistically significant levels.
  • Dodgy Shopping Count (DSC):
    • Tax havens: 25 (meaning, on average, a customer would need to contact 25 providers to get an anonymous shell).
    • Developing countries: 11.9.
    • Rich countries (OECD): 7.8.
    • This implies it is more than three times more difficult to obtain an anonymous shell in tax havens than in OECD member countries, which is the opposite of conventional wisdom.
    • Some stereotypical tax havens like the Cayman Islands, Bahamas, and Seychelles had not a single non-compliant reply.
  • OECD countries were well-represented at the lower end of compliance rates. Among 102 jurisdictions, US law firms ranked 38th, Britain 52nd, Canada 69th, Australia 74th, the US overall 86th, and US incorporation firms were dead last at 103rd.
  • US states: Delaware, Wyoming, and Nevada, along with Alabama and Montana, showed up at the bottom for compliance. Examples of non-compliant US responses were "We don’t need a whole lot of info from you. You can place the order on our website under starting your company. It should only take 10 minutes and that is all the information we need from you," and "All that you need to do is to provide the name you want for you [sic] new company, that’s it".
  • While some jurisdictions (Cayman Islands, Libya, Utah) had high rankings on charts, their patterns of response varied, highlighting that different processes can lead to the same non-provision of untraceable shells.

Explaining the results

Two main explanations are considered for these surprising results:

  • International Pressure on Tax Havens: Firms in tax havens comply at high rates because these jurisdictions have been subjected to intense scrutiny and international pressure to improve corporate transparency since the late 1990s, mainly from clubs of rich states like the FATF and OECD.
    • Historically, many offshore centers had a "Wild West" ethos, with easy anonymous shell companies and lax banking.
    • In response, clubs of like-minded developed states (OECD via "Harmful Tax Competition" and FATF via blacklists) coordinated action, concluding that tax havens were the main problem in financial secrecy.
    • These blacklists proved effective by damaging reputations and raising costs for transacting with listed countries. Tax havens and even unlisted jurisdictions pre-emptively tightened standards.
    • However, the pressure was partial, as targeted non-members were held to higher standards than the rich club members themselves. For instance, the FATF required non-members to ensure authorities could obtain beneficial ownership information ("essential"), while FATF members only needed to "consider" additional measures. This double standard created a gap in regulatory stringency.
  • Deregulation in OECD Countries: Many developed states, particularly English-speaking ones, have attempted to remove business barriers through deregulation, making incorporation quick and easy.
    • This includes moving the incorporation process online and minimizing documentation requirements.
    • In the US, competition between states (e.g., Delaware) further incentivizes quick and cheap company formation, generating substantial fee revenue.
    • The World Bank's Doing Business report measures the ease of starting a business, and countries often prioritize improving their ranking. For example, New Zealand officials were reluctant to reform corporate laws even after untraceable shells were linked to arms smuggling and drug cartels, fearing jeopardizing their Doing Business ranking.
    • In contrast, forming a company in most developing countries remains a bureaucratic ordeal, which might unintentionally make anonymous incorporation difficult.
    • However, the study found very little evidence that general business deregulation directly facilitates anonymous incorporation. There was no significant association between compliance rates and most Ease of Doing Business rankings for OECD countries or all countries. Only the cost of startup was negatively related to compliance, but this might be an artifact of country wealth. Therefore, the general ease or difficulty of starting a business does not appear to significantly affect compliance with global corporate transparency standards.

Chapter 4: Terrorism and corruption

Introduction to Experimental Interventions

This chapter shifts from descriptive findings to experimental research, analyzing the impact of specific interventions on corporate service providers' (CSPs) compliance with global transparency standards. The focus is on the results of three experimental treatments: Terrorism, Corruption, and Premium. The research aims to understand the causes of compliance and non-compliance among CSPs, which is crucial for policymakers combating terrorism financing and corruption, areas where traditional measures of effectiveness are often unknown.

Explaining the Terrorism Treatment

The study acknowledges that knowledge about terrorism financing is often tentative and disputed due to national security secrecy. While terrorist attacks can be cheap and may not always involve shell companies, international standards, particularly those from the Financial Action Task Force (FATF), mandate that financial institutions assess terrorism financing risk based on country of origin. The Terrorism treatment emails signaled risk by having the fictitious customer originate from terrorism-prone countries (e.g., Pakistan, Yemen) and claim to work for an Islamic charity in Saudi Arabia, which is a recognized "red flag" by FATF. The cross-border nature of the activity further heightened the risk signal.

Terrorism Findings

The results for the Terrorism treatment showed some encouraging signs:

  • Increased Non-Response: Non-response rates significantly increased for the Terrorism treatment (from 44.5% to 58.3% internationally, and 73.8% to 83.3% in the US sample). This is interpreted as a form of "soft compliance," where providers avoid engaging potentially risky customers.
  • Decreased Non-Compliance: Among providers who did respond, non-compliance levels (offering a shell company with no photo ID) were significantly lower (from 8.7% to 5.7% internationally, and 11.3% to 5.8% in the US sample). This indicates that potential terrorists find it more difficult to obtain untraceable shell companies.
  • Mixed Results and Caveats: Despite these positive signs, the system is "far from watertight". The Dodgy Shopping Count increased, meaning more approaches are needed to find a non-compliant firm (from 11.5 to 17.5 internationally, 8.9 to 17.2 in the US). However, the treatment also led to a decrease in partial compliance and full compliance in the international sample, and a decrease in refusal rates in the US, suggesting that more law-abiding firms might also be dropping out. Some responses, even from compliant-seeming CSPs, were still concerning, showing a willingness to bypass or ignore identification requirements for a price.

Explaining the Corruption Treatment

Shell companies are "the most common laundering device in major corruption cases," including bribery and embezzlement of state funds. The chapter provides examples:

  • Australian Reserve Bank case: Australian companies involved in printing bank notes paid "consultancy fees" via shell companies in tax havens to agents to win contracts abroad, which turned out to be bribes.
  • Oil industry example: Local officials set up shell companies as "local partners" to extract bribes from foreign oil companies by receiving a percentage of revenues for doing nothing.
  • Zimbabwean diamond mines: Secret filming showed Seychellois CSPs explicitly advising journalists posing as corrupt Zimbabwean officials on how to use shell companies and sham consultancy arrangements to launder looted wealth, even encouraging ignorance of the source of funds. The Corruption treatment signaled risk through two modifications: the customer's nationality (from "Guineastan" countries known for high corruption, like Equatorial Guinea or Kyrgyzstan) and their claimed involvement in "government procurement," an industry notoriously prone to corruption. The combination of these factors was designed to be a "blatantly obvious invitation to aid and abet criminal behavior".

Corruption Findings

The results for the Corruption treatment were largely disappointing:

  • Increased Non-Response: Similar to the Terrorism treatment, non-response rates increased significantly (from 44.5% to 52.6% internationally, and 73.8% to 78.4% in the US sample). This suggests some CSPs avoid inquiries from risky clients.
  • Indifference to Risk: Beyond non-response, the treatment effects were small and largely insignificant. The non-compliance rate showed almost no change, remaining around 8.7-8.9% internationally and 8.9-9.8% in the US sample. This implies that CSPs are "almost completely insensitive to the danger of corruption".
  • Decreased Compliance and Refusal: In the international sample, compliance rates decreased (from 18.9% to 15.0%), suggesting that compliant providers may have been deterred from responding at a higher rate than non-compliant firms. In the US, refusal rates significantly dropped (from 13.0% to 9.8%), further indicating a willingness to enable corruption.
  • Collusion Effect: The findings suggest a "collusion effect" where CSPs, despite clear signals of corruption risk, appear willing to facilitate illicit activities for profit.

Explaining the Premium Treatment

The Premium treatment aimed to test how direct financial incentives influence compliance. It explicitly offered an "unspecified extra payment" if CSPs would waive the Know Your Customer (KYC) rule requiring identification. This tests the "logic of consequences," assuming providers would respond to increased rewards for non-compliance. The chapter provides examples of lawyers and financial intermediaries who were actively complicit in money laundering schemes for financial gain, like Bhadresh Gohil for James Ibori, or Michael Berger for Teodorin Obiang, showing that such complicity can be incentivized not just by direct payment but also by lavish "fringe benefits". Conversely, the offer could also raise suspicion, as FATF guidelines identify "extraordinary fees for services which would not ordinarily warrant such a premium" as a red flag.

Findings for the Premium Condition

The Premium treatment yielded disturbing results:

  • Reduced Compliance: The clearest finding was a "pronounced decrease in the rate at which providers demand full, certified identify documentation". Compliance dropped significantly from 18.9% in the Placebo to just 1% in the Premium condition, a 4.5 percentage point decrease. This suggests that the prospect of enrichment can outweigh moral qualms or fear of penalties.
  • Increased Non-Response: Non-response rates increased (from 44.5% to 49.6%), indicating some CSPs were wary.
  • Mixed Overall Effect: While non-compliance rates slightly decreased, this was not statistically significant. The Dodgy Shopping Count increased, implying it was harder to get an anonymous shell, but this result was also not statistically meaningful. The primary takeaway is that paying extra directly incentivizes CSPs to break international law.

Conclusion for Chapter 4

The chapter concludes that the three treatments elicited mixed responses, demonstrating heterogeneity among providers.

  • Terrorism: There's significant sensitivity to terrorism financing risk, leading to higher non-response and lower non-compliance, but obtaining untraceable shells is "far from watertight".
  • Corruption: Providers showed "general indifference to very unsubtle hints about...corruption". This is particularly alarming given that shell companies are major enablers of large-scale corruption. CSPs seem more alert to the less common risk of terrorism financing than the "much more important one of corruption".
  • Premium: The results indicate that some providers are willing to violate rules for financial gain, confirming that non-compliance can be "bought". Overall, the chapter highlights the challenge of promoting compliance when faced with profit motives, with interventions often having unexpected or counter-intuitive effects.

Chapter 5: Laws and standards

Introduction

This chapter investigates whether informing CSPs about relevant laws and standards influences their compliance. It specifically compares the effects of standards attributed to:

  • International law (FATF).
  • Private standards (Association of Certified Anti-Money Laundering Specialists - ACAMS).
  • Domestic enforcement (US Internal Revenue Service - IRS). The underlying theoretical question is whether non-compliance stems from ignorance of rules, as suggested by "managerial" and "legalization" perspectives in international law.

Private Authority, Non-State Actors, and Transnational Relations

The chapter contextualizes the study within the broader trend of "global governance without government," where private actors increasingly play roles as both regulators and regulated entities. Private authority can stem from market influence, moral standing, specialized expertise, or the ability to "get the job done". Examples include credit-rating agencies and organizations like ACAMS that set standards without direct governmental authority.

Knowledge, Rules, and Compliance

The "managerial approach" to international law argues that non-compliance is often due to ambiguity or lack of clarity in rules, or actors' "incapacity" or "ignorance" of them. If this is true, then providing information about rules should significantly improve compliance. This chapter empirically tests this observable implication.

Explaining the FATF Treatment

The FATF (Financial Action Task Force) is the authoritative global standard-setter for anti-money laundering and counter-terrorism financing. Its Recommendation 24 mandates that authorities have "adequate, accurate and timely information on the beneficial ownership and control of legal persons". The FATF treatment explicitly mentioned these international standards, aiming to test if informing CSPs about them would increase compliance. The authors acknowledge that some CSPs might already know these rules, in which case the treatment would act as a "priming" mechanism. However, a post-experiment survey suggested that a significant proportion (64.1%) of CSP representatives had never been briefed on FATF recommendations.

FATF Findings

The results for the FATF treatment were "surprising and puzzled" the authors.

  • No Significant Effect: Across both international and US samples, informing CSPs about FATF international law had no statistically significant impact on their behavior. Non-compliance rates, partial compliance, compliance, and refusal rates remained statistically equivalent to the Placebo condition.
  • Dodgy Shopping Count: The Dodgy Shopping Count was effectively identical between the Placebo and FATF treatment, meaning it was just as easy to obtain an anonymous shell company. These "null results" challenge the notion that non-compliance stems from ignorance and that simply providing information about international law will motivate greater compliance.

ACAMS Treatment

The ACAMS (Association of Certified Anti-Money Laundering Specialists) is a for-profit private organization that offers certification and training in anti-money laundering. It diffuses and entrenches FATF standards among private financial firms, acting as a private standard-setter without direct delegated authority from governments. The ACAMS treatment tested whether attributing the same identification standards to a private, rather than inter-governmental, body would elicit a different response. The hypothesis was that ACAMS might carry more weight if CSPs were more familiar with it or valued its certification.

ACAMS Findings

The results for the ACAMS treatment were "not strong and generally mixed".

  • Limited Impact: While mention of ACAMS "does produce greater levels of compliance (5.9 percentage points higher)," this result did not reach standard levels of statistical significance in basic comparisons. However, it showed "significantly higher compliance" when compared to FATF and other treatment conditions in more specific statistical tests.
  • No Effect on Non-Compliance: The Dodgy Shopping Count remained similar to the Placebo and FATF treatments, indicating no significant reduction in the ease of obtaining untraceable shell companies. Overall, while there's "some evidence that invoking private certification standards produces greater levels of compliance," the effect is "likely small".

IRS Treatment

This treatment was applied only to CSPs in the US sample and tested the effect of attributing identification standards to the US Internal Revenue Service (IRS), a powerful domestic enforcement agency. The rationale was that, unlike the FATF, the IRS has direct coercive powers (fines, imprisonment) and a widely known public profile. It aimed to determine if the "prospect of penalties" would motivate compliance. Examples of IRS enforcement actions against Swiss banks and tax preparers for assisting tax evasion through shell companies were highlighted as reasons CSPs might be wary. Importantly, the IRS does not actually require providers to collect official identity documents for beneficial ownership; thus, the treatment "primed" providers by raising the salience of prospective punishment rather than accurate factual information.

IRS Findings

The IRS treatment produced "much stronger effects" than the FATF or ACAMS treatments.

  • Decreased Non-Compliance: Non-compliance rates for the IRS treatment dropped significantly from 11.3% in the Placebo to 7.6%, a one-third reduction. The Dodgy Shopping Count increased from 8.9 to 13.2, indicating it was significantly harder to get anonymous shells.
  • Increased Non-Response: The non-response rate significantly increased (from 73.8% to 80.0%), suggesting CSPs were deterred from engaging with a potentially risky inquiry.
  • Decreased Refusal: Interestingly, refusal rates also dropped, which is less normatively positive than compliance, though the impact on overall non-compliance was still positive. These results suggest that raising the "specter of a powerful domestic enforcer like the IRS deters CSPs from offering anonymous shells".

Conclusion for Chapter 5

The chapter concludes that informing CSPs about international standards (FATF) or private standards (ACAMS) had no significant effect on their compliance, but invoking a powerful domestic agency like the IRS did. This challenges the managerial and legalization schools' assumption that ignorance causes non-compliance and that simply providing information will lead to adherence. It suggests that "domestic laws and practice have simply not yet caught up with international standards", and that meaningful enforcement at the domestic level is crucial to curbing anonymous incorporation.

Chapter 6: Penalties, norms, and US origin

Introduction

This chapter delves into a core debate in social sciences and International Relations (IR): whether actors are primarily motivated by penalties (logic of consequences) or social norms (logic of appropriateness). The "daycare experiment" anecdote (where fines for late pick-ups increased tardiness, implying a shift from social obligation to a transactional cost) highlights how penalties can sometimes backfire and that the relationship between norms and penalties can be complex. The chapter tests these logics, along with the impact of US origin on compliance rates, revealing surprising results that often contradict initial expectations.

Penalties Treatment

Drawing on rationalist IR theories, the Penalties treatment aimed to test whether explicitly signaling the "prospect of penalties" for non-compliance with FATF standards would increase compliance. The treatment emails mentioned that "legal penalties may follow violation of these standards," alongside the FATF's disclosure requirements. The expectation was that fear of sanctions would lead CSPs to refuse service or demand certified identity documents. The FATF indeed mandates that states impose "effective, proportionate and dissuasive" penalties for non-compliance.

Penalties Results

The results for the Penalties treatment were surprising and mixed:

  • No Significant Decrease in Non-Compliance: The non-compliance rate (offering anonymous shells) was largely unchanged from the Placebo, and the Dodgy Shopping Count was not meaningfully different. This contradicts the expectation that fear of penalties would reduce the supply of untraceable companies.
  • Decreased Refusal Rate: The refusal rate significantly decreased (from 11.3% to 7.7%), suggesting fewer CSPs were outright rejecting such inquiries.
  • Decreased Compliance (in some models): More sophisticated statistical models (multinomial) indicated a statistically significant decrease in the overall compliance rate.
  • "Collusion Effect": The findings suggest a "collusion effect" where more scrupulous (compliant) firms might have opted to "drop out" (increase non-response or decrease refusal) when faced with the mention of penalties, while less scrupulous firms saw it as an "invitation to collude in prohibited but profitable conduct". This means the threat of penalties either had no positive effect or actually made compliance worse among the responding firms.

Norms Treatment

This treatment drew on constructivist IR theories and the "logic of appropriateness," which suggests that actors comply with standards because they seek to follow social rules and conventions, or "do the right thing". The Norms treatment emails highlighted that FATF standards were "widely accepted" and expressed a desire to "do the right thing by international rules" as "reputable businessmen". This aimed to activate a sense of social obligation or reputational concern among CSPs. The expectation was that such an appeal would increase compliance and reduce the availability of untraceable shell companies.

Norms Results

The results for the Norms treatment were also mixed and not definitively positive:

  • No Clear Increase in Compliance: Providers were "less compliant when presented with the Norms treatment relative to the Placebo" (15.6% vs. 18.9%), though this was just outside standard significance levels. Statistical models also showed no significant relationship to outcomes or even a "statistically significantly lower levels of compliance" in some specifications.
  • Collusion Effect (again): Similar to the Penalties treatment, the results might suggest a "collusion effect" where the appeal to "appropriate behavior" might have reassured less scrupulous firms that the customer was "lower risk" or seeking complicity.
  • Overall Ineffectiveness: The Dodgy Shopping Count was not meaningfully different. The chapter concludes that attempts to foster greater compliance through appeals to shared norms "may actually make the situation worse" or have "little impact".

US Origin Treatment

This treatment tested the impact of US power and extra-territorial jurisdiction. The alias in the email was changed from a "Norstralian" (minor-power OECD country) to a US national ("Mark Brown"). The expectation, based on realist IR theory and well-publicized US enforcement actions against foreign banks for assisting US tax evasion (e.g., UBS), was that foreign CSPs would be less willing to provide anonymous shells to US clients due to fear of US penalties. The treatment acknowledged confounding factors like US wealth and low corruption rates.

US Origin Results

The findings for the US Origin treatment were "surprising" and "directly contrary to our expectations".

  • No Increased Scrutiny: The non-response rate and non-compliance rate remained virtually unchanged from the Placebo condition. This suggests that US customers do not receive "increased scrutiny" and the risk of US extra-territorial enforcement "is not a major concern for international service providers".
  • Lower Compliance: In fact, there was a significant decrease in the compliance rate (from 18.9% to 14.0%), meaning CSPs were less likely to demand full certified ID from US customers.
  • Counterintuitive Findings: The results suggest that any heightened skepticism toward US clients is "counterbalanced by the greater attractiveness of American customers". This challenges the assumption that US hegemony automatically leads to greater compliance with US-driven standards.

Conclusion for Chapter 6

The chapter concludes that the interventions designed to promote compliance (Penalties, Norms, US Origin) "substantially increased compliance with international standards on beneficial ownership" in "none" of the cases. In fact, some "actually seemed to undermine the enforcement of the standard". This is a "sobering discovery" that underlines the challenge of promoting compliance when profit motives favor non-compliance. The findings suggest that expected interventions often have no impact, or even counter-intuitive effects, pointing to the complexity of behavioral change in transnational relations. The results indicate heterogeneous effects among providers, where some might be deterred while others are encouraged to collude.

Chapter 7: Conclusion

Introduction

This concluding chapter re-emphasizes that profit-driven crime relies heavily on financial secrecy, for which untraceable shell companies are crucial. The book advocates for an "experimental science of transnational relations" (Experimental TR) as a new, rigorous approach to studying world politics by focusing on non-state actors through field experiments. It aims to dispel the myth that experiments in IR are impractical or unethical. The study highlights that global rules increasingly depend on non-state actors for compliance, making this research relevant for both scholars and policymakers.

What Have We Learned About Shell Companies?

Overall Compliance Problem

The study reveals troubling news about global financial transparency:

  • Widespread Non-Compliance: Just under half (48%) of all replies received did not request appropriate identification, and almost a quarter (22%) requested no photo identification whatsoever.
  • Ease of Acquisition: Untraceable shell companies are "remarkably easy to obtain". The "Dodgy Shopping Count" (average number of approaches needed to find a non-compliant firm) was 12 internationally and 10.9 in the United States, meaning it takes very little effort.
  • US as a Major Problem: In the US, 41.5% of replying CSPs required no photo ID, which is two-and-a-half times higher than the international average.

Surprising Geographic Patterns

The study debunked conventional wisdom regarding compliance levels:

  • Tax Havens Outperform: Contrary to expectations, tax havens were among the most law-abiding countries concerning beneficial ownership standards.
  • OECD Countries are Least Compliant: Wealthy OECD countries were the least compliant, even when compared to poorer developing countries.
  • US States as "Worst in the World": CSPs in the US, particularly in Delaware, Indiana, Wyoming, and Nevada, were identified as "among the worst in the world," with only 9 out of 1,722 US providers requiring full certified ID. These country and state-level findings are descriptive, not experimental.

Experimental Results Overview

The experimental findings generally showed "insensitivity of providers both to information regarding the rules they should be following, and in response to even obvious red flags about customer risk".

  • Terrorism Treatment: Offered a "glimmer of hope". CSPs were significantly less likely to respond to potential terrorist threats, and when they did, non-compliance significantly dropped. This provides evidence of CSPs' potential sensitivity to customer risk.
  • IRS Treatment: For US providers, mentioning IRS enforcement led to significantly lower response rates and decreased non-compliance.
  • Corruption Treatment: Despite the "magnitude of human suffering" caused by corruption, CSPs were less likely to demand certified ID in response to corruption signals and seemed "indifferent to the damage they cause".
  • Premium Treatment: Offering extra money to preserve anonymity decreased the compliance rate.
  • FATF & ACAMS Treatments: Among the "most sobering discoveries" was that making CSPs aware of international standards (FATF) or private certification standards (ACAMS) "motivates no more compliance than a Placebo condition". There was no evidence that information about international law or private standards increased compliance.
  • Penalties, Norms, US Origin Treatments: These treatments yielded mixed or unclear patterns. The threat of legal penalties had little impact, and in some models, compliance even decreased, possibly due to a "collusion effect". Appealing to norms had almost no effect or a marginally negative one. Approaches from a US alias resulted in lower compliance rates, contrary to expectations of increased scrutiny due to US power.

Overall Implications of Treatments

The results generally point to mixed conclusions, with some treatments increasing non-response but sometimes simultaneously decreasing compliance or refusal rates, suggesting complex "heterogeneous treatment effects". The "important and telling null findings," such as the ineffectiveness of international standards despite vast coordination efforts, highlight that much more work is needed in financial transparency.

Experiments as a Solution to Scholarly and Policy Problems

The study argues that field experiments can help resolve the "intellectual dead end" in IR scholarship regarding compliance, particularly the problems of endogeneity and selection bias. Policymakers, including the FATF, are increasingly dissatisfied with measuring only "technical compliance" (laws on paper) and seek to measure "effectiveness" (actual behavior), which field experiments are uniquely suited to provide. The book asserts that experiments bridge the gap between academic and policy concerns by providing accurate causal inferences. Knowing "what causes what" is essential for both understanding and effecting change in global governance.

Experimental Transnational Relations (Experimental TR)

The book aims to establish the value of experiments for IR scholars and provides examples of other field experiments undertaken by the authors' labs, demonstrating the "general applicability of this approach". These examples, which generally do not involve deception (unlike the shell games study due to its unique context), include:

  • Crowdsourcing development information in Uganda using mobile phones and incentives.
  • Testing confirmation bias of microfinance institutions regarding research effectiveness.
  • Evaluating the impact of an NGO scorecard on accountability and transparency in Uganda.
  • Probing citizen support for foreign aid projects based on funding source in Uganda.
  • Studying municipal incentives for foreign direct investment based on timing and origin.
  • Analyzing online charitable contributions and susceptibility to different appeals.
  • Mapping NGO coordination to improve communication.
  • Investigating foreign origin, status, and gender effects on helping behavior in Uganda. These examples illustrate the wide range of low-cost, practical field experiments possible in IR, emphasizing that researchers should explore beyond these confines.

Conclusion

The study concludes by affirming its contribution to IR scholarship by introducing a new experimental approach, testing key theoretical propositions, and suggesting ways forward for studying transnational relations. It provides "provocative findings" on worldwide financial transparency and the causes of compliance, noting that good social science often raises more questions than it answers. The authors hope their work stimulates discussion and productive action in both academic and policy circles regarding the mechanisms of law-abiding behavior and how to measure it.