25 February 2020
Anti-money laundering: The World's Least Effective Policy Experiment? Together, we can fix it
Anti‑Money Laundering: The World’s Least Effective Policy Experiment — How to Fix It
Anti‑money laundering (AML) regimes have become a near‑universal feature of modern governance. Worldwide, governments have adopted complex laws based on Financial Action Task Force (FATF) standards and compelled banks, lawyers, accountants and other “gatekeepers” to collect information, monitor flows and report suspicious activity. Yet three decades into this global experiment, the evidence suggests the policy’s core design is delivering negligible impact against the criminal proceeds it set out to disrupt. This article summarizes the problem, quantifies the scale of failure as the evidence allows, explains why the failure persists, and outlines what must change to make AML meaningful rather than mostly performative.
Why the question matters
When FATF was established in 1989, the political objective was straightforward: use financial information to detect and prevent serious profit‑motivated crime and thereby reduce its social and economic harms. Despite the global effort that followed, there are no robust, measurable indicators showing the AML regime has achieved those crime‑reduction goals. Where outcomes are reported, they measure actions, inputs or outputs (filings, prosecutions, laws adopted), not impacts on criminal finances or harms to victims. Without clear outcome measures and validated data, policy remains blind to whether it works.
The evidence: trivial interception of criminal proceeds and huge private costs
Estimating precisely how much criminal property AML efforts recover is challenging because data are fragmented, methodologies vary and many jurisdictions report seizures differently. Reasonable aggregation of available, imperfect data produces the following, sobering picture:
-
Global illicit proceeds
Using UN‑based estimates as a baseline, illicit proceeds in recent years can be placed in the low‑trillion dollar range annually (for illustrative purposes, roughly $2.5–$3.1 trillion).
-
Global confiscations
Authorities worldwide appear to confiscate only a few billion dollars of illicit assets per year. A conservative working figure is about $3 billion confiscated annually.
-
Implied “success rate”
If $3 billion is confiscated from $3 trillion of illicit proceeds, the fraction of criminal funds intercepted is roughly 0.1%. Adjusting for the fact that many confiscations result from traditional investigations rather than AML‑driven reporting (empirical work suggests conventional policing can account for a large share of forfeitures), the portion actually attributable to AML systems could be an order of magnitude smaller; a defensible indicative mid‑point is about 0.05% (one‑twentieth of one percent).
-
Compliance costs
Private sector AML compliance costs are enormous. Studies imply global private compliance expenditures in the order of hundreds of billions of dollars annually; one cautious extrapolation suggests roughly $300 billion per year. Other market estimates are larger. Even using conservative estimates, compliance costs are many tens to hundreds of times greater than amounts recovered from criminals.
-
Penalties
Regulators levy multibillion‑dollar penalties on banks and firms for AML breaches; in 2019 fines alone exceeded $8 billion worldwide. That further shifts economic burden onto regulated institutions and, ultimately, customers and taxpayers.
-
Distributional consequences
Ordinary citizens and marginalized groups experience collateral harms: restricted access to banking, higher costs of financial services, and de‑risking that pushes clients into cash or informal channels.
Taken together, these figures portray an intervention that consumes heavy resources and inflicts tangible costs on lawful actors yet recovers a vanishingly small share of criminal proceeds.
Why AML remains so weak: design problems more than implementation
The industry narrative commonly blames banks and uneven national implementation. That story is partial at best. Three interlocking design issues explain why repeated regulatory intensification has not moved the needle materially.
-
Absence of clear, measurable success metrics. FATF’s “high‑level objectives” and its post‑2014 “effectiveness” methodology use the language of outcomes but routinely assess activities, outputs and capacities rather than impact. Without explicit, measurable crime‑reduction targets (for example, proportion of criminal proceeds disrupted or reductions in specific harms), systems cannot be judged successful even in principle.
-
A compliance‑process operating model misaligned with crime disruption. AML policy primarily prescribes complex controls across broad swathes of industry: client due diligence, transaction monitoring and suspicious activity reporting. This model prioritizes process conformity and generates large volumes of data and filings that overwhelm public agencies. Much asset recovery still stems from traditional investigative work, not from the bulk of AML filings or automated monitoring outcomes. That suggests the prescription itself (universal, rules‑driven compliance by regulated entities) is a poor lever for the intended outcome (substantially disrupting criminals’ financial capacity).
-
Weak cross‑disciplinary engagement and group incentives that sustain the status‑quo. AML policymaking evolved largely within a specialized international community with limited integration of public policy design, evaluation science, criminology and cost–benefit thinking. That siloing, combined with powerful incentives (political signaling, industry jobs, regulatory prestige), creates resistance to acknowledging systemic failure. When policy assessments reify process metrics, practitioners are rewarded for activity, not impact.
Consequences of persisting with “more of the same”
Continuing to add rules, increase reporting requirements and punish regulated entities more severely risks worsening net societal outcomes. Costs accumulate (private compliance burdens, public enforcement expenditures, fines, and indirect harms to customers and excluded populations) while criminals continue to hold the vast majority of proceeds, enabling reinvestment in trafficking, corruption, exploitation and violence. Without structural change, the global AML experiment will remain an expensive policy theater with little effect on the social harms it purports to address.
What “fixing it” requires — a pragmatic agenda
Reorienting AML from an activity‑centric to an outcome‑centric enterprise will not be quick or easy, but the following high‑level shifts are necessary and practicable.
Define clear, evidence‑based objectives and measurable outcomes
Replace vague “protect the financial system” language with explicit, time‑bounded goals linked to crime and harm reduction: for example, measurable increases in rates of criminal asset disruption in targeted markets, increases in convictions tied to asset‑based evidence, or reductions in specific categories of criminal activity where finance is a critical enabler. Outcomes must be fit‑for‑purpose, measurable, and owned by policymakers rather than by compliance functions.
Improve and standardize data collection, validation and transparency
Governments, FIUs and law enforcement must publish harmonized metrics: amounts seized, confiscated and returned; attribution of seizures to AML‑driven intelligence versus conventional investigations; public accounting of enforcement and administrative costs; and private sector compliance expenditures. Rigorous, peer‑reviewed methodologies are essential so policymakers can meaningfully compare performance across jurisdictions and over time.
Rebalance expectations of regulated entities toward targeted intelligence value
Rather than universal, box‑ticking monitoring of all customers and transactions, systems should evolve toward risk‑based, intelligence‑led models that prioritize detection where finance materially supports criminal operations. That requires better triage of suspicious activity reporting, investment in analytic capacity at public agencies, and clearer legal pathways for regulators and law enforcement to act on high‑value intelligence.
Conduct cost–benefit and distributional impact analyses before major reforms
Every substantive AML rule should be assessed for net social benefit and for its distributional effects, including impacts on financial inclusion, small businesses and vulnerable populations. Policymakers were counseled early not to focus on costs; that must change.
Expand interdisciplinary governance and independent evaluation
FATF and national agencies should incorporate independent public policy, criminology and evaluation expertise into standard setting, mutual evaluations and strategic reviews. Third‑party, transparent evaluations of the entire AML chain — from private reporting to prosecutorial and forfeiture outcomes — should be routine.
Prioritize high‑value enforcement and asset recovery capacity
Given that asset disruption is the key mechanism to deny criminal recyclability, governments must invest in specialized financial investigations, international cooperation mechanisms that work in practice (not just on paper), and legal frameworks that expedite restraint and forfeiture where evidence supports it. Measured investments that demonstrably increase confiscation rates in targeted areas are preferable to blanket increases in compliance burdens.
Pilot and scale innovations, then evaluate for impact
Experiment systematically with new approaches — for example, enhanced public–private partnerships focused on specific crime types (drug trafficking, human trafficking, corruption), data‑sharing sandboxes, or analytic hubs combining regulated data with crime intelligence — and evaluate them against explicit outcome metrics before broad rollout.
A closing reality check
The empirical picture painted by available data is stark. Across regions and globally, authorities confiscate a minuscule share of illicit proceeds while regulated firms and citizens bear heavy costs. That combination — enormous effort and expense with almost no measurable impact on criminal finances — is a classic sign of policy design failure. A candid admission of that reality is an essential first step.
Fixing AML will require rethinking fundamentals: objectives, measurement, and incentives.
The operational infrastructure and international networks that built the AML regime are powerful assets; repurposed toward outcome‑driven priorities and informed by rigorous evaluation, they can deliver much better public value. The alternative is continued high costs, growing public cynicism, and the moral hazard of an expensive system that, in practice, leaves “Criminals, Inc.” mostly unmolested.
If the goal is to reduce the harms of serious organized crime, then policy must face hard truths, focus scarce public resources where they demonstrably work, and demand evidence of impact rather than volumes of activity. Only then can AML move from the world’s least effective policy experiment to a program that actually disrupts criminal finance and protects vulnerable people and societies.
Dive deeper
What else?
