Haiti's Remittance Fees Are a Bigger Pool Than the Gangs' Tolls, and Easier to Reclaim

One percentage point of what it costs to send money to Haiti is worth about $41 million a year, on an annual inflow of roughly $4.1 billion. The International Crisis Group estimates that the Viv Ansanm coalition takes $60–75 million a year from illegal tolls and cargo shakedowns. Cut the corridor's benchmark price by 1.5 points, from 5.4% to about 3.9%, and roughly $62 million a year stays with households, just above the bottom of the ICG's range. Cut it to 3% and the figure is about $98 million, above the top of that range. Nobody has to fight anyone to collect it. The rest of this column is the fine print.

How big is the pool, really?

Smaller than the raw arithmetic suggests. Multiply 5.4% by $4.1 billion and you get about $221 million a year. That is a ceiling, not a measurement. The 5.4% figure is the benchmark cost of sending $200, as reported from IFAD's Sending Money Home 2026. It is not a volume-weighted average across all transfers. Larger transfers cost less in percentage terms, and an unmeasured share of money moves through informal channels that no price survey records.

The inflow base needs its own caveat. The $4.1 billion comes from that single secondhand report, and the World Bank's macro poverty outlook describes strong inflow growth on a different basis. I have not reconciled the reference years. At a $3.5 billion base, each point is worth $35 million. At $4.6 billion, it is worth $46 million. Across that range, a 1.5-point cut is worth $53–69 million and a cut to 3% is worth $84–110 million. Only the low end of the conservative case falls short of the ICG range. Every other combination matches the toll haul or exceeds it.

Gross is not the same as recoverable, either. The price a sender pays bundles a sender-side fee, the foreign-exchange spread and the payout agent's commission. Since January 2026, cash-funded transfers have also carried a US federal excise tax, and I treat the 5.4% benchmark as excluding it. Some of the fee revenue stays in Haiti, with banks, agents and mobile operators that pay Haitian wages and rent. I do not have a verified split, and I will not invent one. So there are two cases, both on the $4.1 billion base:

• Conservative: a 1.5-point cut, to about 3.9%, is worth about $62 million a year. That is roughly HTG 8 billion at the BRH reference rate of 130.73 gourdes per dollar at the time of writing.
• Target: a 2.4-point cut, to 3%, is worth about $98 million a year, or roughly HTG 12.9 billion.

Both figures are illustrative arithmetic, not forecasts. PeaceBot's feasibility and cost scores for the instruments below were unavailable when this was filed, so the dollar figures are mine, not the model's.

Two warnings on how to read this. First, the comparison does not say that money transfer operators are morally equivalent to gangs. It is about the size of a pool that policy can reclaim without force. Second, the ICG figure covers only tolls and cargo extortion. It leaves out kidnapping, the pass-through to food prices and lost trade, so the gangs' full cost to Haiti is far larger.

What can Port-au-Prince change on its own?

The receiving side belongs to the Banque de la République d'Haïti, and the bank has three levers. The first is payout and FX-conversion rules. The BRH could require agents to post the dollar-gourde rate they use beside the BRH reference rate and to disclose the spread. I do not have a verified street rate, so the spread against 130.73 is a number to publish, not one to guess. Hidden conversion margins tend to be wider than visible ones. El Salvador, which uses the dollar, has no conversion layer at all. Haiti cannot copy that, but it can make its own conversion layer visible.

The second lever is interoperability. A sender who funds a transfer by card should be able to reach a MonCash wallet, a bank account or another operator's wallet without the recipient cashing out first. Today each closed loop is a small toll booth. The third lever is agent-network licensing. Exclusivity clauses between payout agents and transfer companies are widely cited as a cost driver in remittance corridors. A licensing rule that bars exclusivity, and lets any licensed agent pay out for any licensed sender, is cheap to write and costs the treasury nothing.

This is where two PeaceBot instruments would do their work. The remittance_fee_reduction instrument covers regulatory and infrastructure fixes. The digital_peace_dividend_wallet instrument moves cash-pickup customers onto account-based rails. The donor role is narrow: fund the BRH's supervisory capacity and the wallet-to-bank switching infrastructure, and nothing more.

What can Washington's rules and markets change?

The sending side is mostly American. The 2025 reconciliation law, H.R. 1 (Public Law 119-21), enacted a 1% federal excise tax on remittances. The tax applies to transfers funded in cash, by money order or by similar means. As I read the statute, transfers funded from an account at a covered financial institution, or with a US-issued debit or credit card, are exempt. Treasury and IRS guidance governs the details, so check those before relying on my summary. On a $200 transfer the tax is $2, added to a typical fee of about $10.80 at 5.4%. The sender who walks into a shop with cash pays about 6.4%. The sender with a bank card pays 5.4% or less.

That gap is the lever. Every sender who moves from cash to a card-funded rail avoids the tax and usually pays a lower fee as well. Two owners can widen that path. Treasury can say plainly whether US-issued prepaid cards qualify for the exemption, which is a matter of guidance, not legislation. The CFPB's remittance transfer rule already requires itemized price disclosure. A public, Haiti-specific comparison built on those disclosures would put pressure on fees where competition is thin.

The 3% target is not arbitrary. The UN's Sustainable Development Goal target 10.c commits governments to reduce remittance costs below 3% and to eliminate corridors costing more than 5% by 2030. At 5.4% on the benchmark transfer, Haiti sits above the 5% line. Asking governments to meet an existing commitment is more modest than making a new demand.

Why might this fail?

The strongest objection is that the senders paying the highest prices are the least able to switch. Undocumented senders, and anyone without a US bank account or the ID to open one, cannot easily move to card-funded rails. They will keep paying the tax and the higher fee, and a program built only on the exemption misses them.

What reaches these senders is the receiving side, because it does not depend on how a transfer was funded. A posted FX spread, non-exclusive agents and wallet interoperability lower costs for cash senders and card senders alike. Whether these measures are enough is an empirical question, and I would not claim that they are. The corridor average may well fall by less than the full 2.4 points, which is why the conservative case stops at 1.5.

A second risk is that Haitian banks and agents lose revenue and respond by cutting service in the provinces. Disclosure and interoperability squeeze margins without abolishing them, so they carry less of this risk than a price cap would.

The test is simple, and it should have a single owner. The BRH should publish a quarterly US-Haiti corridor average drawn from the World Bank's Remittance Prices Worldwide database, and should weight it by volume once the bank has the data. The interim mark is 3.9%, worth about $62 million a year. The target is 3%. If the number does not fall, none of the above has worked.