The Severed Artery

The shock wave of TPS on a locked-down Haitian economy.

On 27 July, a US administrative decision shook the Haitian diaspora: more than 350,000 Haitian nationals covered by Temporary Protected Status (TPS) lost their legal authorization to work in the United States. While the tragedy is first of all a human and personal one for these families, this article shows that the risk reaches far beyond them. For Haiti, it is a systemic macroeconomic threat.

The money the diaspora sends home, the famous “transfers,” is not just there to help a mother or a cousin with the shopping. These funds quietly irrigate the national foreign exchange market. They supply the whole nation with the foreign currency it needs to buy abroad what it no longer manages to produce on its own soil.

Since it is impossible to calculate in advance exactly how many dollars will stop being sent, this article is not a prediction. It is a test of resilience, what is known as a stress test, similar to the ones large banks undergo to check their soundness. We will simulate two hypothetical declines in remittances, a loss of 25% and a loss of 50%, to understand how the Haitian economy, already weakened and gridlocked, would react to such a shock. These two magnitudes are not derived from a model of migrant behavior: they are conventional bounds, chosen to bracket the range of possibilities. By way of comparison, the adverse scenario published by the IMF in May 2026 assumes a 10% fall in net remittances. The shocks simulated here are therefore severe stress tests, not a central estimate of the direct effect of the end of TPS.

1. The journey of a dollar: the two faces of remittances

To grasp Haiti’s vulnerability, one has to understand what actually happens when a member of the diaspora sends $100 to their family. The transaction acts simultaneously on two very distinct mechanisms.

The household mechanism, or purchasing power. The recipient in Haiti goes to the counter and receives the money in local currency, the gourde. This money pays for food, housing, and school.

The foreign exchange mechanism, or the supply of foreign currency. The $100 handed over to the transfer agency, Western Union or CAM, in Miami or in New York, does not vanish: it enters the Haitian financial circuit and feeds the supply of dollars available to importers.

A single transfer thus acts on two levels at once. It is the second, invisible to the person receiving the money, that makes this a national matter.

This split is not left to chance; it is governed by the regulations of the Bank of the Republic of Haiti (BRH), the central bank. Through its Circular 114-3, the BRH requires transfer agencies to sell to Haitian banks 70% of the dollars that serve as the counterpart to payments made in gourdes. Even after allowing for remittances that go straight into dollar accounts, it is estimated that 57% of all the money sent by the diaspora is channeled into the national financial system.

This foreign currency is Haiti’s war chest. It is what allows Haitian importers to pay for rice, fuel, and medicine on the world market, and it is by acquiring a share of it that the central bank rebuilds its reserves. The table below gives a measure of how overwhelmingly important this resource is.

Table 1. The weight of diaspora remittances, fiscal year 2026. Sources: International Monetary Fund, Country Report 26/107 and April 2026 WEO; Bank of the Republic of Haiti; author’s calculations. The share of GDP sets net remittances, that is, transfers received minus transfers sent in the balance-of-payments sense, against projected nominal GDP for 2026, which stands at $39.2 billion: 11.2% projected, 10.8% actual. The 20% ratios frequently cited rest on earlier and much lower GDP series, on the order of $20 to $25 billion. The gap therefore lies in the denominator, not in the measurement of remittances: it is GDP in dollars that has doubled in three years, through rebasing and exchange-rate effects.

2. The digital twin and the constrained scenario

To anticipate the shock waves from a loss of these dollars, economists use what is called a computable general equilibrium model. For teaching purposes, one can picture it as a vast digital twin of the Haitian economy. In this computer program, every sector (agriculture, industry, households, the state) is linked to the others by equations. If one resource is reduced, in this case the diaspora’s dollars, the program automatically recalculates how prices, wages, and the currency must adjust to reach a new equilibrium.

The problem with standard theory. In economics textbooks, when a country runs short of foreign currency, its currency loses value. As a result, foreign goods suddenly become very expensive. The theory then says that people start buying locally made products, which revives the country’s factories and farmers.

Haiti’s reality in 2026. That theory does not apply today. The Haitian economy is barricaded. Armed-group violence controls the national highways and isolates the major agricultural basins, such as the Artibonite Valley.

As a result, even if imported rice became unaffordable, Haitians could not turn to local rice: the trucks are not getting through. The country is physically unable to replace imports with its own production. This is what we call the constrained scenario.

In a normal, fluid economy, a 50% drop in remittances would shrink national output by barely a tenth of a percent, because local production would shift toward exports. In the constrained scenario, which corresponds to Haiti’s reality, the same drop cuts nearly 10% off the economy.

3. The statistical trap: GDP versus consumption

Here is the table of simulation results. It conceals a formidable trap for interpretation.

Table 2. The consequences of the shock in figures, constrained scenario. Results are expressed as deviations from the baseline economy calibrated on 2026, in which unemployment stands at 14.4%, a sack of rice costs 5,000 gourdes, and availability is 47.9 kg per person per year. The price of a sack of rice is a mechanical illustration: it assumes full pass-through of the depreciation, with local margins and costs unchanged, which overstates its level. Rice availability assumes that domestic production holds at 55,000 tonnes, with the decline falling on imports alone.

Why GDP hides the tragedy. If one looks only at gross domestic product, the most severe scenario produces a decline of 9.7%. An outside observer might consider that manageable. It is an optical illusion. GDP measures the wealth produced on national soil. Remittances, however, are external income that pays for purchases abroad. Cut off that money, and Haiti’s few factories keep running for a while, but families are ruined overnight.

The true measure of the crisis lies in household consumption, which collapses by more than 22%. Put simply, a family would lose the equivalent of nearly three months of consumption over a year.

A nonlinear response. The effects are not proportional to the cause. When the shock doubles, from a 25% drop to a 50% drop, several consequences more than double: for GDP, the effect is multiplied by nearly 2.5; for consumption, income, and the exchange rate, the factor is closer to 2.3 to 2.4. Why? Because the scarcer dollars become, the more value the gourde loses, and the more inflation soars on basic goods that cannot be replaced. The economy is already cut to the bone: today, of every gourde a household spends, 21 centimes go straight abroad to pay for the imported product. People have already cut out everything superfluous. The next shock will strike directly at survival: rice, wheat, and medicine.

4. The crisis in daily life: unemployment and desperate trade-offs

How does this macroeconomic decline play out in everyday life? Mainly through job losses.

The economy obeys the law of the wage curve: when a firm faces a massive crisis, it is legally and socially very difficult for it to cut the wages of all its employees in half. Instead, the real wage falls a little, because prices rise, but the firm adjusts through mass layoffs. That is why unemployment would come close to 30%. The sectors that live off the movement of imported goods would be devastated.

Table 3. Where the shock hits: change in value added by sector.

Faced with this fall in income, households make what we call the destitution trade-off. To keep a minimum of food on the plate, they cut savagely into every other expense.

Table 4. Family sacrifices, severe scenario (minus 50%).

Non-food spending, which includes school, health, and housing, would fall by nearly 28%. Despite this squeeze, the quantities of food consumed would still fall by nearly 15%, the equivalent of almost two months of food. Rice falls further, by nearly a quarter, because it is almost entirely imported, whereas the average food basket still contains a share of local production.

5. The illusions: textiles and the providential diaspora

Table 2 shows a surprising figure: Haiti’s exports would rise by 30%. How is that possible in a gridlocked country?

The explanation lies in the textile assembly plants, such as those in Ouanaminthe. These factories sell their T-shirts to Americans in dollars but pay their workers in gourdes. If the gourde loses half its value against the dollar, the factories automatically grow richer and can hire.

But the illusion is twofold. First, the sector is far too small, 2% of GDP, to save the economy. Second, it depends entirely on a US trade arrangement, the HOPE and HELP Acts. These preferences expired once already, on 30 September 2025, before being restored retroactively through 31 December 2026. A bill passed by the House of Representatives, the Haiti Economic Lift Program Extension Act (H.R. 6504), would extend them through 31 December 2028, but it is still awaiting consideration in the Senate. Another bill, H.R. 5209, introduced but not passed, contemplates an extension through 2037.

The other illusion is to believe that the diaspora will find a way to make up the difference. Part of the literature shows that migrants send more money when a disaster, such as an earthquake, strikes their country of origin. But here, the TPS shock hits the migrants themselves, in the United States. You cannot ask someone who is losing the right to work legally to send more money.

6. The boomerang effect: American farmers hit

A globalized economy works both ways. Haiti no longer produces enough and has turned into a huge import market. Among other things, the country is the world’s third-largest buyer of American rice. If Haitians suddenly grow poorer, they will stop buying.

Table 5. The threat to US agricultural exports. Sources: USDA Foreign Agricultural Service, “U.S. Trade with Haiti in 2024,” for the first three rows; author’s calculations for the last. Lost sales are valued at the average 2024 unit price, about $755 a tonne, at constant prices.

A loss of 90,000 tonnes in sales would come as a direct blow to the farmers of the Arkansas and Mississippi valleys, whose operations are already at the limit of profitability. Set against the $2.4 billion in US rice exports, the loss remains modest: less than 3%. But it is concentrated in a handful of producing states, which makes it an argument about domestic politics far more than about foreign trade.

7. The four absolute priorities for the state

In the face of this imminent danger, public policy must concentrate on four vital pillars.

1. Security and logistics

This is the absolute economic priority. As long as the roads are in the hands of armed groups, no substitution by local production is possible.

2. Rigorous management of reserves

The central bank will need to preserve its strategic reserve buffer and avoid artificially defending a parity that has become unsustainable. In the event of a severe shortage, any prioritization of foreign currency for essential imports will have to follow transparent rules, so as to limit rents and distortions.

3. A targeted social shield

The state must protect the most vulnerable by securing school canteens and by reducing the fees charged on the remittances that still come in.

4. Active economic diplomacy

Washington must be persuaded to renew the HOPE and HELP trade laws before 31 December 2026, the date of their next expiry, using the argument of the losses facing American rice farmers.

Technical appendix: the engine room

This section is for readers who want to understand the mathematical logic of the model. The computer model solves 45 equations simultaneously. It works in comparative statics: it compares two equilibrium states without describing the path that leads from one to the other, so the deviations reported here are differences between two snapshots, not changes over time. Here are the six fundamental relationships, explained simply.

Two limitations must be flagged at the outset. The detailed internal structure of the economy is still that of 2013, since no supply-and-use table has been published since. And beyond a 50% drop in remittances, the model stops converging: this is a direct consequence of the capped elasticities of the constrained scenario, documented as such, and not an economic result.

Architecture. The model takes up the general equilibrium framework of Cicowiez and Filippo (twenty-two activities and commodities, five factors), fully reimplemented by the author and aggregated to twenty industries. It is built on a 2013 social accounting matrix for Haiti, updated to the International Monetary Fund’s 2026 aggregates by biproportional scaling. The closure rule fixes foreign savings and government consumption: all external adjustment is therefore shifted onto the real exchange rate, which weighs on the results as much as the elasticities themselves. The model replicates the original matrix to machine precision, with the forty-fifth equation, redundant by Walras’s law, holding to fifteen decimal places, and the updated baseline solves with residuals on the order of 10⁻¹², for unemployment of 14.4%. All results are measured against this solved baseline. This numerical validation guarantees the internal consistency of the model, not the economic certainty of its results, which remain conditional on the parameters and closure rules chosen.

Household income

It is the sum of wages (YF), assistance (CPI), and, above all, diaspora remittances (trrow) multiplied by the exchange rate (EXR).

The external constraint

This is the golden rule. The value of imports (QM) cannot exceed the sum of exports (QE), remittances received, and foreign savings (FSAV). If remittances fall, imports must mechanically decline.

The trade-off between local production and exports

An equation that determines whether producers prefer to sell at home or to export. The letter sigma represents flexibility. For Haiti, we have locked it at 0.2, which is very rigid, because of insecurity.

Substitution between imported and local goods

How consumers respond to prices. Here again, since local foodstuffs cannot be transported, flexibility is set very low, at 0.25 for food.

The wage curve

This equation captures the fact, explained above, that wages (W) hold up in the crisis, so that it is unemployment (U) that absorbs the shock, through layoffs.

The survival instinct, or the Stone-Geary equation

A mathematical formula that forces the virtual households to buy a vital minimum of food before they can spend the rest of their budget on anything else.

The parameters used

In economics, an elasticity measures how a behavior adjusts to a change. Each parameter falls into one of four categories: observed, calculated, drawn from the literature, or an explicit assumption.

Table 6. The behavioral parameters used, known as elasticities.

Three readings of the same shock

Table 7. Change in real GDP depending on how much freedom the economy is given to adjust. The gap between the last two rows defines a conditional interval: it is an analysis of sensitivity to the capacity to adjust, not a probability range. If the standard variant does not worsen when the shock doubles, it is because the real depreciation there triggers an export boom that grows with the shock and offsets most of it, precisely the mechanism that the current blockages rule out.

Adding further crises

Table 8. The severe scenario (minus 50%) serves as the baseline; the next two columns each add a crisis to it. The model tested the addition of two other potential misfortunes to see how the system reacts. The supplementary assumption of forced returns, amounting to 3% more people in the labor force, comes against a backdrop in which the International Organization for Migration (IOM), the National Migration Office (ONM), and the Support Group for Repatriated Persons and Refugees (GARR) recorded some 412,000 returns between January 2025 and June 2026. Nearly all of them, however, came from the Dominican Republic: this series measures the general pressure on the labor market, not the US deportations that might follow from the end of TPS.

One result deserves to be singled out: real investment falls by 1 to 4%, depending on the scenario. This decline signals a longer-term risk, since in a dynamic framework slower capital accumulation would weigh on future productive capacity. The static model used here, however, can neither date nor quantify this effect.

Sources

International Monetary Fund (Country Report 26/107 and April 2026 WEO), World Bank, Bank of the Republic of Haiti, Department of Homeland Security (Federal Register notice of 28 November 2025), IOM, ONM, and GARR, IPC, USDA. Data on the financial system are public aggregates; no data from any individual institution are used.