REMITTANCES TO CUBA REVISITED: IMPACT OF NEW MEASURES[1]
Luis R. Luis
As many other developing nations Cuba depends on personal remittances as a sizable source of foreign exchange and support for families and small enterprises in the informal economy. This does not escape the government of Cuba which tries to exert control over these flows of money. Cuba is setting up special hard currency stores for appliances and electronic equipment as well as a complementary system for channeling foreign cash through its banking system. The United States in turn changed regulations in September 2019 setting a limit on family remittances and banning donations to persons other than close relatives.
Cuba does not publish data on remittances in its balance of payments. Central banks in most countries report remittances as part of receipts in the current account of the balance of payments. Remittances come through both formal and informal channels. Formal channels include money transmitters and international banks. These are registered by central banks. Other remittances come through unregistered firms and personal delivery. Since there is no official public Cuban data on remittances estimates are derived from source country statistics. I attempt to shed light on transfers to Cuba from the U.S. using a model of remittances estimated from U.S. and international data.
DETERMINANTS OF REMITTANCES
Remittances from a source country are basically driven by the income of immigrants in the source country and their links to families and friends. There are other elements relating to payments channels and restrictions that may exist in destination countries. These economic foundations allow fashioning and estimating a remittance equation. Data on U.S. transfers to persons in 20 developing countries in Latin America, Eastern Europe and Asia enable estimation. Remittances data is compiled by The World Bank. These statistics in turn come from the IMF as reported by central banks and in national balance of payment statistics. The data comes mainly from two items in balance of payments reporting: 1) personal transfers and 2) compensation of employees. The first contains estimates for informal transfers and may be biased downwards given incomplete coverage of flows. Compensation of employees is income derived from temporary employment in source countries and all of it is reported as remittances in World Bank estimates. These are biased upwards as a large proportion of earnings remains in the source country. We need to keep in mind these opposing biases in the data in evaluating remittance flows.
A remittance equation is estimated from a cross-section sample of 20 developing countries in three continents. Data on remittances and migration are from the World Bank (2017), exchange controls from the IMF (2018), international banking assets from the BIS (2018). Data on characteristics of immigrant groups are from the U.S. Census Bureau (2011, 2017).
ESTIMATED REMITTANCE EQUATION
R = .784 + .076M7*** + .106POV* – .001LIQ – 1.973XC*** + 1.354REG** (t) (.656) (3.165) (1.935) (-1.287) (-2.643) (2.257)
R2 = .606 SE = 1.488 F = 4.008
R = Remittances per immigrant to country of origin
M7 = Proportion of immigrants arriving in the last 7 years
POV = Proportion of immigrants below the poverty line in the U.S.
LIQ = Cuban assets in international banks per capita
XC = Variable for exchange controls in destination country
REG = Regional variable for Latin America, Eastern Europe and Asia
* Significant at 90% level ** Significant at 95% level *** Significant at 98% level
The dependent variable is remittances per U.S. immigrant to the home country. The main explanatory variable is the ratio of immigrants in the last 7 years to the immigrant population from each country. Recent immigrants send remittances at a higher rate than earlier ones. As there is no consistent income data for immigrant groups, I use a poverty ratio from the U.S, Census as an indicator of relative income. The estimated parameter for this variable has a positive sign – poorer immigrant groups send more remittances per person – which is a puzzle. Additional statistical analysis using education levels as a proxy for income show a negative relationship – more highly educated immigrants send fewer remittances. There are a number of possible explanations for these puzzling but consistent results. One hypothesis is high correlation of income levels within families with better-off relatives receiving smaller transfers.
A variable for exchange controls, which takes values of 0, 1 and 2, represents the restraining influence of controls on family remittances. There is also a measurement problem as controls drive remittances into informal channels which are under-reported. The regional variable represents higher remittances from Eastern European and Asian immigrants versus those from Latin America. Lastly the variable representing international liquidity was not significantly different from zero. Immigrants are seemingly unaffected by foreign cash conditions in their home country as an element in the decision to send family transfers.
The equation is used to provide an estimate for remittances to Cuba from the United States. Estimated annual remittance per immigrant is $3256 in 2017, slightly less than the average remittance of the 20-country sample. Total flow to Cuba is $4141 million including both cash and in-kind transfers. These estimates are subject to wide confidence intervals and need to be used with caution. A one standard error interval (68% probability) places U.S. remittances between $2248 million and $6034 million. Emilio Morales (2018, 2019) estimates 2017 cash remittances at $3575 million and total remittances, including in-kind, at $6565 million. The estimates from the model refer to U.S. remittances, both cash and in-kind. Cuban-born immigrants in the U.S. account for around 85% of the world total and for more than 90% of household income. This suggests the model estimate underestimates worldwide remittances by 5% to 10%.
IMPACT OF NEW MEASURES IN THE U.S. AND CUBA
The U.S. Treasury (2019) announced in September 2019 a limit of $1,000 per person per quarter for family remittances to a single close relative in Cuba. The average remittance estimated above falls below the new annual limit by about $750. However active remittance senders are a fraction of the overall immigrant population, and the average effective remittance is well above the new limit set by U.S. Treasury regulations. Hence the new regulation will have a negative impact on the flow of family remittances.
The new rules announced by the U.S. Treasury include further authorization of “remittances that encourage the development and operation of private businesses by self-employed individuals.” This will induce new money to cuentapropistas and other self-employed persons.
Cuba’s cash crunch worsened in the last two years by the collapse of the Venezuelan economy and subsequent curtailment of oil shipments to the island. New measures by the Trump administration will also hit tourism with the abrogation of cruise ship sailings and reduction in flights from the United States. All of this is moving the Cuban government to seek new ways to capture for the state some of the family transfers.
New Cuban measures (Granma 2019) aim at channeling family remittances through the Cuban banking system and special stores selling valuable products purchased via dollar accounts in Cuban banks. In this scheme remittance money is transferred directly to the bank accounts which in turn can be used via magnetic cards in the hard-currency stores. These stores also aim to divert some in-kind remittances by offering similar products to those shipped by relatives from the U.S. and other places.
It is early to assess the quantitative impact of the new scheme. The government aims at capturing a high percentage of remittances. Independent reports that the dollar has risen significantly in the Cuban parallel market versus the convertible peso (CUC) shows there is demand for dollars to fund the new bank accounts. However, it is likely that the new accounts will be funded carefully by persons with remittances. Cuba’s long history of restrictions, freezes and confiscation of hard-currency accounts argues against a sizable capture of transfers by state banks. Avoidance of the hard currency stores and bank accounts will rise if prices in the stores are maintained well above those of comparable items in neighboring countries.
The government expects to benefit both from additional dollar liquidity in the banking system and from selling high margin products in state stores. There is an inherent contradiction in this system implying disappointing returns for the government from this scheme. Additional restrictions such as forcing family remittances into the new dollar accounts will displace transfers towards informal channels.
DATA SOURCES AND REFERENCES
Bank for International Settlements 2018, Quarterly Review, Basel, March.
Global Migration Group 2017, “Handbook for Improving the Production and Use of Migration Data for Development,” Global Knowledge Partnership for Migration and Development. The World Bank, Washington D.C., pp. 68-72.
Granma 2019, “Gobierno Cubano informa nuevas medidas económicas a través de la mesa redonda,” Granma.cu, Noviembre 8, 2019.
International Monetary Fund 2018, Annual Report on Exchange Arrangements and Exchange Restrictions. Washington.
Morales, Emilio 2019, “Remesas, ¿una ruta de inversión para los cubanos?” The Havana Consulting Group, September 17.
Morales, Emilio 2018, “Remittances to Cuba diversify and heat up the payment channels.” The Havana Consulting Group, March 3.
Remittances, migrationdataportal.org.
The World Bank 2017, Migration and Remittances Data, Bilateral Remittance Matrix 2017, worldbank.org.
The World Bank 2016, “Migration and Remittances: Recent Developments and Outlook.” Washington, April.
U.S. Census Bureau 2011, Current Population Survey.
U.S. Census Bureau 2017, “The Hispanic Population in the U.S.” Washington.
U.S. Department of the Treasury 2019, Office of Foreign Assets Control: Frequently asked questions related to Cuba, September 6, pp. 14-15.
[1] This note benefitted from comments on an earlier version by Ernesto Hernández-Catá.