According to International Monetary Fund (IMF) data, personal remittance volumes to Haiti exhibited a strong year-on-year uptrend in the decade following the 2010 earthquake, as measured by the percentage of Gross Domestic Product (GDP), rising from approximately 11.7% in 2012 to 23.1% in 2020 (World Bank, 2021a). According to the World Bank (2022), remittances to Lowand MiddleIncome Countries (LMICs) have outpaced the foreign direct investment (FDI) since 2016. and the official development assistance (ODA) has been recorded at three times for at least a decade. Since 2015, remittances to the LAC region have grown at least 6.7% per year, reaching an estimated 25.3% in 2021 (ibid). In 2023, remittances to the region are expected to reach an estimated USD 153 billion. According to the World Bank (2020), in 2018, remittances (USD 3.1 billion) significantly outpaced both ODA (USD 992 million) and FDI (USD 105 million). In 2020, FDI decreased further by 12% due to a decline in global economic activity, while remittances proved resilient, increasing by just under 1% globally that year.

In April 2020, the World Bank´s 35th Migration & Development Brief (2021c), which focuses on the impacts of COVID-19 on the global remittance economy, projected that remittances to the LAC region would fall by 19.3%; instead, in that year, it increased by 9.6% (World Bank 2020; 2021a; 2021b; 2021c). In fact, remittance flows to LMICs in the LAC region were among the most resilient in 2020 and 2021, growing 6.2% and 21.6%, respectively (World Bank, 2021c). According to the World Bank, these increases are related to several factors, including the US stimulus package and economic recovery in 2021; increases in transit and stranded migrants in the region, including Haitians, and an increase in remittances they receive; increases in remittance flows due to disaster events (e.g., hurricanes and earthquakes) in countries of origin; a depreciation of local currencies against the US dollar; and a marked increase in digital nomads working remotely from other countries in the region (World Bank, 2021c). Remittances to Haiti have remained relatively steady between 2019 and 2021, moderately decreasing from USD 3.2 billion to USD 3.1 billion during this period (World Bank, 2021b). Globally - as in the LAC region - remittances are projected to grow further in 2022 (World Bank, 2022).

While global remittance amounts continue to outpace ODA and FDI, their impacts have been greatly debated. Some argue that remittances do not foster economic development but encourage unnecessary consumption rather than providing necessities (Jahjah et al., 2003). In line with this pessimistic view, Clément’s (2011) microeconomic study of remittances in Tajikistan found their use unproductive at the household level (Clément, 2011). Yang (2011), however, argues that remittances used for consumption expenditures should be carefully evaluated - as they may be optimal for poor households with lower consumption levels. The more positive view of remittance spending suggests that they are critical to household survival and, in some cases, cover school fees, improve livelihoods through savings, and facilitate investments (Adams & Cuecuecha, 2010; Cardona Sosa & Medina, 2006). Further, remittances have been credited with helping poor households mitigate against and recover from disasters (Mohapatra et al., 2009).

Nowhere is this research more relevant than in the Latin America and Caribbean (LAC) region, where the share of international migrants and remittance transfers sent home is significant compared to other regions (Ratha & Xu, 2008). Many studies have therefore sought to determine the impact of remittances in the LAC. In Guatemala, Adams & Cuecuecha (2010) found that remittances increased spending on education and housing. Mohapatra et al. (2009) report that remittances help families in disaster preparedness and recovery. In Haiti, Bredl (2011) found that remittances positively impacted educational outcomes among low-income families. However, Amuedo-Dorantes et al.’s (2010) research in Haiti found that although there were gains in human capital development, as in school attendance, the disruptive effects of family out-migration mitigated these gains. Despite these concerns, in the Haitian context, as elsewhere, remittances have proven critical in mitigating the effects of political, economic, and social upheaval, as well as disasters (INURED, 2020a).

The extant literature on the role of remittances in times of crisis in the homeland presupposes that international migrants are not directly affected by the same crisis, given their physical distance from it (Wendelbo et al., 2016; Yang & Choi, 2007). The COVID-19 pandemic and resultant economic shocks are clearly different, with global, cross-cutting impacts. Unlike localized crises (such as civil war, natural disasters, and famine, among others), the pandemic has affected every node in the transnational schema, including migrants in host nations and their families in the homeland.

In host nations, migrant populations have borne the brunt of the socioeconomic burdens of the pandemic (Freier, 2020; Martin & Ferris, 2017; Orozco, 2017). In the US, for example, employment fell by 9.8% for foreign-born workers between 2019 and 2020, compared to 5.4% for the native-born population. Data from the Brazilian Ministry of Labor show that the percentage of registered Haitian workers hired and terminated in the country was 3.3% and 3.5% in 2019 and 2020, respectively. However, this belies the fact that Haitians are among the lowest-paid migrant workers in Brazil’s formal economy (Observatório das migrações internacionais, 2020), and that they are overrepresented in the informal economy. Given their marginal status, often serving as agro-industrial, lowwage, and/or low-skilled laborers in the destination country, their vulnerability is likely to have been compounded by the pandemic. These factors certainly impact their ability to remit.

With few exceptions (Ratha & Shaw, 2007), studies of international remittances generally privilege transfers from countries in the Global North, often neglecting those from the Global South, despite the fact that by 2005, it was estimated that South-South Migration (SSM) accounted for almost half (47%) of all global migration, or at least 74 million migrants. More recent arguments suggest that by 2015, SSM “constitute(d) the bulk of international migration” (Short et al., 2020). Remittances have, in many ways, served as the cornerstone of studies on migration impacts on the economies and development of countries of origin, particularly in the Global South. However, shifts in migratory trends suggest a need for studies of remittances in the context of SSM. Ambrosius et al. (2020) contend that SSM flows account for one-third of remittances sent globally. Over the past decade, Brazil and Chile have emerged as prominent destinations for Haitians migrating within the Global South. However, to our knowledge, no notable studies on the impact of these new migration flows on Haiti’s remittance economy have been published.

Haiti provides a unique opportunity for further examination of the impact of SSM on the remittance economy, not only due to its history of migration within the Global South but also in light of shifting geopolitical realities and regional labor demands that have redirected post-earthquake migration toward South America. Brazil and Chile’s prominence in Haiti’s new migration landscape are key to understanding the nation’s remittance flows over the past decade, a period in which the country has experienced several major disaster events (ASFC, 2019; Marcelin & Cela, 2017a; Marcelin et al., 2016; Paz, 2021) and significant political, economic, and social upheaval (INURED, 2020b; Marcelin & Cela, 2017b).

This article draws on data collected during a study on the impact of COVID-19 on Haitian households implemented during the summer of 2020 and remittance data obtained from Haiti’s Central Bank in order to answer our central research question: What role have remittances from South America played in Haitian households leading up to and during the first wave of COVID-19 pandemic? We begin by describing the methods employed in the study; then, we present the results from the Haiti Central Bank data and COVID-19 study, followed by a brief discussion and conclusion. The data reveal that Haitian migrant commitments to the homeland - as evidenced by remittance transfers - are vital to household survival and remain constant, particularly during times of crisis.

Publicly available data on remittance transfer volumes to Haiti is problematic for four reasons: 1) the IMF and the World Bank only provide quarterly estimates; 2) official BOP statistics ³ are known to underestimate remittances; 3) the most recent nationwide household survey ⁴ contains indicators on migration and remittances but does not fully capture post-earthquake migratory destinations that have emerged over the decade; and 4) such data fail to account for informal, though critical, transfer channels.

To overcome these challenges, the research team acquired access to monthly remittance transfer data covering a specific time within the overall study period: from October 2017 to August 2020. Although the series is small (n=35 months), it contains data for the period leading up to, during, and immediately following the first wave of the COVID-19 pandemic. Further, it disaggregates remittance amounts by source country. Due to the sample size, the analysis of the impact of COVID-19 on households in Haiti is limited and not generalizable; however, it does provide insights that may help orient future studies of migration and migrant commitments to the origin country in times of crisis.

According to BRH data, between October 2017 and August 2020, remittance transfers were relatively constant, with March 2019 and December 2019 serving as the peak transfer months and a notable decline in March 2020 when the pandemic was officially declared in Haiti. Following a dip in transfers in April 2020, remittance volumes to Haiti peaked from May to August 2020, recording the highest amounts transferred during the 35-month period (Figure 1 ). The decline in remittances during the pandemic was not significant, as 2019 had already seen sharp declines due to the ongoing political crisis in Haiti referred to as Peyi Lok ⁵ . During the five months that exhibited a negative year-on-year change in the series, October 2019 and March 2020 were the lowest, at -8.2% for each month, compared to the same month of the previous year. By May 2020, year-on-year increases were positive and in the double digits until the last month of available data (August 2020), with June 2020 representing an increase of more than 50% over June 2019.

Figure 1 Source: Authors’ calculations based on Banque de la République d’Haïti data.

Figure 1 shows the six-month simple moving average (SMA) as an indicator of the general upward trend. From May 2019 to November 2019, all monthly remittance volumes to Haiti were below the six-month SMA, indicating a weakening trend through 2019. Thus, when examining remittance flows to Haiti during the first phase of the pandemic, it is essential to acknowledge the sharp decline in transfers in the latter part of 2019. Except for January, all months in 2020 exhibit a higher transfer volume than the six-month SMA, and a significant upward trend is evident, especially when considering the months between May 2020 and August 2020.

Figure 2: Source: Authors’ calculations based on Banque de la République d’Haïti data.

Figure 2 shows the total number of transfers received and their USD average amount. In April 2020, while the total number of transfers decreased significantly, the average transfer amount increased approximately 2.5 percent from USD 151.17 to USD 154.95. COVID-19-related public health measures (e.g., lockdowns, stay-at-home orders, business closures) adopted in major sending countries, particularly the US and Canada, seemingly disrupted money transfer services. Therefore, migrants may have made up for these disruptions by sending larger amounts. A stepwise increase in average transfer amounts occurred in the first six months of 2020, from USD 141.93 in January to USD 171.43 in June. With the maximum in December 2019 reaching USD 214.24, however, it seems that the recovery in remittance volumes was largely sustained by an increase in number of transfers rather than their amount. The extent of the recovery in terms of the total number of transfers is remarkable, indicating the engagement of remittance senders during the pandemic.

The study of COVID-19’s impact on households in Haiti allows for a unique comparison of differential remittance flows from countries in the Global North and the Global South. Firstly, the central bank data are disaggregated by source country, and secondly, the study sample contains data on the country of migration (source country), remittance amounts, and basic spending indicators, as well as other relevant socio-demographic information. Here we provide the results of the descriptive analysis of both datasets, differentiating between source countries in the Global North and Global South. We pay special attention to the first wave of the COVID-19 pandemic, providing a primary analysis of the impacts of the global crisis on remittance flows to Haiti.

For the study period, remittance inflows from countries in the Global North differed greatly from those of the Global South. Previous work (OECD & INURED, 2017; Orozco, 2006) has highlighted the importance of remittances from the Global North, particularly the US, where most Haitian migrants reside. To our knowledge, there has been no prior research in Haiti explicitly analyzing differences in remittance flows and impacts from countries in the Global South. Further, as many studies have relied on the Haiti Living Conditions Surveys from 2001 or 2012 (e.g., Jadotte, 2009; Justesen & Verner, 2007), they do not capture contemporary surges of post-earthquake migration to the Global South. The present study contributes to the growing literature on SSM and remittances by outlining differences in household-level outcomes by country of destination.

Monthly remittance transfer data by country between October 2017 and August 2020 suggest that although the US still accounts for between 70% and 84% of all remittances to Haiti (see Figure 3), more recent destination countries such as Brazil and Chile have emerged as newer, yet still critical, remittance sources. For example, in August 2020, Chile accounted for 8% (USD 22 million) of total remittance volumes to Haiti, almost equivalent to the total combined contributions of Canada, France, and Brazil.

Remittances from countries in the Global North, including the US, Canada, and France, tend to peak in March in thetime series (See Figure 3). In 2020, however, this peak occurred in April 2020. Destination countries in the Global North accounted for 89% of remittance flows as per the central bank data in April 2020. The US alone contributed USD 238 million, amounting to 83% of remittance flows to Haiti that month, despite measures adopted to curb the spread of COVID-19, such as closing businesses and schools and restricting population movements. These measures were not equally adopted for major remitting countries in South America, including Brazil and Chile.

Figure 3: Source: Authors’ calculations based on Banque de la République d’Haïti data.

Although Brazil was the first South American country to emerge as a major migration destination for Haitians, remittance volumes from Brazil were lower than those from Chile in all 35 months of the BRH time series. On average, remittance inflows from Chile were 165% greater than those from Brazil, while transfers from Brazil appeared more stable than those from Chile. Remittances from Chile exhibit a general downtrend as 20 out of 23 months exhibited yearon-year decreases and greater variance, while remittances from Brazil remained steadier, with only 10 out of 23 months exhibiting decreases. Furthermore, during the peak of the first wave of the pandemic, from March 2020 to May 2020, when Haiti instituted various lockdown policies, including closing its airports and borders, remittance volumes from Chile experienced much larger decreases than those from Brazil as year-on-year changes. The recovery in Chile was only seen in August 2020, while Brazil showed year-on-year increases between June 2020 and August 2020. This may indicate Chile’s waning status as a destination country for Haitian migrants. Moreover, we hypothesize that as Brazil emerged first, individuals’ time in a country may be longer on average, allowing for more integration and, thus, greater financial security than that experienced by migrants in Chile ⁶ .

The Haitian central bank dataset is also not without its limitations. One limitation is the inability to capture remittance transfers made through informal channels, such as person-to-person delivery or intermediary networks. In this regard, household-level data may provide important insights into informal exchange patterns in Haiti’s remittance economy.

The survey of 511 households in five administrative departments of Haiti comprised 361 (71%) femaleand 148 (29%) male-headed households. More than half (51%, 259) of all respondents were between 32 and 45 years of age, 24% (124) were aged 18 to 31, while 21% (106) were between 46 and 59 years old, and 4% (22) 60 years or older. This is consistent with Haiti’s youthful demographic profile and with World Bank (2019) estimates that 62% of the Haitian population is between the ages of 15 and 64. Eighty percent (410) of households were urban, and 20% (101) were rural. More than half (56%, or 286) had between four and six family members living in the home; 19% (97) had one to three members; 19% (98) between seven and nine members; and 6% (30) 10 or more. Sixty-nine percent (353) of participants reported having one to three offspring, 22% (114) four to six; 4% (20) between seven and nine, and 1% (7) 10 or more children. Before the pandemic, 46% of households (236) were engaged in petty commerce, 26% (135) were engaged in wage labor, 17% (85) were entrepreneurs, 6% (33) were inactive but not seeking formal employment, and 4% (21) were unemployed.

More than half (52%, or 265) of all households reported having no family or close friends living outside of Haiti, 27% (137) had only a close friend abroad, and 21% (109) were migrant households, defined here as having a family member living abroad. Among migrant households, 64% (70 of 109) reported only having a family member living abroad, and the remaining 36% had at least one family member and a close friend living abroad. Among non-migrant households, more than one-third (34%, or 137) had only a close friend living abroad.

A significant number of study respondents indicated having some university education or having completed all university-level studies at 7% (34) and 8% (43), respectively ⁷ . Over one-third of participants (39%, or 197) reported having some secondary school education, 16% (83) cited having completed secondary school, and 6% (29) reported no formal schooling. For migrant households in the sample (n=109), the proportions of higher education were slightly greater than the general sample at 10% (11) for some university education and 9% (10) for all completed university-level studies. Migrant households with close friends abroad reported higher education levels on average than all other categories, with just under one-quarter (23%, or 9 of 39) having completed some or all university-level studies. Almost one-fifth (17%, or 27 of 137) of migrant households with no close friends abroad had completed some or all university studies. Non-migrant households were among the least educated, with almost 10% (25 of 264) having no formal schooling. It has been shown that migrants from Haiti are “positively selected” when it comes to socio-economic status and exhibit a greater propensity to migrate as per capita income and educational attainment increase (Orozco, 2006) ⁸ .

Households with migrants in the Global North exhibit higher levels of educational attainment on average than those with migrants in the Global South. Slightly more than one-quarter (26%, or 14 out of 54) of households with migrants in the Global North reported having completed some or all universitylevel studies, whereas only 13% (7 out of 55) of households with migrants in South America reported the same. In addition, 12 household heads (24%) with migrants in South America reported having had no formal schooling or some primary school. This is the case for only 3 (6%) heads of households with migrants in the Global North, all of whom reported at least some primary school. Notably, there were no cases of no formal schooling for household heads with migrants in the Global North. These findings align with previous works (INURED, 2020b; Schlabach, 2020) showing the proliferation of unskilled and semi-skilled laborers migrating to Brazil, recruited as workforce for the 2014 World Cup and 2016 Olympic Games.

These statistics show the importance of migration capital in Haiti for migrant and non-migrant households. Furthermore, households with more than one migrant abroad (one household reported having as many as eight) proportionally received higher numbers of remittances. Eighty-six percent (86%, or 31 of 36) of households with two or more migrants abroad reported receiving remittances in the previous 24 months. This figure jumps to 95% (18 out of 19) when considering households with three or more migrants.

Consistent with the analysis of Haiti’s central bank data, we find stark differences in remittance receipt per North or South migration corridor. Eighty-three percent (83%) of households with migrants in the Global North reported receiving remittances in the previous 24 months, while the proportion for those in the Global South was only 64%. When asked about the amounts of remittance transfers received between March 2020 and May 2020, respondents with migrants in the Global North reported higher amounts than those with migrants in the Global South. Well over half (59%) of respondents associated with a migrant(s) in the Global North reported receiving more than USD 100 as compared to 37% with a migrant(s) in the Global South. Forty-one percent (41%) of non-migrant households reported receiving more than USD 100 from a close friend between March 2020 and May 2020. More than twice as many (59, or 69%) households with close friends living abroad that reported receiving remittances within 24 months of the survey had migrants in countries of the Global North versus 26 (31%) with close friends sending remittances from the Global South. Notably, 54 of the 59 households receiving remittances from the Global North were linked to the United States.

For remittance transfer amounts below USD 100, proportions for households with migrants in the Global South were almost double (63%) that of the Global North (35%), indicating that households with migrants in the Global South received lower amounts. According to this indicator, even non-migrant households receiving remittances from close friends in the Global North fared better than migrant households with family members in the Global South. The vast differences in remittance transfer amounts received from the Global North and South at the household level are consistent with Haiti’s central bank data reporting.

The vast majority (88%, or 94 of 107) of COVID-19-related remittance disruptions were linked to livelihood activities such as employment loss or disruption. Eleven percent of respondents cited transfer service interruptions either in Haiti or the source country as the reason for disruptions. The extent or duration of these disruptions is unknown. However, in households that had received remittance transfers within 24 months of the survey, only 59% (85 of 145) reported receiving transfers early on after the official declaration of a pandemic and lockdown period between March 2020 and May 2020 in Haiti. As expected, of these households, there appeared to be a positive relationship between the amount received and the number of migrants a household had abroad, including close friends. This may suggest that, in times of crisis, households not typically receiving remittance transfers may activate migrant networks that are otherwise inactive. Interestingly, a few (8) households reported sending, rather than receiving, remittances during this period. Among these households, 75% (6) had migrants in the Global South: 2 in Brazil, 2 in Chile, and 1 in Ecuador.

Figure 4: Source: Authors’ calculations based on INURED household survey data.

Out of 109 migrant households, there were nearly equal distributions of family members in the Global North (54) and South (55). More than three-quarters (78%, or 43 of 55) of migrants in the Global South were in South America, overwhelmingly Brazil and Chile, accurately reflecting changes in Haitian migration destinations in the post-2010 earthquake period. Of the 39 migrant households with close friends abroad, households with migrants in South America tended to have more close friends in the US and Canada than in these regions. This may indicate recent trends in onward migration from South America to the United States ^{9 , 10} . Of the 19 households that had migrants in continental North America and a close friend abroad, 13 reported the close friend living in North America and six in the Caribbean. Of the 14 households with migrants in South America, six had a close friend in the US or Canada, four in the Caribbean, and the remaining four in South America. Thus, even for migrant households participating in the new SSM routes, the North American and Caribbean corridors remain a crucial dimension to their transnational composition.

Rural migrant households tended to migrate to destination countries in the Global South, particularly Brazil and Chile. Of the 24 rural migrant households, 15 (63%) were in South America, and six (25%) were in continental North America. In contrast, of the 85 urban migrant households, more than half (52%, or 44) were in continental North America and only one-third (28) in South America. These data are consistent with historic migration trends in Haiti in which those from rural parts of the country migrate within the Global South, whereas urban dwellers tend to migrate to the Global North ¹¹ . In an examination of household transnational status vis-à-vis gender, there was a higher proportion of female-headed households with migrants in the Global South (42 out of 55; 76%), as compared to the Global North (30 out of 50; 56%). It is known that most Haitian migrants in Brazil and Chile are male, which may contribute to these differences.

Regarding migration within the Caribbean, 8 out of 11 households had a migrant in the Dominican Republic (DR) at the time of the survey. The other three households reported having a migrant in Turks and Caicos (1) and the Bahamas (2), and all three reported receiving remittances. Only one-quarter (2 of 8) of households with a family member in the DR reported receiving remittances; both households were urban. As previously mentioned, two households with migrants in the Caribbean showed extreme values for some key indicators, suggesting that they may not be representative. Moreover, all households with migrants in the Caribbean were female-headed households, reflecting the intense circulation of male laborers across this corridor.

When comparing the remittance spending patterns of migrant households with those of non-migrant households, we found higher proportions reportedly spent on food, health, rent, land, and education in the former (see Table 1). A significantly higher proportion of migrant households responded positively to these indicators than non-migrant households, particularly the education variable. The double-digit differences in proportions between migrant and non-migrant households with respect to expenditures on schooling (-25%), rent (-18%), and health (-14%) could indicate that in households with migrants as opposed to friends living aboard, there is more stability and consistency in remittance reception which translates into a more productive use of remittance transfers. There were stark differences in remittance spending on health along gender lines; 67% (32 of 48) of women compared to only 31% of men (8 of 26) reported directing these resources toward health expenditures.

Table I:

Consumption area	Migrant	Non-migrant	Difference
Food	95%	87%	-8%
School	62%	37%	-25%
Health	54%	40%	-14%
Rent	33%	15%	-18%
Land	8%	3%	-5%

Source: Authors’ calculations based on INURED household survey data.

Across some areas of spending, Global North migrant households exhibited similar reported proportions as those migrant households with family members in the Global South. For example, 30 out of 31 (97%) households with migrants in the Global South reported spending remittances on food compared to 40 out of 43 (93%) households with migrants in the Global North. Reported proportions were also similar for education expenditures, although a greater proportion, 28 out of 43 (65%), of Global North transfer recipients paid for schooling versus 18 out of 31 (58%) for the Global South. The greatest differences were in health and rent expenditures. While a greater proportion of Global South migrant households reported spending on health, there was an equally high difference in the other direction when looking at spending on rent. Just under half (20 out of 43) of Global North migrant households reported spending on health, while almost two-thirds (20 out of 31) of Global South migrant households did the same, a net difference of 18%. This supports the assertion that female-headed households are more likely to spend on health, as more than three-quarters (76%, or 42 out of 55) of households with migrants in the Global South are headed by women. This is compared to 30 out of 54 (or 55%) of Global North households that are female headed. Forty percent (17 out of 42) of Global North migrant households reported using remittances to cover rent expenses, as compared with fewer than one-quarter (7 out of 31) of Global South migrant households.

When cross-tabulating the proportion of positive answers to the indicator for child #1 by transnational status, there was a higher proportion of remittances supporting educational expenditures from migrants in the Global North (49%) than the Global South (37%). As private school tuition is paid periodically throughout the year, this may partially explain the greater frequency and larger transfer amounts sent from migrants in the Global North, as they may have made explicit commitments to cover school tuition costs. However, financial support for school tuition received from the Global South is significant, and over one-third of migrants have made similar commitments.