Remittances have been a key income source for many vulnerable households in developing countries, enabling them to meet vital expenses such as foods, healthcare, and education, and substituting for lacking public social safety nets (Bettin et al. 2014). In 2019, remittances to low- and middle-income countries amounted to $540 billion. In sharp contrast to the dire predictions at the onset of the COVID-19 pandemic, remittances have been surprisingly resilient, declining by only 1.7% in 2020 to reach $549 billion for low- and middle-income countries. This strong resilience documented in the literature (Kpodar et al. 2021) has resurfaced the debate on the high cost of sending remittances as it has become critical to address impediments to remittances at a time where poorer countries are grappling with the economic and health fallout of the pandemic.
Against this backdrop, we explore in a new study the cost- and risk-based constraints and market structure predicting variations in remittance fees, using data across 365 corridors over the period 2011-2020 (Beck et al. 2022). We also examine variation in costs across different types of remittance service providers, thus adding more granular evidence to the existing literature (Freund and Spatafora 2008, Beck and Martinez Peria 2009, Cecchetti and Schoenholtz 2018, Bersch et al. 2021).
Remittance fees have been on the decline, but remain high…
Figure 1 shows that the fees for a $200 remittance have come down significantly over time, although they seemed to have stagnated in the last three years. The median remittance fee has decreased from 7.7% in 2011 to 5.7% in 2020. The reduction was also broad across the distribution of remittance fees, with the 75th percentile decreasing from 11.1% to 7.7%, and the 25th percentile falling from 5.2% to 4%. The larger drop at the top of the distribution has contributed to a narrower dispersion of remittance fees. Nevertheless, most of the corridors still have median remittance cost above 5% in 2020, far from the 2030 Sustainable Development Goal target of average fees of less than 3% and no remittance corridors with costs higher than 5%.
The data also reveal significant variations in remittance fees within the same corridor (across firms), between the same sending country and different receiving countries, or between the same receiving country and different sending countries. Similarly, remittance fees vary with the payment instrument, the access point, and the speed of the transfer. This suggests that the drivers or remittances are multiple, complex, and specific to the characteristics and policies of the sending and receiving countries, as well as the corridor itself.
Figure 1 Median price across corridors over time
Sources: Authors’ calculations based on Remittance Prices, Worldwide, World Bank
… reflecting cost- and risk-based constraints and market structure
Teasing out the key drivers of remittance fees requires combining the price data with several country- and corridor-specific variables in an empirical framework exploiting three different dimensions: (i) variation across corridors, (ii) variation over time, and (iii) variation across different remittance service providers within corridors. Five results stand out:
- Higher GDP per capita in the sending country and easier geographic access to financial institutions is associated with lower fees, especially for banks.
- Scale economies matter: a larger market for remittances (as proxied by closer economic ties and a larger migrant population from the receiving in the sending country) is associated with lower costs as is a shorter distance between sending and receiving countries.
- The market structure is important: banks charge higher fees than money transfer operators (MTOs), but a larger share of banks among remittance service providers is also associated with higher fees charged by MTOs. Unlike banks, MTOs’ fees react to competitive pressures, with more market players being associated with lower MTO but not bank remittance fees.
- In corridors where the sending country has a pegged exchange rate, both banks and MTOs charge lower fees.
- There is some evidence that cash payments attract higher fees, while payments over the Internet are charged lower fees. Nonetheless, there are no conclusive results regarding the impact of the regulatory framework, which leaves unsettled the debate that compliance to regulatory requirements (e.g. on AML/CFT issues) heightens the cost of remittance services.
Decisive policy actions are needed, although structural factors present challenges
Overall, these findings point to both cost- and risk-based constraints and market structure as barriers to lower remittance fees. Higher transaction costs as result of a more rural population in the sending country and lower scale can explain high remittance fees in some corridors. These factors are largely structural, implying a limit to the extent to which remittance fees can be lower with policy actions.
However, decisive policy actions are needed in areas that are directly under the control of policymakers and where the yield from reforms can quickly materialise. For instance, stronger competition through easing entry to the remittance market, especially for non-bank providers, and digitalisation might help reduce remittance costs. Similarly, exchange rate stability (or better hedging possibilities) might contain remittance costs, more so because exchange rate margins (often not fully disclosed by remittance service providers) can make up a significant portion of the remittance fees in corridors where exchange risks are the highest. More work is needed to understand how the regulatory framework affect remittance fees, notably by exploiting exogenous changes in regulation (e.g. due to international pressure on AML-CFT frameworks).
Aycinena, D, C Martinez, and D Yang (2009) “The Impact of Remittance Fees on Remittance Flows: Evidence from a Field Experiment among Salvadoran Migrants”, Department of Economics, University of Michigan, Ann Arbor.
Beck, T and M S Martinez Peria (2009), “What explains the cost of remittances?”, VoxEU.org, 28 September.
Beck, T, M Janfils and K Kpodar (2022), “What Explains Remittance Fees? Panel Evidence,” IMF Working Paper (forthcoming)
Bettin, G, A Presbitero and N Spatafora (2014), “Remittances and vulnerability in developing countries: Results from a new dataset on remittances from Italy”, VoxEU.org, 10 February.
Bersch, J, J F Clevy, N Muhammad, E Pérez Ruiz and Y Yakhshilikov (2021), “Fintech Potential for Remittance Transfers: A Central America Perspective”, IMF Working Paper WP/21/175.
Cecchetti, S and K Schoenholtz (2018), “The stubbornly high cost of remittances”, VoxEU.org, 27 March.
Freund, C and N Spatafora (2008), “Remittances: Transaction Costs, Determinants, and Informal Flows,” Journal of Development Economics 86: 356-66.
Gibson, J, D J McKenzie and H Rohorua (2006), “How Cost Elastic Are Remittances? Evidence from Tongan Migrants in New Zealand”, Pacific Economic Bulletin 21: 112–28.
Kakhkharov, J, A Akimov and N Rohde (2017), “Transaction Costs and Recorded Remittances in the Post- Soviet Economies: Evidence from a New Dataset on Bilateral Flows,” Economic Modelling 60: 98–107.
Kosse, A and R Vermeulen (2014), “Migrants’ Choice of Remittance Channel: Do General Payment Habits Play a Role?”, World Development 62: 213-227.
Kpodar, K, M Mlachila, S Quayyum and V Gammadigbe (2021), “Defying the odds: Remittances held up during the COVID-19 pandemic,” VoxEU.org, 27 September.
1 Reducing the costs of remittances has been on the policy agenda for more than a decade. In 2009, the G8 member countries made a public commitment to reduce the cost of remittances by five percentage points over five years (the ‘5x5 Objective’). The Sustainable Development Goals include the objective to “by 2030, reduce to less than 3 per cent the transaction costs of migrant remittances and eliminate remittance corridors with costs higher than 5 per cent” (SDG 10.C). Studies also document that higher fees can reduce remittance flows (Gibson et al. 2006, Aycinena et al. 2009, Kosse and Vermeulen 2014, Kakhkharov et al. 2017).