1. The Question
Kenya is one of the biggest remittance-recipient countries on the African continent, where remittances have been estimated to amount to approximately USD 4.4-4.8 billion in 2024, equating to 4.6% of GDP and surpassing income from coffee and tourism. This outstrips the total ODA income received by Kenya, which has been estimated at USD 2.6 billion for 2023 (2.5% of GNI) and is likely to go even lower in the wake of reduced foreign aid globally. Anecdotally, and this is exactly what a research project aimed at investigating the claims of philanthropy trips would aim to question rather than accept at face value, remittances can be considered superior development finance to aid due to their efficiency and directness.
This paper asks: do Kenyan households connected to diaspora remittance networks actually show different growth and welfare trajectories than households dependent primarily on NGO or government aid, and if so, why? We treat this explicitly as a question about mechanism, not just outcome, because the answer has direct implications for how a philanthropically-minded research initiative should think about its own theory of change.
2. The Microeconomic Mechanism
Both remittances and foreign aid are, at bottom, forms of external transfer into the household budget constraint, which means money that the household received without generating it through local production. Economic theory suggests that they would increase consumption and investment, if productive uses are made of the funds. What makes the puzzle is the difference in the observed impact of each form of foreign transfer, and this is because not of the quantity but four things about the transfer itself.
Directness (intermediation layers). A remittance normally involves only a single chain, from sender to recipient, with a mobile money platform, and not an intermediary that may take a decision for the household. Whereas traditional aid is normally intermediated along several channels, namely, donor government, implementing partner such as a multilateral organization or NGO, local government or community institutions, and finally the household. Each stage can be a point of leakage or misappropriation or principal-agent problem, which has always been the main criticism of aid effectiveness ("Aid has conventionally been an indispensable source of finance, though in many cases ineffectual in precipitating real development, besides being highly volatile").
Conditionality. Remittances come with absolutely no conditions attached, the recipient is free to use them in any way he chooses. In many cases, traditional development aid is highly conditioned in that the particular use (e.g., building a school, a hospital, or an input pack) of the aid is predetermined by the donor or implementing agency, and it may or may not be the most valuable use for the household at the time.
Frequency and predictability. Remittances to Kenya are quite regular, in terms of monthly payments to family members. This enables the household to have better consumption patterns and plans. Assistance, especially project or emergency assistance, may just be one-time or short-term help, and this makes it ineffective in helping with long-term planning and investments.
Channel efficiency. Kenya’s remittances are increasingly conducted through mobile money rails, which are equally inexpensive and widespread as the rail system that has been analyzed earlier in this study series, more than 70% of remitted funds from the diaspora to the home country go through formal institutions such as mobile money service providers, and mobile money account ownership among the recipients is extremely high (close to 82.5% in the most recent remittances survey of the country).
The falsifiable hypothesis of the framework is not that remittance and aid are different simply due to their names, but because of where each one lies on the four dimensions outlined above. For example, an aid project that happens to fit within all the criteria for being a remittance will have far more in common with a remittance than project-aid.
3. Evidence
Remittances are large, growing, and increasingly channeled efficiently. Remittance receipts into Kenya increased by 18% in 2024, rising to approximately USD 4.4 billion, already surpassing the revenue from two key traditional export products of the country, namely, coffee and tourism. The national Remittances Household Survey conducted in 2025 indicated that 22.3% of the remittance-receiving households consider remittances as their primary source of livelihood, while a further 78.7% use them as a supplementary income source, which indicates the deep reliance of some households on remittances and its supplementation in most others. Remittances receivers demonstrate relatively high rates of financial inclusion (82.5% having mobile money accounts and 55.4% having bank accounts), which is in accordance with the channel-efficiency mechanism discussed above, although relatively low investments in securities and microfinance accounts indicate that they are used mainly for consumption.
Traditional, intermediated aid in Kenya shows modest measured impact. In a 2023 analysis, the World Bank estimated that social assistance cash transfer schemes in Kenya provided by the national government had reduced the national poverty headcount rate by just 1.2 percentage points and the poverty gap by 0.5 percentage points, owing to poor coverage, insufficient cash transfers, and budgetary constraints. This is a useful benchmark exactly because it is a cash transfer scheme and not an in-kind scheme, but it is even less successful compared to the example below because of the multiple layers involved in its administration.
Aid that mimics remittance mechanics produces remittance-like results. The most obvious validation of the mechanism discussed above has come from the randomized studies of the GiveDirectly program that provides unconditional cash transfers to residents of rural Kenya through direct mobile money transfers to their households without any strings attached. Large effects were observed, a rise in household consumption by around 23%, an increase in the value of assets held by 61% and an increase in monthly income by 33% among the recipients, as well as spillover effects in terms of increased consumption levels in non-beneficiary households and additional revenues earned by local firms, refuting concerns regarding crowding out of other people in the local economy as a consequence of generous transfers to selected households. Another study linked to the GiveDirectly organization showed that a single cash transfer of about $1,000 decreased infant mortality rates by 48% in the treated areas, especially among women who lived in close proximity to health facilities thanks to higher ability to afford prenatal care. Mechanistically, this transfer program is very similar to a remittance program as far as the process of delivery goes, it is unconditional, direct, carried out using mobile money and initiated by the household itself.
4. Argument
The data provides sufficient reason for a reinterpretation of the title of this very paper. There is not a question of “remittances versus aid” as two distinct types of resource flows; rather, there is a question of whether the flow is direct, unconditional, and informationally efficient on the one hand, and intermediated, conditional, or underresourced on the other, and just happens to be the case that most remittances are direct and unconditional and most traditional aid is not. The redesign of an aid project to make it more similar in structure to remittances, as the GiveDirectly project has done by introducing unconditional transfers via mobile money, makes its impact resemble the effects reported by remittance recipients, sometimes surpassing even those of Kenya’s own state-subsidized cash transfer scheme.
It follows that there is a clear implication for what one can reasonably expect “diaspora-linked” communities to be different from “NGO-dependent” ones. In the case where an NGO-dependent community has its dependence on NGOs in project aid (such as schools, boreholes, in-kind distribution, technical help), the above mechanism suggests that growth impact will be less pronounced and will vary more widely at the household level than it would in a comparable diaspora-linked community that receives direct remittances. However, if the community’s experience with NGOs lies with direct cash transfers, then the difference will be considerably smaller.
5. Implications
- For a philanthropy-minded research initiative: Consequently, the weight of evidence suggests assessing aid programs, which could include those faced by members of the initiative when traveling for the inspiration to create it, based on the mechanisms by which they are delivered, as opposed to the intentions behind them or the good name of the entity responsible for the programs.
- For Kenyan development finance policy: As remittances now amount to twice the size of aid and continue to grow even as aid decreases, it may be more beneficial to pursue policies that will lower the cost of sending remittances (which averages at 6% worldwide but is higher in Sub-Saharan Africa) and create more financial services for recipients.
- For primary research: This has been the most obvious gap to emerge in all of these papers, which is the lack of any study that has compared matched diaspora-based communities and NGO-reliant communities with regard to the growth of households. An empirical household survey which compares a community that relies on remittances with an equally similar community based on NGOs and aid would add much originality to the literature.
6. Limitations
This paper does not have, and, to our knowledge, none currently exists in the published literature, an analysis that causally compares the growth metrics of matched diaspora-connected and NGO-reliant Kenyan communities; the comparison developed in this paper is drawn from two literatures (on remittances and on aid modalities) which differ in methodology, timeframe and geographical focus, making it difficult to consider the “convergence” theory presented in Section 4 as anything more than suggestive. In terms of the methodologically rigorous (randomized controlled trial) GiveDirectly evaluations, the study was focused on specific programs implemented in specific counties (Rarieda and Siaya, western Kenya). The data on remittances, specifically their use in households in detail (rather than just general categories such as school fees, healthcare and business), is less abundant compared to the literature on aid evaluation studies which have the advantage of being based on randomized trials. Finally, the claim made in this paper that “delivery mechanics, not funding source, drives outcomes” is a reasonable hypothesis based on the existing body of evidence, but not a causal finding, exactly the gap that would be closed by a household survey conducted by this research initiative.
References (selected)
- KNBS, CBK & FSD Kenya (2025). 2025 Remittances Household Survey (RHS).
- Central Bank of Kenya (2021). Diaspora Remittances Survey Report.
- World Bank (2024-2025). Migration and Remittances data; Kenya net ODA received data.
- OECD (2025). Final statistics on official development assistance (ODA) and other resource flows to developing countries in 2024, and Cuts in official development assistance report.
- Development Initiatives (2024). Improving the impact of ODA on social protection in Kenya.
- Haushofer, J. & Shapiro, J. (2016). Household response to income changes: evidence from an unconditional cash transfer program in Kenya. Quarterly Journal of Economics.
- Egger, D., Haushofer, J., Miguel, E., Niehaus, P., & Walker, M. (2022). General equilibrium effects of cash transfers: experimental evidence from Kenya. Econometrica.
- Innovations for Poverty Action (IPA) & J-PAL, GiveDirectly Kenya evaluation summaries.
- Misati, R. N., Kamau, A., & Nassir, H. (2019). Do migrant remittances matter for financial development in Kenya? Financial Innovation.
- USIU-Africa (comparative study). A Comparative Study of the Effects of Foreign Remittance and Foreign Aid on Kenya’s Economic Development.
Prepared as part of the East Africa Microeconomics Research Series. This is a student working paper that synthesises existing research and published data rather than presenting original fieldwork; feedback and data corrections are welcome via the contact page.