With rigorous economic research and practical policy solutions, we focus on the issues and institutions that are critical to global development. Explore our core themes and topics to learn more about our work.
In timely and incisive analysis, our experts parse the latest development news and devise practical solutions to new and emerging challenges. Our events convene the top thinkers and doers in global development.
What if taxpayers could decide for themselves how some of the UK’s aid budget is spent?
Allocating funding would let taxpayers engage meaningfully with development issues, potentially reinforcing support for tackling poverty and deprivation overseas.
Competition for funding would give international development organisations an incentive to offer an explicit value proposition. This could catalyse a race to the top in becoming transparent, measuring impact, and delivering value-for-money.
AidChoice, as set out below, would be revenue neutral, would not lower the UK’s overall spending on foreign aid (or the amount scored as ODA), and might generate modest but meaningful savings, all while increasing public support for development spending and improving accountability.
Our thanks to Caitlin McKee for excellent research assistance and Michael Anderson for helpful comments on an earlier version of this proposal.
AidChoice is our name for a simple idea that would empower taxpayers to allocate funding to international development projects listed on a website up to a ceiling (say, £100 per taxpayer) and deduct that payment in full from their tax bill. Taxpayers would simply quote their national insurance number at the time they make the donation through a recognised platform, and HMRC would reduce their tax bill accordingly.
AidChoice would reallocate aid without reducing the UK’s overall spending on aid: it would not reduce the total size of the UK’s commitment to tackling poverty overseas. It would be inexpensive to implement, and an initial trial with a low cap could easily scaled up if the scheme proves to be popular and successful (based on evidence from a rigorous evaluation). Pairing the trial with an evaluation framework would enable officials to track efficacy and observe where taxpayers chose to allocate aid when it is in their control.
Enabling taxpayers to allocate their share of the development budget would enhance the relationship between organisations that use aid and the people who pay for it. It would encourage greater local support for the foreign aid agenda and offer unconstrained contributions to innovative organisations. And competition for this valuable source of funding should encourage implementing agencies to provide a compelling proposition to their stakeholders by showing that they can deliver value for money.
AidChoice makes aid more accountable
The beneficiaries of UK aid are a long way—geographically and politically—from Whitehall. Foreign aid is an unusual public service in that the people who provide the funding—and who have the power to hold delivery organisations to account—have little or no direct experience of the services that aid supports. This means that it is hard for the UK government’s aid spending to be accountable, despite DFID’s impressive efforts to be transparent. Oversight of aid spending by parliament, the Independent Commission for Aid Impact, and the National Audit Office provides valuable scrutiny but these cannot substitute for the feedback from direct experiences that citizens have of services like transportation, education, and health.
Private actions and attitudes do not match the narrative of too much government generosity for fighting poverty. British people gave over £2.5 billion for international development in 2013, equal to nearly a fourth of the UK’s total aid budget and second only to Americans’ private giving. Clear majorities think that international development’s objectives matter: 62 percent of UK respondents believe they can “play a role in tackling poverty in developing countries” and 70 percent think that doing so also “has a positive influence on EU citizens.” However, recent opinion polls suggest that the UK public trusts charities more than the government and only 19 percent of people think politicians are a credible source of information about international assistance. Something about the intermediation of aid by government (more pejoratively, the “nationalisation” of aid) makes it less popular than aid that people themselves control.
The costs of aid are small but real enough to make them a political target: spending on overseas development assistance (ODA) worked out to £385 for every taxpayer in the UK in 2014. For more than a decade, DFID has been aware of the need to build support for development activities, reaching back to a strategy paper in the 1990’s outlining how to raise public awareness and laying foundations for incorporating the “Global Dimension” in school curriculum in the 2000’s. In an era of pressure on budgets, the 0.7 percent of GNI target for spending on foreign aid is a laudable commitment, but the political consensus behind it is fragile.
Creating a sustainable national consensus that giving aid is a good thing means changing this conversation. AidChoice proposes to do this by recognising that part of the problem is disengagement with the government’s aid programme, not Britons’ lack of interest in alleviating poverty and hardship. AidChoice addresses this not by offering more publicity or clearer information, nor by investing in education programmes in schools, but by giving taxpayers more control over how their money is spent.
At the moment, most bilateral aid agencies are, in effect, vertically-integrated monopolies. They use dedicated budgets to develop and deliver aid programmes that they select and (to some extent) control and manage. They may spend some of this money through other agencies or through the multilateral system (of which they are the principal shareholders), but aid agencies have been largely insulated from the move towards a purchaser-provider split which has been part of public management reform of other public services since the 1980s. The idea of this split is that there should be a clear separation between those who advise the government on policy direction and those who implement the policy: otherwise the policy advice will tend to reflect too much the interests of the current implementers.
Where there is a purchaser-provider split, service delivery outcomes are defined and made transparent in contracts or service agreements, an approach which is mainly absent in the aid system. In the absence of a purchaser-provider split, and with the inherent problems of accountability of aid, aid agencies can appear to be monolithic, unresponsive, secretive and self-serving, adding to public distrust of public aid programmes.
AidChoice would help bridge the gap between how aid is spent and the people who pay for it. It would do this by giving taxpayers a direct say in how some of their aid is spent. This would create a feedback loop—now absent—between organisations implementing projects to tackle poverty overseas and the UK taxpayers who fund it, and become a platform for UK taxpayers to become emotionally and intellectually engaged with how foreign aid is programmed and disbursed.
How AidChoice works
A simple website could be set up in collaboration with HMRC in which eligible charities (see determining charity eligibility section below) enter their charity number. Through this website the charity submits to HMRC a list of Taxpayer Identification Numbers or National Insurance numbers and the amount of the donation. The taxpayer lists the donation on their tax form, which is then automatically cross-checked against the charity’s list of donors and the amounts. If the donation details match, the taxpayer’s tax liability is reduced by that amount.
For example, a taxpayer chooses to donate £70 to Oxfam GB via the AidChoice website. This choice would be logged into the HMRC website, and £70 would be transferred directly to Oxfam GB, while the DFID budget would be reduced by £70. The amount of money paid by the individual would remain unchanged, it would just be spent differently.
Users could choose how much to allocate to one or several programmes up to their individual cap. (A user could voluntarily give more, but the total tax rebate would be capped; she could also give less, and she can spread her giving across more than one recipient). By associating each transaction with a national insurance number or taxpayer identification number, the user receives a pound-for-pound reduction in their tax bill. This in turn leads to an equal reduction in DFID’s budget. Financial transaction costs would be kept very low by batching the donations and processing them together.
Charities must use this money in a way that is consistent with OECD’s definition for ODA spending that “is administered with the promotion of the economic development and welfare of developing countries as its main objective”. AidChoice can be organised in a way that still counts as ODA and towards the government's commitment to spend 0.7 percent of GNI for this purpose.
Determining charity eligibility
The eligibility standards for which organisations could receive AidChoice funding would have at least three characteristics:
minimal to allow a large number of agencies that are eligible for funding under the programme,
cheap, so that they do not require setting up a large, separate monitoring bureaucracy, and
easily monitored, so that they do not create an additional reporting burdens for implementing agencies, which would discourage smaller and potentially more innovative NGOs from participating.
AidChoice should use definitions that are already in place, widely trusted, and associated with a legal definition and tax status. The UK Charity Commission regulates organisations and evaluates whether they fulfill a charitable purpose. To begin with, eligible organisations could include UK-registered charities whose primary purpose is classified as Overseas Aid. This extensive list of almost 12,000 organisations would be narrowed down to only those UK charities who publish their aid spending to the International Aid Transparency Initiative (IATI). The resulting list of less than 200 charities would be automatically eligible for AidChoice funding.
In this way, any organisation that would like to be eligible can meet these standards to be included (they do not need to be vetted by a selection process). Yet due to the rigorous standards of IATI, it still keeps the total number of potential organisations to choose from manageable for the taxpayer, and it ensures that only organisations that are transparent in their use of money are eligible. Both the UK charities registry and IATI are already established, so little additional monitoring mechanisms on the platform would be needed. These selection criteria ensure that HMRC can easily trace the funds to the charities but also that the spending of those funds is publicly visible.
In addition to funding well-known programmes or projects, AidChoice could be used as a Kickstarter-like platform to finance newer, original initiatives, so that taxpayers could act like investors making small bets on disruptive innovations. (And those that fail—as some inevitably will—should reflect choices by taxpayers choosing to make a bet with their own money, rather than decisions by government-appointed gatekeepers which then need to be publicly defended.)
By making the threshold transparent and simple to monitor, the costs of operating AidChoice and determining eligibility would be small.
What AidChoice delivers
What kind of an impact would this have on UK aid? UK ODA was £11.73 billion in 2014, and the most recent public data from HMRC show 30.4 million individual taxpayers in 2013-14. If contributions are capped at £100 and a fifth of taxpayers use the platform (which would be a lot—more than 6 million people), it would mean a £608 million reallocation, roughly 9 percent of the £6.8 billion in bilateral aid that DFID spent on its own programmes in 2014. (Bilateral spending is less than total ODA because some of the aid budget is pre-committed to the World Bank, EU, the UN system, and other multilateral organisations), and some money is spent by other government departments. Any unallocated money would go to DFID by default.
This is not a trivial amount, but nor does it hollow out the aid budget. AidChoice would reallocate spending without reducing the total amount spent by the UK on tackling poverty. There would be a reduction in the money passing through DFID’s accounts, to the extent that taxpayers chose to decide for themselves how their aid money was being used: it is possible that this money would be used at least as well as if it had been allocated by DFID Ministers and officials. An initial trial with a relatively low cap could easily be scaled up if the scheme proves to be successful.
How might the competition work?
Consider the airline industry: carriers like Air France and Ryanair compete for passengers on price, quality, and unique attributes like routes and departure times. At the same time, a secondary industry—from travel agents to flight comparison websites like Expedia or Kayak—exists to help passengers navigate these offers to find the ones that appeal to their needs and budgets.
AidChoice encourages the same two-tiered competition.
First, it gives agencies the incentive to compete for taxpayers’ attention and resources.
Second, it facilitates competition between various aid platforms to add value by comparing those agencies and evaluating their impact.
At the agency level, requiring that all applicants meet the same eligibility threshold would create a level playing field, putting ‘outsider’ programmes and policies on an equal footing with smaller NGOs and other agencies.
A competition to attract funding would raise our collective expectations about the quality and quantity of information that aid organisations should provide about what they are doing and how effective they are. GiveDirectly, for example, has a simple proposition: it motivates donations because it is committed to a remarkable degree of transparency about its operations and finances, extending to updating its website with impact evaluations literally as the data come in.
The second tier of competition is between the third-party websites and portals that add value by comparing potential funding recipients. These would combine the research into impact done by organisations like GiveWell with the comparisons provided by websites like Expedia (travel) or Confused (insurance). This is not a novel concept. International development organisations already market themselves to solicit donations, while websites like CharityChoice and GiveWell help people decide which organisations deserve their money. AidChoice creates opportunities for these parallel competitors.
AidChoice builds on existing success
In its mechanics, AidChoice would draw on popular programmes that already exist. The mechanics of associating private donations with tax rebates is similar to the HMRC’s Gift Aid scheme, which provides tax offsets by allowing charities to reclaim the income taxes that donors already paid on their donations. AidChoice instead allows a taxpayer to control for themselves how their aid contribution is used (up to a ceiling) and funds the 100 percent tax rebate out of the aid budget.
Similarly, AidChoice’s emphasis on spending choices by users has a precedent in participatory budgeting trials– experiments with giving communities more determination over how their local councils spend relatively small budgets. An early evaluation suggests these schemes increase people's control over the allocation of resources, expand the number of people involved in decision-making, open up new channels of engagement, disrupt the status-quo by allowing new or different services to receive funding, and increase confidence in local service providers.
AidChoice also builds on the concept underlying the distinct but analogous AidMatch scheme, which allows a small group of hand-picked NGOs to earn matching grants from DFID for private donations they receive. But AidChoice provides a direct link between taxpayers and development programmes; this is in contrast to AidMatch which appoints DFID as gatekeeper, selecting eligible NGOs and therefore benefits a far smaller number of charities, causes, and innovations. AidChoice would give taxpayers much more choice, with no incremental cost to them to choosing to direct some of their tax contribution themselves.
AidChoice is not wholly novel. A similar concept was proposed in academic literature as early as 1968, suggesting US citizens could get tax credits by directing foreign aid through using non-governmental channels. Like AidChoice, the goal was to rekindle public interest and trust in aid. AidChoice features several of the same innovations, including a cap on contributions (in the earlier formulation, this was designed to limit disproportionate influence of disproportionately wealthy people), encouraging competition amongst recipients, and creating pressure to be more efficient—all while delivering more impact for fewer dollars.
Can we “trust” taxpayers?
The twin principles underlying AidChoice is that taxpayers who fund our development spending should have the opportunity to be more invested in the causes their spending supports, and that government policy should reflect the electorate’s preferences.
Take the example of cash transfers. Though it seems counterintuitive to many, cash transfers in emergency and humanitarian situations are quickly emerging as the new standard to beat. Giving people tiny, unrestricted grants enables them to buy what they need locally, thereby supporting local markets and repositioning people displaced by catastrophe and violence as paying customers rather than drains on local services.
Despite these creditable features, aid agencies have not reacted by rapidly scaling up their use of humanitarian cash transfers. Though data are opaque—itself a longstanding problem in development spending in general and humanitarian aid in particular—experts’ best estimates suggest that just 6 percent of humanitarian aid is disbursed as cash and vouchers to individuals.
Agencies’ hesitance is not because we lack evidence about efficacy or efficiency. Growing from early pilots, so-called humanitarian cash transfers are now amongst the most rigorously evaluated development interventions ever. Studies conclude that those who have received cash transfers from aid agencies rather than goods (often called ‘aid in kind’) overwhelmingly do not waste money on alcohol or other sin goods, but use it to invest in their homes, educate their children and manage their household budgets. And delivering cash, rather than food or other equipment, creates significant savings for stretched humanitarian budgets.
In contrast to aid agencies, taxpayers seem to have a much stronger preference for allocating meaningful shares of aid as unconditional cash transfers. On average, people choosing how to split up a notional £5 in aid between cash and traditional in-kind aid would spend £1.9 of it as a cash transfer when the money is spent on ‘good’ things.
But the data also show a steep decline in support if there’s a perceived risk that even some of that money will be misspent. When asked to divide aid between ‘cash’ and ‘in kind’ when some of the cash could be spent on alcohol, the average choice of how much to allocate as cash drops to just £0.9.
Though this is an abbreviated summary of our in-depth analysis of the survey, there are three stand-out points relevant to AidChoice:
The average preference for spending money on a valuable, innovative way to invest foreign aid for those in urgent need is orders of magnitude greater amongst people than amongst aid agencies: as a share of £5, a choice of £1.9 going to unrestricted cash transfers is nearly 40 percent, over six times the six percent that frontline agencies are currently estimated to spend this way.
Even when we put our thumbs on the scale by suggesting that ‘some’ money could be misspent on something bad (alcohol), support dips to £0.9, which, at 18 percent of £5, is still three times amount agencies spend this way.
Nevertheless, the difference in public tolerance for cash transfers suggests that the people who fund development programmes are extremely sensitive to risks that money is wasted or misspent. Since evidence suggests only a vanishingly small fraction of humanitarian cash transfers are wasted like this, it reinforces, rather than undermines, the benefits of educating taxpayers about how their money is spent through giving them a voice in how to spend it.
On balance, narrow evidence from asking a large cross-section of European respondents about a particular aid innovation suggests taxpayers may well have preferences for smart and timely innovations in how foreign aid is spent—and that aid agencies, left to their own devices, may not develop or share these preferences, or may be averse to adopting these innovations for fear of what their stakeholders may think.
Benefits that AidChoice could deliver
Empowering taxpayers to invest may directly encourage buy-in for development spending: taxpayers may identify with the charities they have chosen, providing a counterbalance to political pressure to limit the UK’s spending on tackling poverty.
This could help combat the common misperception that spending on aid is much higher that it is. The UK’s aid budget was £7.9 billion in 2012, but an Ipsos Mori poll that year found that more people chose aid as the government’s single largest expenditure than chose pensions (£74 billion) or education (£51.5 billion). The opportunity to allocate a small part of the aid budget will remind taxpayers that there are very many worthy causes from which to choose.
Most importantly, enabling taxpayers to allocate scarce funds also creates competition amongst development organisations for a share of the UK aid budget. (DFID’s capacity will not be unduly eroded because the total tax rebate value would be capped). By agreeing and publishing standards for inclusion on the roster of eligible agencies, this competition raises the bar for the wider development sector.
Overall, AidChoice would:
generate public interest in and support for foreign aid, fabricating a feedback loop between taxpayers and aid allocation without political constraints;
provide a cheap, transparent way to give programme support to charities—including small or experimental ones—that is difficult for a large aid bureaucracy;
raise the bar for transparency and communication amongst the agencies that spend aid through eligibility requirements and competition for AidChoice funding.
Costs of implementing AidChoice
AidChoice is revenue neutral from the point of view of public finances: users, who are uniquely identified by their national insurance number, allocate funding to recipients and are refunded on their tax statements, with the rebate financed by a reduction in DFID’s budget.
From a Treasury perspective, the proposed tax rebate would reduce the measured tax burden, and the increase in spending through non-government channels would count as a reduction in government spending as a share of national income (and could save some civil service costs by reducing the total amount of aid spending that needs to be programmed and accounted for by bureaucrats).
There would be implementation costs for running the website and administering the programme (including website development and maintenance, creating content, and informing users), but these would be modest. IATI is a reasonable comparator; their budget is less than £175,000 annually to operate, maintain and upgrade their website, including staff time. Some initial fixed costs should be added to this, but overall variable costs would be modest.
These services for AidChoice should be done outside government, if necessary with some modest government or philanthropic financing. Charities seeking additional funding would also launch their own campaigns to attract taxpayer attention (much as beneficiaries of AidMatch and Gift Aid have done). As a result, government spending on marketing to inform the public about AidChoice would be minimal.
While AidChoice is not wholly novel and parallels can be drawn between it and existing UK programmes, it is indeed experimental. Careful evaluation would need to be done to determine if it did shift public attitudes towards aid support. If it did not meet its aims or caused other undesirable disruptions, then the programme could be discontinued.
Assessing take-up of the programme would be the first step to see if this market does spring up and taxpayers use it: testing whether the feedback loop is created. Brief surveys to participants could be added to the AidChoice platform to see if it leads to a better-informed, more engaged public with increased support for foreign aid. Spending on eligible charities could be easily aggregated to see the level of programme support provided by size and type of activities. And if more organisations began publishing to IATI standards and communicating cost-effectiveness, then this would be a win for development beneficiaries and funders.
Questions and answers
In the section below, we consider other possible challenges for the successful implementation of AidChoice.
Won’t taxpayers make worse choices than development professionals?
Taxpayers typically want to support people in need in the most effective possible way, but are not necessarily well informed about what that is. Development organisations are typically better informed, but also have other interests which they must pursue (for example, broader government objectives, institutional incentives, etc.).
Appointing development specialists—for example, by deputising DFID to control aid spending—may make some aid more effective, but may also create a bias toward the status quo that prevents disruptive innovations from flourishing, as in the case of humanitarian cash transfers. And retaining specialists as the only gatekeepers of aid spending does little to secure broad-based political support for foreign aid amongst the electorate.
So shifting some spending decisions from development agencies to taxpayers could be better (because the incentives may be more strongly aligned with the interest of the beneficiaries) or it could be worse (because the decisions may be less well-informed).
Two factors suggest that, over time, it may well be better. First, AidChoice creates incentives for agencies to be transparent about their impact, so taxpayers who opt to make these decisions would have more opportunities to become better informed. Second, taxpayers may already prefer to support investments and spending that incumbent agencies do not—as in the case of humanitarian cash transfers, above—but have not been given a voice in budgeting to demonstrate this.
Will this cause a “race to the bottom”?
Creating interest in AidChoice might drive organisations to favour short-term, appealing projects of limited impact instead of experimental, challenging, less sexy, or longer-term engagements. There is certainly evidence that people’s decisions about giving to charity are driven by ‘warm glow’ effects—emotional narratives increase giving but information on rigorous evaluations, paradoxically, decreases it.
These risks are minimised in the following three ways:
worthwhile initiatives which do not attract direct public funding can be financed by DFID, which will continue to manage the vast majority of the aid budget;
the platforms would be data-rich sources of accessible information on aid impact and transparency, and would help to educate the public (who are smarter than policymakers sometimes give them credit for) about the value of these initiatives; and
the organisations that are eligible for AidChoice are already vetted by existing platforms, thereby excluding organisations that do not satisfy the Charity Commission’s public interest tests or IATI’s transparency requirements.
What about large-scale collective funding or other ways of financing development, like soft loans, equity stakes, guarantees or development impact bonds?
AidChoice works by providing a way for individuals taxpayers to choose where their money goes, and batches those allocations together to benefit the organisations that taxpayers choose. This lowers the transaction costs of collective giving and, in this respect, is similar to crowdfunding platforms like Kickstarter and Indiegogo.
This is not limited by design to any particular funding instrument. This means, for example, that taxpayers might choose to spend their money on a grant to a programme, but could also choose to allocate funding to investments, provided they are managed by organisations that meet the criteria (namely, UK-registered charities which publish their information to IATI).
Will AidChoice inspire organisations to spend more on marketing and promotion, so increasing their overheads and wasting money that should be spent on their charitable purpose?
AidChoice increases the rewards to better outreach, so organisations will have a stronger incentive to communicate with the public about what they do. This is a good thing unless it becomes excessive (i.e. it turns into rent-seeking). We know that the public are skeptical of organisations that spend a lot of money on administration, marketing and other overheads (probably too skeptical).
Since eligibility for AidChoice requires a high level of transparency about how the money is spent through publishing financial data to the IATI aid data standard and the charity commission registry, organisations will have to disclose what proportion of their income they are spending on marketing and overheads. This means that organisations will have an incentive to increase their spending on communications up to the point at which the public considers that they are spending “too much” on it.
Gift Aid, which we mention above, also raises the returns to marketing by NGOs, but has not been linked to a wasteful increase in spending on self-promotion (though many do advertise that they benefit from the scheme).
Will this create uncertainty for NGOs?
Funding from DFID (and other government donors) can be very uncertain: it can be arbitrarily cut, including in-year, based on a decision by one Minister or senior civil servant. Crowdfunding by taxpayers, by contrast, is the cumulated sum of many decisions by many people. The law of large numbers suggests that this kind of funding might be more predictable and more stable, and come with less administrative hassle, than funding delivered via DFID. Furthermore, the money obtained by NGOs through AidChoice may come with fewer strings attached than money from aid agencies.
Will this create uncertainty for DFID?
Large changes in the number of taxpayers who decide to exercise their own choice, rather than leaving decisions to DFID, could result in a significant change in DFID’s room for manoeuvre. However, the number of taxpayers who choose to allocate their aid spending themselves is unlikely to change much from one year to another, so once the scheme is up and running, it will be reasonably predictable from one year to the next.
Will this undermine DFID?
The majority of DFID’s budget would still be earmarked for spending by the agency itself, which could be prioritised for the difficult or protracted programmes that should be an essential part of the UK’s support for international development.
AidChoice will disintermediate DFID with respect to the relatively small share of the aid budget that taxpayers choose to allocate through AidChoice. But this is a feature, not a bug: it increases competition by building domestic support for aid spending and could help build a closer relationship between taxpayers and aid programmes. The aim of the development community should be to increase public support for international development, not public support for DFID.
The spending will be outside the government’s financial systems. Does this increase the risk that the money will be misused?
Individual taxpayers will be deciding for themselves where to allocate their own money, and can decide for themselves what level of safeguards they want for their money. They can allocate money to long-established and respectable organisations if they have a low risk appetite, or a newer organisation with less of a track record if they prefer. Under AidChoice, allocation will be by ‘many eyes’ looking at a wide range of information about organisations, including a choice of platforms competing to provide the most useful information about the different programmes. This could well lead to a higher level of scrutiny than ‘expert’ scrutiny from an overworked official in Whitehall.
Why can’t taxpayers just keep the money they earn and decide for themselves how much to give away and for what?
Some people challenge the whole premise of giving money for development through government, believing that it should be a choice for individual taxpayers. The premise of AidChoice is that there is a sound case for requiring everyone with sufficient means to make a modest contribution towards development, and to prevent free-riding: taxpayers would prefer to live in a more prosperous, safe, and healthy world, but might be inclined to let someone else pay for it. AidChoice balances these views by giving gives taxpayer more choice about how some of their money is used but without giving them a choice about the overall level of giving.
Won’t AidChoice crowd out giving to domestic charities?
This seems unlikely: the cap on allocations will be modest, likely lower than the £120 that the median household gave to charity in 2012. And AidChoice will clearly be limited to funding for international development, without competing with domestic charities.
While we can argue about how much to spend on the national health service or repairing roads, both are ultimately government services that we benefit from. Foreign aid’s stakeholders, in contrast, are generally out of sight and out of mind. This distance—physical and political—makes it difficult to catalyse support for, or understanding of, foreign aid.
AidChoice aims to help fill this gap, using existing tools to give taxpayers direct and personal ownership of the foreign aid agenda by partly disintermediating foreign aid agencies. In doing so, it creates an incentive for organisations to provide a compelling proposition to those who fund it, rewarding transparency, value-for-money, and results.
AidChoice is a new way of allocating a portion of aid funding and will be evaluated after being rolled out. Because it is revenue-neutral, an experiment with AidChoice would not have large-scale budget consequences and, given the possibility of success, it is worth trying.
AidChoice could shift the national conversation about spending to tackle poverty overseas, creating vital, stable, and sustainable political support for UK aid.
 In this proposal we focus on UK income taxpayers as the most easily identifiable participants for AidChoice. However, for the democratic objective of allowing non-wage earning citizens to have a voice in how aid is spent, the scheme could be expanded to include the 44.7 million British citizens registered for parliamentary elections. This would require further exploration of a system for identifying these individuals without a unique identifying number (a National Insurance number isn’t required to register to vote) and therefore likely a higher cost for running the scheme.
 Throughout, we focus on DFID because it is a principal recipient of UK aid, but acknowledge that it is not the only government entity which receives aid. In 2015 DFID spent 82 percent of the ODA budget.
 This could later be enlarged so that charities registered overseas, multi-laterals, or other relevant organisations could be eligible for AidChoice funding if they also publish their information to IATI.
 Note that this is an additional requirement than being “registered” under IATI. Registration does not require organisations to publish their spending.
 Hirschman, Albert O., and Richard M. Bird (1968). "Foreign aid - a critique and a proposal." Princeton, New Jersey: Princeton University.
 Of course, DFID is not the only domestic means through which the UK spends ODA, but it is the main one accounting for 82 percent in 2015; we use it here as a placeholder for ‘domestic avenues through which ODA can be spent.’
This paper illustrates the tradeoff between country ownership and funders’ priorities with a formal model in which aid is governed by a contract to produce a jointly desired outcome. The model generalizes the Principal-Agent approaches for studying aid which treat countries as having multiple objectives.