David Miliband, President and CEO of the International Rescue Committee, has published a new article in the New York Review of Books today.

Refugees and displaced people are the greatest victims of failed political leadership around the world today. In this article I argue that there is a need to rethink aid policy, not least because all the evidence suggests that the root causes of the current refugee problem, including both religious and ethnic politics and geopolitical divisions, are unlikely to be addressed in the short term. I also argue that the United States and other wealthy countries need to play a stronger leadership role in terms of humanitarian financing and resettlement. The summits this week are rightly shining a spotlight on these important issues but they must go beyond stating challenges.

The Best Ways to Deal with the Refugee Crisis

In July 1941, Albert Einstein, ten months a US citizen, wrote Eleanor Roosevelt from his Saranac Lake retreat to register “deep concern” at the policies of her husband’s administration. A “wall of bureaucratic measures” erected by the State Department, “alleged to be necessary to protect America against subversive, dangerous elements,” had, he wrote, made “it all but impossible to give refuge in America to many worthy persons who are the victims of Fascist cruelty in Europe.”

Einstein asked the First Lady to raise this “truly grave injustice” with the president, but his appeal had limited effect. Paranoia that refugees would, if granted entry to America, turn on their host and spy for its enemies persisted. The annihilation the following year of some 2.7 million Jews—nearly half of all Jewish victims of the Holocaust—could not dispel this prejudice. Nor did the killing in 1942 result—amid economic depression, the battle against the Axis, and strains of popular and political xenophobia—in a US response to the refugees’ plight. The American “wall” against refugees would remain largely standing until the beginning of 1944, the year before the Allied victory.

The source of Einstein’s vexation that summer has returned to public life. We are again seeing a double assault against some of the world’s most vulnerable people. Their character and intentions are often impugned and they are denied dignified refuge. A day after American-born Omar Mateen’s June 12 attack on the Pulse nightclub in Orlando, Florida, Donald Trump, the Republican presidential nominee, warned of “a better, bigger version of the legendary Trojan horse,” declaring: “We have to stop the tremendous flow of Syrian refugees into the United States—we don’t know who they are, they have no documentation, and we don’t know what they’re planning.”

Trump’s claims are myth, not fact. Of the nearly 5.5 million people who have fled the conflict in Syria during the past five and a half years, around 10,000—less than 0.2 percent of the total Syrian refugee population—have been resettled in the US from Syria’s neighbors this year. We know who they are, because refugees are the single most vetted population entering the US. The UN Refugee Agency (UNHCR) registers, documents, and verifies the claim of all those whom it refers to the government for resettlement. The Department of Homeland Security, the Department of Defense, the Federal Bureau of Investigation, and multiple intelligence agencies then conduct interviews, gather detailed biographical and biometric data, and carry out a range of background checks on every candidate before they receive clearance to travel to the US. The entire process takes between eighteen and twenty-four months. There is no harder route into the US.

Similarly, the metaphors of a Trojan horse and a fifth column do not bear scrutiny. According to international and federal law, families and individuals who qualify for refugee status do so by proving “a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion.” Once they arrive in the US, they receive federally funded support for housing that covers no more than three months’ rent. After a year, refugees go through another security check to receive their Green Card. After five more years, they are eligible to become citizens.

Some 800,000 refugees have resettled in the US since September 11, 2001. Their record is not one of terrorism. A very small number have been charged with terrorism-related offenses, mainly concerning material support to groups abroad, rather than attacks in the US or on US citizens abroad. None have been able to carry out their plans.

Elie Wiesel, who died this past July in New York, told me two years ago: “I am a refugee, but the word ‘refugee’ is not popular. But everyone likes the idea of refuge. Fight for refuge. We all need refuge.” With America and the world now facing what can only be described as a global exodus of people fleeing war, supporting refugees is more necessary than ever.

According to the UNHCR, in 2015 there were some 65.3 million people throughout the world who had been uprooted from their homes by conflict and persecution. Over 20 million of these people are refugees, i.e., they have fled from a well-founded fear of persecution, crossed a national border, and received refugee status from either the United Nations or a state. Between three and four million of them are now in the process of claiming asylum out-side their home country. The rest are “internally displaced persons” who have not crossed national borders.

Such large numbers of displaced people have not been seen since World War II: were they a nation, it would be the twenty-first-largest on earth, the size of California and Texas combined, the same as the United Kingdom. On average, 34,000 people were forced to flee their homes every day of 2015.

The violence in Syria, which has rendered homeless half the country’s population of 23 million, bears responsibility for a large number of the world’s displaced. So do new wars and local fighting in places such as Yemen, South Sudan, Burundi, Iraq, Ukraine, Nigeria, and the Central African Republic.

In defiance of international humanitarian law, many of the belligerent governments and rebel forces target civilians and aid workers. In August several aid agencies reported deaths of civilians and medical staff in bombings of hospitals in Syria by the Syrian government and its Russian backers. Shocking acts of cruelty toward civilians have become commonplace, drastically increasing the numbers of people in flight.

The more than a dozen conflicts that have broken out or reignited since 2010 are behind much of the growth in global displacement as a result of conflict; the persistence of old wars explains the rest. Today’s conflicts burn on for an average of thirty-seven years, making return for the uprooted an ever-distant prospect. Some 2.7 million Afghans and 1.1 million Somalis have been exiled for decades.

What is perhaps less known is that the burden of sheltering today’s historic numbers of refugees has fallen, with wild disproportion, upon a small number of relatively poor states. According to the UNHCR, 86 percent of all the refugees for whom it is responsible are in low- and middle-income countries close to states in which there is violent conflict. Because of their geographical location, a mere seven countries house more than half of the world’s refugees.

The exodus from Syria has transformed Turkey into the country with more refugees than any other: over 2.7 million people have sought protection there. Lebanon shelters more refugees compared to its national population than any other state—one uprooted Syrian for every five Lebanese citizens. This would be the equivalent of the United States taking in the whole of the United Kingdom. Jordan, with nearly 660,000 Syrians known to be on its territory, and according to the Jordanian government a similar number who have yet to identify themselves to the UN, ranks third in the number of Syrian refugees. Pakistan and Iran have, respectively, 1.6 million and 950,000 Afghan refugees. A single country, Kenya, provides refuge for almost half of the more than one million people who have fled Somalia since 1991.

The wider international response to this situation—the sheltering of increasing numbers of refugees by a small group of relatively poor countries—has been singularly deficient in three ways. The first is funding. The amounts of money requested by the UN and its partners to meet the needs of the displaced, the communities that receive them, and others who rely on humanitarian aid have risen tenfold since the year 2000 (reaching an all-time high of $21.9 billion this year). But the funds actually provided by governmental and private donors over the same period, while also reaching record levels, have not kept pace.

The gap between needs and re-sources is widening, not narrowing. In 2015, the UN appealed for $20 billion in order to address global humanitarian needs; it received just $11 billion, an unprecedented 45 percent shortfall. This threatens the ability of organizations like the International Rescue Committee (the IRC, of which I’m president) and their partners to fulfill crucial responsibilities: for instance, to deliver lifesaving assistance to millions of people inside and around Syria; to provide health care, protection of women, and water and sanitation to those in need across sub-Saharan Africa; and to address both urgent and longer-term needs in more than four thousand villages across Afghanistan, where we have been working for thirty years.

Second, wealthier states—with several exceptions, such as Germany—have failed to take in meaningful numbers of refugees. Germany’s extraordinary effort during the last year, described recently in these pages, will involve processing the cases of 800,000 asylum seekers—people seeking, but not yet granted, asylum status. Notably, Eastern European countries have been drastically limiting the number of refugees.

Of the 21.3 million refugees in exile, the UNHCR estimates that 1.2 million urgently need to be moved from the countries to which they have fled, for reasons of security, dire poverty, or other special vulnerability. Despite this, the richer nations resettled just 107,100 refugees last year—less than a tenth of those deemed to be in need of additional protection. The United States took some 60 percent of these— nearly 70,000 people—and is on track to provide shelter to 85,000 refugees in 2016. The scale of today’s displacement crisis is such that the US and other advanced countries need to do much more. In a recent article in Foreign Policy, George Soros showed why Europe must work out more coherent policies for external aid and internal resettlement; and he pointed to a variety of ways in which additional funds could be made available.

Third, the practice of humanitarian aid has been undermined by the fiction—sometimes convenient for donors in the midst of financial stress and host countries concerned about taking in permanent new residents—that the problems they seek to address are temporary. In fact, refugees are displaced for an average of seventeen years, and those displaced within their own country for an average of twenty-three years. In 2014, more than 90 percent of countries appealing for humanitarian aid had made appeals for three years or longer and 60 percent had made appeals for more than eight years. Meanwhile the median length of a grant from the IRC—we manage around seven hundred different grants at any one time—is twelve months.

This fixation of donors on short-term solutions has led to a systematic under-investment in programs that re-quire longer-term planning. Refugee education has been a prime victim. In Lebanon over 200,000 Syrian children remain without education five years after the outbreak of the civil war that has driven them from their homes. Around the globe, of the 61 million children under eleven who cannot go to school, it is estimated that 35 percent are living in countries affected by conflict.

The consequences of political and policy failure are not only felt by refugees. Neighboring countries become destabilized by their presence, as do parts of Europe. Humanitarian needs arise from political instability, but when those needs are not addressed they cause further political disruption. The need to rethink aid policy is over-whelming, not least because all the evidence suggests that the root causes of the current refugee problem—including both religious and ethnic politics and geopolitical divisions—are unlikely to be addressed in the short term.

Aside from the fact that people are being displaced for long periods, two other striking conditions should affect planning for the future. The first is that 59 percent of refugees are currently found in urban areas. Refugees are often living among the local population, not separate from them. The second is that at least 43 percent of the world’s extreme poor are now found in fragile countries that are either in the midst of internal conflict or suffering from the effects of nearby conflicts. So the focus of humanitarian policy and development policy, traditionally seen as separate by aid donors and workers—one to deal with crisis and the other with poverty—increasingly over-lap. There should be a major change of approach in four principal ways.

First, donors need to accept the idea that the best approach to helping refugees is often not to deliver tents or food but instead to give them money. A study in 2015 by the Overseas Development Institute summarized well the case for greater use of cash assistance. The institute concluded that “cash transfers give people choice and make humanitarian aid more account-able” to the people who need the aid. The transfers can also “help to make scarce resources go further,” and can provide a base for increasing “opportunities created by the global expansion of financial services.” Yet today less than 6 percent of humanitarian spending is given out in some form of cash, which now includes debit cards, secure electronic forms of financial support, or non-electronic vouchers.

In Lebanon, the IRC conducted the first study to compare refugees receiving cash with those who did not. The results were striking. Child labor went down; school attendance went up; prices did not rise; and each dollar given to refugees generated a $2.13 in-crease in GDP in the local economy.

Cash is not the answer everywhere. But where markets are functioning it can have powerful effects. In rural Pakistan the IRC has just tested a new method of identifying beneficiaries and distributing cash that links delivery of aid to a government database. This method has the potential to result in major reductions in the time taken to distribute money as well as in the cost of setting up the system. With this new model we will be able to reach 38 percent more people with the same amount of aid. “Why not cash?” should be one of the first questions in any humanitarian situation.

Second, refugees need to be given the chance to earn a legitimate living. Uganda, which has the eighth-largest number of refugees in the world, shows the benefits of treating refugees as productive residents. Over half a million Congolese, South Sudanese, Burundians, and Somalis live there. As the Refugee Studies Centre at Oxford and other sources have documented, they are granted a plot of land to farm and are able to legally work, move where they want, and choose where to live. They can use public services, send their children to public schools, improve their skills, and contribute to the Ugandan economy. In Kampala, 78 percent of refugees need no aid, and nation-wide only 1 percent are completely dependent on aid.

It is crucial to bring into close alignment the interests of refugees and host communities. In the future, international financial institutions such as the World Bank need to deliver major economic support to countries hosting refugees, and those countries should allow refugees to do paid work, including through the granting of work permits. The president of the World Bank, Jim Yong Kim, has responded strongly to the need for such reforms, agreeing that the bank should freely conceive its mandate so that it can help middle-income countries that have admitted large numbers of refugees. Such a man-date should link broad “economic and local development” directly with humanitarian aid. In February, Kim said:

Jordan and Lebanon—and the rest of the region as well as Turkey— are faced with a development challenge, not only a humanitarian crisis.... We should be exploring how we can use innovative financial tools...for middle-income countries [i.e., financing at lower than market rates], and insurance products for low-income countries.

The World Bank Group is now leading a project to support Jordan and Lebanon with special “concessional” loans, which come with much lower interest and longer grace periods than those available on the market. Donor grants will be used to lower interest payments on loans to Jordan and Lebanon, for example, from development banks. The plan is to raise $1 billion from donor countries that will make possible between $3 billion and $4 billion of concessional finance. In return, at the London Conference on Syria in February, Jordan, Lebanon, and Turkey promised to remove barriers so that nearly one million refugees will be able to hold jobs.

The “compact” struck with Jordan in particular offers lessons for other countries with large numbers of refugees. The World Bank Group and the EU offer Jordan a comprehensive assistance package, including interest-free loans and tariff- and quota-free exports to the EU for manufacturers willing to employ an agreed-on percentage of Syrians. In return, the government in Amman will be expected to work with international partners to create 100,000 job opportunities over the next five years for Syrian refugees. The IRC is currently working to examine job creation strategies that could help Jordan meet this goal.

Third, there needs to be a concerted drive to determine what types of aid programs deliver maximum value for money. This may seem obvious, but to date there have been only around a hundred evaluations of the impact of aid policies and practices in places ridden with conflict, compared to more than four thousand such evaluations of aid to poor but stable countries. This means organizations are relying too much on anecdotal experience and in-tuition rather than hard evidence when it comes to reducing child mortality, preventing family violence, promoting refugee employment, or providing education.

The power of effective research was shown by the IRC’s recent study of ways to prevent violence against children. The prevailing approach had been to use “community committees” to educate adults and protect children. Studies by the IRC in Burundi, Liberia, and Thailand, however, showed that working directly with individual caregivers to improve parenting practices and their understanding of their children’s needs could reduce violence against children by up to 64 percent.

Fourth, those working in countries facing humanitarian crises need to channel their efforts toward achieving a limited number of specific, clear, measurable outcomes. The effectiveness of such an approach was underscored this year in UN Secretary General Ban Ki-moon’s report “One Humanity, Shared Responsibility.” Referring to the need for a system that brings different organizations with different concerns together, he called for “collective outcomes that are strategic, clear, quantifiable and measurable.” He went on to say:

The articulation and achievement of such collective outcomes will allow a range of diverse actors— national and local authorities, humanitarian, development, human rights and peace and security actors, and possibly even private enterprises—to work together towards a common goal.

The idea that quite different organizations will focus on common goals and create accountability and pressure for results has recently proved its worth, as can be seen in the record of the fifteen- year drive by the UN to reach the eight Millennium Development Goals (MDGs) agreed-on by member states in 2000. The MDGs not only created a clear set of internationally agreed-on goals, but also identified numerical targets to achieve them. For example, the goal to “eradicate extreme poverty and hunger” had a target: “Halve the proportion of people whose income is less than $1.25 a day.” Alongside strong economic growth in India and China, this global commitment by a variety of organizations contributed to important progress. The target of reducing extreme poverty rates by half was achieved in 2010, five years ahead of the 2015 deadline. The successor to the MDGs, the Sustainable Development Goals—adopted one year ago this September—are intended to continue this progress.

The danger, however, is that those living in countries affected by the crises of war and homelessness will be left behind because, for them, there are no specific targets of the kind advocated by Ban Ki-moon. I have seen this in practice. Earlier this year I met some of the 150,000 Burundian refugees in Tanzania. An average classroom in Nyragusu camp has 150 children. As women use the camp’s unlit paths to collect firewood or use the lavatories, they are often subject to sexual violence. One explanation for such dangerous conditions is that the overall plan to provide humanitarian assistance for Burundian refugees is only 40 percent funded. Another is that the roughly twenty-eight humanitarian agencies working in Tanzania—with over one thousand aid workers—still had not yet agreed on a shared set of measurable objectives for their efforts and how they would achieve them.

Specific goals for countries in the midst of crisis would improve global accountability for the world’s most vulnerable: those affected by war and disaster. It is high time that the UN and member states agreed on specific tar-gets for safety, health care, education, and income for those directly affected by conflict. As one student in the Nyragusu camp told me, “I don’t want my future to end here.” For him, and the millions more like him, we must try to make sure it doesn’t.

On September 19, the UN Summit for Refugees and Migrants, called by Secretary-General Ban Ki-moon, will offer UN members the opportunity not only to increase their commitments to collective goals, but to articulate an agenda for improving aid. Alarmingly, however, the draft conclusions of the conference have been watered down by some of the participating nations. Phrases like “We commit to working towards solutions from the outset of a refugee situation” are common-place in the document but do nothing to concretely improve the situation of refugees. It is therefore even more important that the refugee summit for heads of government called by President Obama for September 20, the day after the UN meeting, have tangible effects. It would be a positive start to secure commitments for an additional two million refugee children to receive education. We should also call on the US to raise the number of US refugee admissions to 140,000 for next year.

Ask most refugees what they want most and they will likely say: “To go home.” The long-standing humanitarian aid system was designed for a world where wars between states displaced refugees for short periods of time into refugee camps before they did go home. Today none of these assumptions hold. Many refugees are victims of civil wars, not wars between states. They are dis-placed for long periods. They do not stay in refugee camps. They are unlikely to go home.

Fewer than one percent of the world’s refugees were able to return to their countries of origin last year. The peace-making and peacekeeping efforts that have been recently successful in Sierra Leone and Timor-Leste and between Serbia and Kosovo are complex, expensive, risky, and time-consuming—and exceptional. Protracted conflict is the new norm. In these cases, there is every reason radically to change our approach to direct support for the victims. More of the most vulnerable refugees need to be relocated into richer countries. But most displaced people need far better help in the nearby poorer countries to which they flee. As in the 1940s, the longer the delay, the worse the reckoning.