every aid worker in Africa

comes to a time in her career

when she wants to take all the money for her project —

maybe it's a school or a training program —

pack it in a suitcase,

get on a plane flying over the poorest villages in the country,

and start throwing that money out the window.

Because to a veteran aid worker,

the idea of putting cold, hard cash

into the hands of the poorest people on Earth

doesn't sound crazy,

it sounds really satisfying.

I had that moment right about the 10-year mark,

and luckily, that's also when I learned

that this idea actually exists,

and it might be just what the aid system needs.

Economists call it an unconditional cash transfer,

and it's exactly that: It's cash given

with no strings attached.

Governments in developing countries

have been doing this for decades,

and it's only now, with more evidence

and new technology that it's possible

to make this a model for delivering aid.

It's a pretty simple idea, right?

Well, why did I spend a decade doing other stuff

for the poor?

Honestly, I believed that I could do more good

with money for the poor

than the poor could do for themselves.

I held two assumptions:

One, that poor people are poor in part

because they're uneducated and

don't make good choices;

two is that we then need people like me

to figure out what they need and get it to them.

It turns out, the evidence says otherwise.

In recent years, researchers have been studying

what happens when we give poor people cash.

Dozens of studies show across the board

that people use cash transfers

to improve their own lives.

Pregnant women in Uruguay buy better food

and give birth to healthier babies.

Sri Lankan men invest in their businesses.

Researchers who studied our work in Kenya

found that people invested in a range of assets,

from livestock to equipment to home improvements,

and they saw increases in income

from business and farming

one year after the cash was sent.

None of these studies found that people

spend more on drinking or smoking

or that people work less.

In fact, they work more.

Now, these are all material needs.

In Vietnam, elderly recipients used

their cash transfers to pay for coffins.

As someone who wonders if Maslow got it wrong,

I find this choice to prioritize spiritual needs

deeply humbling.

I don't know if I would have chosen to give food

or equipment or coffins,

which begs the question:

How good are we at allocating resources

on behalf of the poor?

Are we worth the cost?

Again, we can look at empirical evidence

on what happens when we give people stuff

of our choosing.

One very telling study looked at a program in India

that gives livestock to the so-called ultra-poor,

and they found that 30 percent of recipients

had turned around and sold the livestock they had been given

for cash.

The real irony is,

for every 100 dollars worth of assets

this program gave someone,

they spent another 99 dollars to do it.

What if, instead, we use technology to put cash,

whether from aid agencies or from any one of us

directly into a poor person's hands.

Today, three in four Kenyans use mobile money,

which is basically a bank account that can run

on any cell phone.

A sender can pay a 1.6 percent fee

and with the click of a button

send money directly to a recipient's account

with no intermediaries.

Like the technologies that are disrupting industries

in our own lives,

payments technology in poor countries

could disrupt aid.

It's spreading so quickly that it's possible

to imagine reaching billions

of the world's poor this way.

That's what we've started to do at GiveDirectly.

We're the first organization

dedicated to providing cash transfers to the poor.

We've sent cash to 35,000 people across rural Kenya

and Uganda

in one-time payments of 1,000 dollars

per family.

So far, we've looked for the poorest people

in the poorest villages, and in this part of the world,

they're the ones living in homes

made of mud and thatch,

not cement and iron.

So let's say that's your family.

We show up at your door with an Android phone.

We'll get your name, take your photo

and a photo of your hut

and grab the GPS coordinates.

That night, we send all the data to the cloud,

and each piece gets checked

by an independent team

using, for one example, satellite images.

Then, we'll come back,

we'll sell you a basic cell phone

if you don't have one already,

and a few weeks later,

we send money to it.

Something that five years ago

would have seemed impossible

we can now do efficiently

and free of corruption.

The more cash we give to the poor,

and the more evidence we have that it works,

the more we have to reconsider

everything else we give.

Today, the logic behind aid is too often,

well, we do at least some good.

When we're complacent

with that as our bar,

when we tell ourselves that giving aid

is better than no aid at all,

we tend to invest inefficiently,

in our own ideas that strike us as innovative,

on writing reports,

on plane tickets and SUVs.

What if the logic was,

will we do better than cash given directly?

Organizations would have to prove

that they're doing more good for the poor

than the poor can do for themselves.

Of course, giving cash won't create public goods

like eradicating disease or building strong institutions,

but it could set a higher bar

for how we help individual families

improve their lives.

I believe in aid.

I believe most aid is better than just

throwing money out of a plane.

I am also absolutely certain

that a lot of aid today

isn't better than giving directly to the poor.

I hope that one day, it will be.

Thank you.

(Applause)