Poverty in the Era of Data

The more you know about your enemy, the easier it is to beat it. This is true for wars against armies, diseases or corruption — good information will tell you where to best deploy your soldiers, doctors and auditors. The same happens in the fight against poverty. You want your resources to go where they help the poor the most. For that you need accurate, frequent, timely, comprehensive, comparable, consistent, and accessible data. And that is exactly what we are beginning to get. In fact, data is transforming the development profession — you can call it the revenge of the statisticians.

Here is how that transformation is happening. First, the funding of multilateral institutions — like the World Bank — is now more closely linked to the results they promise to achieve. To get money from taxpayers, they have to commit to specific “goals.” How much will infant mortality fall? How many children will be vaccinated? How many girls will learn to read and write? What proportion of women will have access to contraceptives? By when? All this is creating a culture of monitoring and evaluation — “M & E,” in development parlance — that is bringing light to what works and what does not work. For that, you need data.

Second, data is doing to public subsidies today what privatization did to public enterprises two decades ago: it is lifting the veil of inefficiency. With better household surveys, we can identify who exactly benefits from every dollar the government spends — not surprisingly, this is called “benefit incidence analysis.” Take education. Most developing countries spend more bankrolling free public universities than building primary schools. But the main beneficiaries of that subsidized college education are the rich (who could pay tuition) not the poor (who could not). You see the inefficiency? The same applies for subsidies to gasoline (who owns cars?), electricity (who has larger houses?), or pensions (who has formal jobs?). Statistics lets you quantify these aberrations — and argue that the money should be redirected towards those that really need it.

Third, better data is allowing us to focus on poor people’s non-cognitive skills. You see, whether you get a good job — or any job — does not only depend on how many exams you pass, how much you know, or what your IQ is. It also depends on things like how conscientious you are, how you react to new experiences, or how well you interact with others — think of it as the “non-cognitive” side of your resume. Is it better to be smarter or to be on time? To know more or to listen more? To be trained or to be trainable? Household surveys are beginning to gather information that will, one day, allow us to answer those questions — Peru is a leader in this among emerging economies. And when we get the answers, we will be able to design educational curricula to teach not just the concepts, but also the behaviors that make people more productive.

Fourth, it is possible to determine how personal circumstances affect human opportunity. We all know that children have no control or responsibility over their gender, skin color, birthplace, or parents’ income. And yet, those kind of circumstances are sure-shot predictors of a child’s access to vaccination, potable water, kindergarten, the internet and many other platforms without which her probability of success is close to nil — well before she can make any choice by herself. This can now be measured, something that was impossible only a few years ago. The measure is called “Human Opportunity Index,” and is beginning to change how social policy is designed.

Finally, we have broken the taboo of experimenting with people. It is no longer unusual for researchers to walk into a slum, offer child care to a sample of mothers, and then monitor whether they work outside the home more hours than those with no child care at all. (FYI: they don’t always do). These type of “randomized trials” are proving really useful to assess what policies and what projects work best — and which are a waste of time and money. From giving cheap fertilizer to farmers to making cheap loans to female micro-entrepreneurs, you can evaluate anything, as long as you have — or create — the data.

By now, you are probably wondering where all this data is going to come from. Isn’t it true that most national statistical offices in the developing world are somewhere between weak and very weak? After all, those all-important household surveys, when they exist, get published years after they are collected. Millions are rightly being spent to upgrade statistical capacity. But it takes a long, long while before you see results — which explains why politicians rarely care about it.

Is there a short-cut? Is there a fast way to get the data we need to help the poor? Yes, and it is probably sitting in your pocket, in your purse or on your belt. It is your cell-phone. It turns out that people will happily sign up to answer a couple of short phone surveys a month in exchange for “free minutes” of phone use. How many minutes? On average, less than five dollars worth of minutes per month. (Yes, that’s how cheap we all are.) This is a bargain because you don’t have to call more than a tenth of 1 percent of your population to get a good reading of how your country is doing.

As the use of cellular telephony expands among the poor — at flash speed in places like Kenya –the possibility of turning them into data sources becomes real. In fact, some of this is already happening in Latin-America, and may soon catch on in Africa. Others will surely follow. How ironic that, in the end, the war against poverty may be won when those who try to help the poor get to literally listen to them.

Follow Marcelo Giugale on Twitter: www.twitter.com/@Marcelo_WB

#economicdisparity #PovertyData