​History and literature are littered with frustrated attempts to do the right thing. Famously, Dorothea in George Eliot’s Middlemarch has the good intentions of improving the lot of tenant farmers. She fails because she is insufficiently pragmatic. How to make sure that the same doesn’t happen to the aspiration to ‘leave no one behind’, which underpins the Sustainable Development Goals (SDGs)?

The number one job is to keep it on the political agenda. UN Secretary-General Ban Ki-moon recently proposed that the first moment on SDG accountability focus on ‘leave no one behind.’ This is a good sign. It suggests the international community is serious about turning a lofty vision into practice.

Here at ODI, we’re making a start next week by hosting a high-level event where the Dutch Minister for Foreign Trade and Development Cooperation will speak about the importance of donors taking up this agenda. The Costa Rican vice president will set out what her government is already planning to do differently since the UN member states enshrined ‘leave no one behind’.

But once it’s on the agenda, the next job, as Dorothea found, is to convert the noble aim into real life deeds.

The good news is that in several important ways, we have evidence that interventions work to reverse entrenched poverty. Since 2004, ODI’s Chronic Poverty report series has explored successful pro-poorest policies.

Social protection will be a significant part of the solution. So will free universal health coverage, even if countries only start with a very basic package of medicines and treatment at every clinic.

Often we know how to reach the income-poor, but not the marginalised.

We need more analysis of how to overcome the multiple obstacles, discrimination and barriers faced by excluded groups, and how to make sure people in those groups are integrated into the economy and can earn their own living.  This is research that institutions like ODI will need to do in the next few years to give governments and NGOs the evidence required.

The first stage for these, and any other actions, will be to identify just who is being left behind and what that marginalisation looks like in practice.

Our new research offers a start.

We’ve just developed a methodology that attempts to do just this in a few countries around the world. It’s yielded some interesting findings. Differences between peoples’ lived experiences of deprivation and exclusion may seem fine-grained in theory. In reality, they call for differentiated policy responses.

For instance, in Nigeria, Fulani people were eight times less likely than Yoruba to have access to sanitation, three times less likely to have had a substantial education, and more than twice as likely to belong to the bottom wealth quintile in 2013.

In 2004, Afro-Brazilians in the northeast had more than a 20 percentage point higher probability of being in poverty than those in the centre-west. By 2012, this gap had narrowed in absolute terms, but the relative gaps widened.

In Bangladesh there are significant differences in outcomes between poor, single women (de jure heads of household) and poor women whose husbands are usually away (de facto ones).

More interesting still were the holes the research uncovered. The authors weren’t able to draw conclusions about the difference between ethnic groups in Vietnam because the sample sizes were too small. Nor were they able to tell whether they were focusing on the right people and issues: mental health problems may have been holding back progress in Guatemala or Benin, but we didn’t have the data to tell. The exercise was, in other words, constrained by data gaps.

To leave no one behind, this counting exercise will be essential.

For governments to be able to fully count their left behind, they need to combine this kind of data work with big data; monitoring and evaluation by Civil Society Organisations working with minority groups or with difficult-to-reach communities; and using participatory research listening to the people in those communities.

Of course, there are countries where the government itself institutionalises marginalisation (think of Uganda’s anti-homosexuality law, for example). Here the global nature of the SDGs will be particularly important, as the international scrutiny and civil society pressure they bring to bear should make it harder for governments to actively discriminate against, or just ignore the needs of, significant parts of their population.

Delivering affirmative action on a global scale can’t, of course, really be compared to housing George Eliot’s fictional agricultural workers. But the latter’s message is a universal one: ambitious aims need a workable plan if they are to succeed. If the world really is to leave no one behind, this kind of counting exercise will be vital.

For more information about how we can achieve the 2030 sustainable development agenda visit deliver2030.org