Many readers of Nature will take it for granted that they have a birth certificate, and that when they die, their death, and its cause, will be officially recorded, as will their health problems in the intervening years. When aggregated, such data allow researchers to estimate disease burdens and risks to help shape public-health policies and investment in everything from high blood pressure to infectious diseases — and to monitor the impact of disease control efforts.

Yet more than 100 countries, and not just the poorest, lack even basic birth and death registration systems. Furthermore, only 34 nations — covering just 15% of the global population — generate decent cause-of-death data, and even some of those data are unreliable because doctors have not correctly assigned the cause of death.

There is a shocking lack of national and international political will to invest in the basic statistical systems needed to track this most fundamental information. Bodies such as the World Health Organization (WHO) continue to push out charts of global trends. These are handy for advocacy purposes, but the underlying data are often scarce and poor.

Initiatives such as the Global Burden of Disease study — published in The Lancet last December by an international consortium led by the Institute for Health Metrics and Evaluation in Seattle, Washington — have helped. They have sucked up what data are available from demographic health surveys, papers and other sources, and brought unparalleled scientific expertise and advanced modelling to bear on extracting meaning from the sparse and heterogeneous data — and filled in gaps where no data exist at all (see Nature 492, 311–312; 2012). But even the researchers involved are the first to admit that this situation is far from ideal, and that what is really needed is more and better raw data.

The issue of how to improve global health estimates was the subject of a two-day meeting convened in Geneva, Switzerland, last week by the WHO. Many people thought the meeting was constructive, although the consensus recommendations that emerged — for the WHO and academics to collaborate more closely; increased investment in registration systems and training; and better sharing of data and methods — will need to be accompanied by consolidated political commitment to gathering health metrics.

The size of the field is incommensurate with the immense task at hand.

Although their intergovernmental nature and direct contact with ministries mean that the WHO and other United Nations (UN) agencies are essential players in getting better registration systems, they can also be part of the problem. Numerous agencies are involved in health metrics, but they are largely uncoordinated, overly bureaucratic and politicized and too oriented towards defending their turf. No one agency is responsible for promoting civil birth and death registration.

The latest disappointment is the Health Metrics Network (HMN), a WHO-hosted partnership of international organizations created in 2005 to boost civil registration health data with US$50 million from the Bill & Melinda Gates Foundation. Despite a promising start, observers say that there have been few accomplishments to show for the money, and the WHO dissolved the network last November.

That makes the goal to boost civil registration systems more necessary than ever. The new reality is that most of the expertise in health estimates is no longer within the UN; it is in academia. Nature has learnt that at the same time as the WHO meeting in Geneva, other leading scientists in the field were meeting with philanthropists in New York on how to replace the HMN with a new organization — one that would not be hosted within the WHO.

A fundamental problem is that the size of the field is incommensurate with the immense task at hand, and that is further complicated by intense competition for limited funds. The community must work to better present its very justified case for greater political attention and funding — and for a much needed injection of fresh blood and expertise, especially with a national focus.

Given the information technology of the twenty-first century, it is simply unacceptable that the relatively cheap and simple registration systems needed to gather data on births and causes of death on a continuous basis are absent across much of the planet. The development of such systems is largely the responsibility of individual nations, but greater political attention is needed at both the national and international levels to make it happen. A good place to start would be placing the seemingly mundane, yet crucial, issue of civil registration systems higher on the agenda of organizations such as the G20.