Whose needs did we meet? Disaggregating humanitarian data

By Kate Burns on 8 October 2012.

View existing comments or leave a new one!

Whose needs are being met in humanitarian action? Should be simple to answer, but in reality we just don’t know. Why don’t we know? The short answer: humanitarian actors are primarily output-focused people. How many latrines were built? Shelters constructed? Patients seen at a health clinic? These are our main success indicators. While we collectively move to try and measure the impact of our humanitarian action, we must also answer the question: whose needs did we meet? This requires the collection, analysis and use of sex and age disaggregated data (SADD) so we can say whose needs were met, and who fell through the cracks.

I am constantly mystified by an absolute unwillingness to collect SADD. “It’s too hard,” people say. It makes me upset but also angry because without the collection of SADD we cannot know the impact of our efforts in meeting the needs of the people affected by crises. A recent study – Sex and age matter by the Feinstein International Center and Tufts University – supported by OCHA and CARE, gives several powerful examples of how gathering this data early can make a real difference (Mazurana et al, 2011). 

The collection and analysis of SADD is key to gender analyses and gender-sensitive humanitarian programming and, therefore, to measuring the effectiveness of humanitarian response. A lack of SADD makes it more difficult to understand people’s differentiated needs and hampers the ability of aid workers to monitor the impact of assistance. If, for example, a health centre simply reports seeing 10,000 clients a month, humanitarian responders cannot tell whether there are more women than men accessing its services, or whether there are issues to be resolved around men’s or women’s access to health care. Similarly, if a school simply says it has 20,000 students, it will be unclear whether there are more boys than girls attending that school, or if more girls than boys are dropping out and at what stage. That makes it much more difficult to address emerging gaps and to tailor services to address them.

Without SADD, humanitarian organisations will be unclear about the effectiveness and, accordingly, the appropriateness of their assistance. In DRC, in 2011, for example, data on malnourished children was initially not broken down by girls and boys. A Gender Adviser urged a closer look. The new analysis showed that more boys than girls were malnourished, but more girls than boys were coming to supplemental feeding centres. Aid agencies working in the nutrition sector were surprised at this finding and plan to revise their plans accordingly. To design gender-sensitive projects and to deliver humanitarian relief effectively, aid workers need to understand how needs differ by sex and age and more emphasis will be required to build skills and hold actors accountable on not only generating such data but using it to inform programming approaches and measuring their failures and successes.

Examples abound on the lack of data disaggregated by sex and age. I recently read the latest ALNAP State of the Humanitarian System Report and was drawn to page 48. Initially excited to see data from aid recipients regarding their opinions on the services provided and the degree to which they were consulted. The results are bleak – in many cases more than 60% of people affected by the crisis were not consulted. But the data is not disaggregated be sex so we have no idea if both men and women were asked these questions.

My gut tells me that if this data were disaggregated that we would see a very different picture. Women are usually not included in assessments of needs and therefore their opinions are not registered. It would be great to hear from people on the ground about their challenges in collecting and using sex and age disaggregated data and examples of when they did use it - how that changed the way they delivered services.

It is our responsibility to meet everyone’s needs. The only way to be sure is to collect, analyse and use sex and age disaggregated data – if not we are left in the dark about the effectiveness and impact of our efforts.

• Share this page:
• Tweet
• Email
• Print

To:
From:
Message:	I found this on the ALNAP website and thought you might find it interesting.

3 comments

Jerri Ann Husch (2Collaborate Consulting) 8 October 2012, 13:37

Kate! What a great blog piece! So, so good to see you online! You have stated very well the kinds of issues that a team of us have been seeking to "find an answer" to---and we have come up with a method---"action intelligence" that we have been field testing with UNDP in Africa around the issue of climate change adaptation. Action Intelligence is a research method that allows for "disaggregation"---it is a process that takes data and allows us to see (literally, using a variety of visual analytic tools) who is doing what, where and when, in time and place---at a micro, mid and macro level---thus creating a reality based understanding of the social context. I would love to share more about this with you as you have so succinctly defined the need! see http://www.inthum.org/ for a beta view of the process...

Meanwhile--it's been too, too long---time to catch up!!

best, Jerri

Abby Stoddard (Humanitarian Outcomes) 10 October 2012, 15:23

As Kate correctly points out, for the State of the Humanitarian System (SOHS) report we did not include questions on age and sex in our original survey template for recipients, although we do have sex-disaggregated data in the surveys from Pakistan, because it was done as part of a larger, post-distribution survey by the NGO partnering with us.

In attempting the first such "satisfaction" survey done for aid recipients in SOHS, we were primarily concerned with how to access recipient populations and collect responses in a methodologically sound way within a limited timeframe and budget. As it happens, the breakdown of responses from Pakistan (50:50 men and women) were fairly consistent between men and women. There there were no notable sex-based differences in the proportions of those answering yes or no to the broad questions we asked about aid - but of course that is just one case.

We certainly recognise the importance of collecting disaggregated data whenever possible, and would support disaggregating response data by sex in any future expansion of the the pilot recipient survey in the SOHS. Our only caveat would be that for such surveys to be effective they need to be short as possible (especially the SMS-based ones), so each question must be weighed carefully for its potential to add important information to the analysis.

Abby Stoddard and Paul Harvey, Humanitarian Outcomes, members of SOHS Research Team

Kim Scriven (ALNAP) 11 October 2012, 10:34

Thanks very much for this interesting and important blog Kate. One of the things that has always struck me is how, in the absence of good data on the populations they seek to assist, agencies and policy makers often make assumptions about these populations, their behaviours and their needs. This was reinforced for me yesterday when I was alerted by Laura Seay (who blogs as Texas in Africa) to the Simon Fraser University Human Security Report for 2011/12. It's worth quoting Laura at length:

[i]"Essentially, SFU found that the data shows much of the conventional wisdom on such issues as rape as a weapon of war, who is committing rape in wartime, and negative effects on education is completely wrong. Among the report's findings:

   • Conflicts in which extreme sexual violence is committed (think DRC) are exceptional outliers, not the norm.
   • While reporting of sexual violence in wartime has increased, there is no evidence to support the oft-repeated-by-high-level-UN-officials claim that incidences of wartime sexual violence are increasing.
   • Strategic rape incidences, aka "rape as a weapon of war" are not increasing, either.
   • Domestic (household & intimate partner) sexual violence is by far the most prevalent form of sexual and gender-based violence in wartime.
   • Male victims and female perpetrators of rape in wartime may be greater than previously believed.
   • That statistic that 3 in 4 Liberian women were victims of sexual violence during the country's war? No evidence whatsoever for the claim. The real rate of lifetime sexual violence in Liberia is more like 18% - exactly the same as the rate of SGBV in the United States - which means it's impossible that 75% of women were raped or otherwise sexually assaulted during the wars.
   • There is no evidence to support the UN claim that sexual violence committed against children in conflict-affected countries is increasing.
   • Conflict doesn't have a net negative effect on educational outcomes."[/i]

As Laura points out, these findings are more than a little controversial, undermined many of the prevailing narratives about sexual violence in times of war, and will doubtless be the subject of much debate. What it underlines for me is the importance of making the effort to gather credible evidence on which to base policy decisions, whether dealing with SADD, SGBV or for that matter any other issue.

Leave a comment