18 April 2015

Why Has Stunting Declined So Fast in Tanzania? Poverty Reduction & Nutrition Spending Increases Matter

I just returned from a week in Tanzania, exploring options for evaluating the impact of m-nutrition initiatives on nutrition behaviours and status.  More on that as it develops.

During the trip I found out about a recently released 2014 SMART national nutrition survey that shows under 5 stunting rates at 34.7% in 2014, down from 42% in 2010.  That is 7.3 percentage points in 4 years. Good news.  This is an annual average rate of reduction of nearly 5% -- well above the nearly 4% required to meet the World Health Assembly targets.  If this can be maintained then this is even better news.  But there seems to be no consensus as to what is driving the decline.

I don't know of a study that looks into this for Tanzania (my IFPRI colleague Derek Headey has done these kinds of studies in Ethiopia, Bangladesh and Nepal), so I went to the Global Nutrition Report's Tanzania Nutrition Country Profile.  The profile shows that the rates of food security, improved water and improved sanitation are all increasing, but very slowly.  Unimproved sanitation + open defecation rates remain very high (79% in 2012).  Economic growth is steady but not spectacular.  The big change seems to be $1.25 a day poverty rates which have almost halved in 12 years: 85% in 2000 to 43% in 2012. 

As you can see from the link to the 2014 survey results, infant and young child feeding has not improved and for exclusive breastfeeding rates have actually decreased.  Iron Folate supplementation rates have increased, Vitamin A supplementation rates for under 5's have increased and women's thinness has declined:  but none of these positive changes are huge.

The Government of Tanzania's health expenditure has increased substantially between 2008 ($383 million) and 2014 ($622 million).  The 2014 Nutrition Public Expenditure Review is even more revealing. As the picture below shows, the total budget for the nutrition sector has increased rapidly.  Equally important the share of the government in that expenditure, although quite low at about 30%, has held steady, so real GoT expenditure on nutrition is also increasing rapidly.

This is the really good news story--an increasing government commitment to nutrition, not only in words, but also in cold hard cash that has multiple alternative uses.  Of course, money is not the only important resource, and it has to be spent on things that can improve nutrition, but scale up of programmes is impossible without it.

So, while we need a proper study to confirm it, maybe, just maybe, it is the declining poverty rate combined with increased nutrition spending by government and external partners, that is responsible for the decline in stunting in Tanzania.  And more good news: there is plenty of room for improvement in water, sanitation and infant and young child feeding programme coverage rates. 

Can Tanzania meet its WHA targets? If these trends are maintained the situation seems very positive. 


Lawrence Haddad said...

Hi Lawrence,

Very interesting case study! And also very encouraging!! I fully agree that one very good news is the poverty reduction and another is the increase in the public nutrition expenditure and that they may well be the main drivers of the decline in stunting rate.

I'd like to add 3 comments on the Tanzanian situation that you described:

1. As you know, I'm quite interested in following stunting figures in Africa, and we often have troubles to get comparable data. In fact, as far as I know the latest national stunting rate available for Tanzania doesn't com from the DHS in 2010 (42.5%) but from the Tanzania National Panel Survey of 2011, which showed a 34.8% stunting rate (i.e. almost the same as the 34.7% figure from the latest MICS in 2014, but also, then, a dramatic decrease of 7.7% between 2010 and 2011).
Once again this raises the question of the comparability of survey figures and methodologies. Both the 2010 and 2011 estimates are in the Joint child malnutrition estimates produced by the Unicef-WHO-The World Bank consortium (http://www.who.int/nutgrowthdb/estimates/en/).
Colleagues in this consortium are really doing a great job but they lack means to undertake further analyses that could help explain this kind of stituation and I definitely think they should be supported more actively by donors.

2. Whatever the survey we consider and the actual timing of the stunting decline in Tanzania, there seems to be a real decline and this is very good news. However, efforts are still needed to reach the WHA target. According to the tool available at http://www.who.int/nutrition/trackingtool/en/ if current trend is maintained there will still be more than 1 million stunted children in excess of the WHA target, because of the demographic trend. The Government of Tanzania has made many efforts but should be encouraged and supported to even accelerate them. And as you mentioned there is room for improvement in many of the underlying causes of malnutrition (hygiene and sanitation, IYCF practices, to name a few).

3. The last point of your post resonates like a call for further exploration of the drivers of the improvement in the nutrition situation. I cannot agree more. For this, I would suggest to take advantage of the National Evaluation Platform run by the Goverment of Taanzania, with support from colleagues at Johns Hopkins University, in a project funded by Canada (http://www.jhsph.edu/research/centers-and-institutes/institute-for-international-programs/current-projects/national-evaluation-platform/tanzania/index.html and http://www.jhsph.edu/research/centers-and-institutes/institute-for-international-programs/current-projects/national-evaluation-platform/documents/NEP-brochure-Tanzania.pdf)
As it was flagged in the GNR 2014, those National Evaluation Platforms have a real potential for informing decision making in nutrition (http://globalnutritionreport.org/files/2014/11/gnr14_pn4g_20bryce.pdf) and it would be really interesting to hear from this platform about changes in the context, child health and survival.

Lawrence Haddad said...

Previous comment was from Yves Martin Prevel, IRD and Nutripass, also a member of the Independent Expert Group of the GNR

Biram Ndiaye said...

Dear Lawrence,

Thank you very much for your comments and questions on the results of the 2014 Tanzania National Nutrition Survey. Indeed, I was informed by UNICEF ESARO that you were in Tanzania last week, but unfortunately I did not had a chance to meet you.

This survey was initiated by UNICEF and Tanzania Government because we could not rely on DHS surveys every five years. If we want to be serious on stunting reduction, we need more frequent surveys between two DHS as you rightly point it in the GNR. Therefore the objectives of the survey were to assess nutritional status of children aged 0-59 months and level of infant and young child feeding practices and coverage of micronutrient interventions in Tanzania.

A SMART methodology was used with a budget of 460,000 USD covering 30 regions i.e. less than 15,500 USD per region. To me, it is important to have a cost effective approach if we want to have some buy-in on more frequent national nutrition surveys between DHS. The objective is to have a National survey every two years. With this approach, we deliberately collected a limited number of data and limited the level of representativeness to regions. Some people were pushing to include more variables and to have data representatives at district level. There is 180 districts in Tanzania, if we want to cover all of them, the budget would be 2,700,000 USD. Including additional data to collect would further increase the budget and delay the reporting. Therefore the objectives were more to assess trends in key nutrition indicators. For more comprehensive data collection it is better to wait for DHS 2015.

Given all the questions on the drivers of the decline of stunting in Tanzania, UNICEF has decided to support TFNC to carry out more in depth analysis on the determinants of stunting reduction. However, given the limited number of indicators collected during the 2014 national survey, the analysis will be incomplete as the survey was not designed for that purpose. Parameters like assets ownership, socio-economic status, maternal or paternal education, ante natal care, birth order, birth interval, open defecation were not included in the national nutrition survey.

It is true that there was no change in the national level for minimum acceptable diet among children 6-23 months between 2010 and 2014. However preliminary analysis of the relation between stunting prevalence per region and minimum acceptable diet per region has shown that shown that region regions with highest level of minimum acceptable diet have the lowest level of stunting. The same is also seen for minimum meal frequency and minimum dietary diversity. But these assumptions need to be confirm with a rigorous statistical analysis. I can share with you the draft ToR for the in depth analysis of the national nutrition survey. We welcome any support or ideas to carry out this in depth analysis.

Best regards


pauline kisanga said...

Despite very high level commitment to scaling up nutrition in Tanzania, there were not many nutrition specific interventions between 2010 and 2014 except for Vitamin A supplementation, the flagship Mwanzo Bora nutrition project in Manyara, Dodoma and Morogoro and the UNICEF program in iringa, Mbeya and Njombe; and some other smaller area specific projects. What was significant and widespread during this period however, was the Government supported TASAF's direct cash transfers to poor households.Now that you mention the poverty reduction angle I feel this could have played a key role. I read somewhere in the past that cash transfers to poor households increased purchasing power for everyone at the local level regardless of how the funds are spent at that level.The challenge in many African countries is that despite economic growth, there is never a change in the income of the poor because the extra growth in terms of income never trickles down. Therefore there is minimum investment to generate income for other livelihoods. When poor households have nothing to spend on other necessary livelihoods, the funds meant for food/or even food is sold for cash to purchase the other livelihoods-school fees, hospital, uniform, etc. But if there is some money to meet these important other needs, the food grown at home is eaten by the family not sold to get cash. So this could be the reason.
Pauline Kisanga, Nutritionist, COUNSENUTH, Tanzania