1 The importance of good data for policy analysis: an example from Tanzania Cheryl Christensen Economic Research Service (ERS) United States Department.

1

The importance of good data for policy analysis: an example from Tanzania

Cheryl ChristensenEconomic Research Service (ERS)

United States Department of Agriculture (USDA)May 10, 2013

2

The Global Strategy from the perspective of data users

Input from data users is critical to the assessment process • Improved statistical systems are not an end in themselves—

they must be used in order to have impact• The needs and priorities of users should be built into the

process of creating improved statistical systems • Analysis and strong analytic capacity are key to creating value

from better data The costs of poor data

• Bad data lead to ineffective or even harmful policies• Lack of accurate and timely data reduce the efficiency of

markets and raise transactions costsRecognizing the cost of poor data can strengthen the demand for better data

3

An example from Tanzania

USDA engagement with Tanzania • ERS and the National Agricultural Statistics Service (NASS) conducted an initial

USDA assessment July 2011 to evaluate the agricultural statistical system • ERS extended the assessment to the food security information system ( )• Developed strong tie to policy analysis—ERS participated in a USAID SERA Project

analysis of Tanzania’s export ban

Commitment from the Government of Tanzania• This is a country driven model• Good links to data users both among high level officials and working level offices

• Prime Minister’s Office• Department of Food Security at the Ministry of Agriculture, Food Security and

Cooperatives (MAFC)• Recognition by the Prime Minister’s Office that there was a need to change the way

that food security is measured

Coordination among stakeholders and donors • USAID – USDA – USAID consultants – FAO – World Bank

4

Linking better data to policy analysis: Tanzania Food Crops Export Ban Export bans • Since the 1990s Tanzania has periodically used export bans to address food

security concerns• Bans have strong regional impacts in Tanzania as well as eastern Africa• They are incompatible with export-led growth model of Southern Agricultural

Growth Corridor (SAGCOT)

Link to data • Production data collected from extension agents not statistically reliable• ERS evaluation found that the method used to compute food security

requirements overstated the need for maize• The food security requirement calculation was used to determine if national

supplies were adequate• When national maize supplies were inadequate export bans were imposed

5

Key steps in changing the export ban

• Concept Note sent to the Government of Tanzania in October, 2011• Government approved work plan in November• Research prepared by ERS, international and local consultants

during March-June, 2012• President committed to removing export ban in May as part of G8

Implementation Framework• Workshops for Government and all stakeholders in June• Policy Brief prepared in August• PM Announced an end to export bans on Sept. 6, 2012 and an

openness to developing more accurate measures of food security.

6

7

Next steps: Short term strategies

Develop a measure of food security that more accurately reflects Tanzania’s diverse consumption—existing information could be used to estimate a food basket

• Measures changes in cost of acquiring representative food basket• Measures access rather than availability• Data needs

• Calorie shares of foods important people’s diets• Retail prices • Per capita income

• ERS has conducted pilot food basket estimates in two districts (Mara and Mbeya) working with USAID and the Department of Food Security at the Ministry of Agriculture, Food Security and Cooperatives (MAFC)

8

Next steps: Short term strategies Construct supply and use balances for products important to agricultural markets and food security

• At the beginning use existing information supplemented by information gleaned from NGO surveys and interviews with the private sector

• Update and improve balances as new and better data are available from statistics offices

Improve the collection, dissemination and use of key market prices

• Wholesale prices are disseminated for selected crops, but not key food security crops such as cassava

• Both wholesale and retail prices are collected for a much wider range of crops at the district level, but are not systematically organized and disseminated in a timely way

9

Conclusions

• The quality (good or bad) and use (appropriate or inappropriate) of data can have significant implications for policy

• Better data can create opportunities for re-evaluating existing policies, as well as laying a foundation for better future policies

• Data linked to analysis and research can support policy change• Government engagement and support is critical• Short term improvements in data and analytic methods can lead

to improved outcomes even as work to establish a better overall agricultural statistical system is ongoing