Suggested Citation. MIT GOV/LAB Research Brief. 2018. “Testing Access to Information in Tanzania with Mystery Shoppers.” Cambridge, MA: Massachusetts Institute of Technology Governance Lab.

Results have been internally replicated, but may undergo further revisions. MIT GOV/LAB reserves all rights over data, methods, and results for publication.


Accessing public information, like government plans, budgets, and activities, is an important mechanism through which citizens can hold government accountable for providing basic services. From accessing water and medical care to starting a business or enrolling in school, citizens need open and public information to make day-to-day decisions about their lives.

Access to information is enshrined in the Tanzanian Constitution and in the 2016 Access to Information Act. Yet even with laws and principles on paper, an active culture of freedom of information has been slow to develop. Two years after passing the Act, the government released regulations and must now work to implement the rules across different offices and districts. In 2017, Twaweza data showed that 86% of Tanzanian citizens believed that information held by public authorities is a public resource and 13% had contacted a local government office to ask for information, yet only 5% had heard of the Access to Information Act in 2017.

In this context, MIT GOV/LAB and Twaweza partnered on a “mystery shopper” experiment in early 2016, to assess how citizens’ information requests were received and processed at the local level before the law was passed (the law was tabled in mid-2015 but its passage was widely expected in 2016). The goal of this research was to create a baseline assessment of government transparency, which can enhance implementation of the law and help measure progress in the future.

Key Takeaways

What percentage of information requests were fulfilled?

The mystery shopper experiment indicates that in 2016 there was significant variation in how government offices responded to information requests overall. Moreover, the information requested was released in just one out of three instances.

    • Tanzanian researchers made information requests at 131 offices in 26 randomly sampled districts (of 169 total districts in the country) to get a nationally-representative baseline assessment of current government practices.
    • Of those requests, 33% were either fully or partially fulfilled, while 67% were denied.
    • Requests were most frequently fulfilled on paper (44%), as opposed to verbally or online.
    • When requests were fulfilled, 67% were provided in Kiswahili.
    • Government officials responded to 37% of requests on the same day, though 13% of requests took 4 days to receive a reply. Fulfilled requests were processed nearly one day earlier than denied requests.
    • The Department of Public Works fulfilled the most requests (10 out of 21) while the District Executive Directors’ (DED) Offices fulfilled the fewest (2 out of 25).
How did information fulfillment rate vary by office?
*DED = District Executive Director

Featured research project: Tsai, Lily L. and Alisa Zomer. “Evaluating public information provision and government transparency using a mystery shopper methodology.” Work in progress.

Partner: This research collaboration was undertaken with Twaweza, a civil society organization that works on enabling children to learn, citizens to exercise agency and governments to be more open and responsive in Tanzania, Kenya and Uganda (

For more on the research partnership and process: “MIT GOV/LAB Learning Case. 2018. “Navigating Access to Information with Twaweza and MIT GOV/LAB.” Cambridge, MA: Massachusetts Institute of Technology Governance Lab.