> Promoting and protecting open civic space which enables citizens to freely assemble or organise, speak and act.


A selection of Twaweza’s work bringing governments and citizens together.


Twaweza partnered with Compass Communications and Sahara Media Group to produce and broadcast a live interview show infused with data. Questions collected from citizens via SMS and social media were asked in studio to high-level government officials including four ministers.

Officials who participated: January Makamba (Minister – Vice President’s Office, Union and Environment Affairs), Harrison Mwakyembe (Minister for Information, Culture, Arts and Sports), Ummy Mwalimu (Minister for Health Community Development, Gender, Seniors and Children), Selemani Jaffo (Minister, President’s Office – Regional Administration and Local Government), Kitila Mkumbo (Permanent Secretary, Ministry of Water and Irrigation), ACP Barnabas Mwakalukwa (Spokesperson, Police Force of Tanzania), Ludovick Utuoh (former Controller and Auditor General, National Audit Office of Tanzania).

Monitoring data show that over 35% of people watched the show, 95% thought the show was a good way to bring citizens’ concerns to ministers, 58% said it was the first time they had seen a minister answer questions from citizens and that viewers were much more likely to believe that citizens know enough to be able to question officials (79% vs. 64% of non-viewers).


  • As with previous Twaweza initiatives, live TV in particular seems to be impactful and more accessible than TV viewership figures suggest.
  • Securing Ministers who were willing to participate could be quite challenging resulting in some postponements of the show.
  • Citizens feel that they have limited opportunities to engage with ministers directly and so initiatives that build more space for interaction and exchange should be emphasized.


Twaweza is partnering with Maa Media to produce #MbungeLive, a profile of Members of Parliament – their election promises and implementation, how much they interact with constituents, and how well they represent constituent and national interests in Parliament.

In the first stage, we produced two pilot episodes featuring one MP from each of the two major parties in the country: Hussein Bashe (MP, Nzega Urban), John Heche (MP, Tarime Rural).

From the research and pilot screenings, we find that the show appeared to increase belief in MP’s willingness and qualification to fulfil campaign promises, increase knowledge of those promises, and increase belief in MP’s interest in listening to citizen concerns. There is no evidence that respondents thought the MP sent us to do the research.

On the MP side, we have anecdotally seen an increase in constituency level implementation of commitments for the show and accountability actions from the screenings whereby issues raised there were addressed by the MP soon after the screenings.

Given the success of the pilot, we are now seeking to pursue a scale up – to produce a full season of the show for airing on national TV.


The filming of the show acts as an intervention. During the pilot, we already saw MP Bashe in particular stalling recording in his constituency to enable him to finish projects. Indeed, the selected MPs are already making plans to do the same. And when Hussein Bashe was lobbying other MPs to participate he was very clear that they had to make sure they had done work in their constituency.

The MPs (based on the interviews conducted in Dodoma) are feeling quite downtrodden, unheard, unrecognised. This makes them more enthusiastic to participate than we had anticipated. We had a queue of about 30 MPs who all wanted to be part of the show; one MP refused the leave the location until he had been interviewed; some of those selected got on their knees in gratitude at having been selected. MPs see the show as a good way to communicate to their constituents and showcase their achievements.

Twaweza will continue to engage with this group of MPs over the longer term. We will in particular emphasise interaction with citizens as an important means to ground their national-level interventions in Parliament. This is the most important function MPs play – leveraging the local to the national (and national back to local), and so we will emphasise interaction with constituents as the most important way to play a meaningful and successful national role. This will motivate the MPs as it is in essence advising them on how to make the best use of the platform they have nationally, and it will help nudge the system as currently MPs are viewed as sources of support for individual challenges rather than conduits to amplifying citizens’ voices.

Research Brief: #MbungeLive

Full research report: Evaluation of #MbungeLive


Following a qualitative study and nationally representative findings showing that 70% of citizens were keen for MP candidates and Presidential candidates to debate each other, we ran four manifestoes focused debates with senior representatives from five major political parties in the election. We ran one debate between candidates vying for the ruling party nomination. And we ran candidate debates for the posts of President of Zanzibar and President of the Union. All debates were broadcast live on TV and radio, and online. Citizens could submit questions online or via SMS. Each debate began with a short data animation highlighting key challenges and successes in the relevant sectors. These data were also shared in advance of the debates via social media.

The Mkikimkiki debates were watched and listened to by five million people cumulatively. Monitoring data show that the debates reached 40% of citizens in Tanzania, generated in excess of five million Twitter impressions and were watched online by over three hundred thousand people. In total, we received over 250 questions from citizens. The debates generated over 30 pieces of media coverage and were credited with having contributed the only issue-focused discourse to the campaign period and the only opportunity for citizens to have their questions answered by political party representatives and candidates.

We also worked with Well Told Story to produce radio and comic content encouraging young people to consider issues when voting, generating significant response on social media. In particular, young people appeared better able to articulate their political priorities following the campaign. And we produced animated comic clips with caricatures of political stereotypes with Vuvuzela Media which generated close to two million impressions on social media.

And finally, we designed an extensive in-depth monitoring plan to look at the reach, quality and effects of this work. The monitoring included a household survey, focus groups and questionnaires at the debates, the conjoint experiment in four districts in the country and investigation of reach through representative surveys, web analytics and media consumption data.

Read more about the evaluation and learning components


We confirmed some of our hypotheses about the priority areas for citizens during elections through an early year scoping study. The scoping study found that citizens were skeptical of politicians’ promises during the campaign period and that they were hungry for interaction, the opportunity to question candidates. Thus we dropped a planned ‘ten commitments’ intervention and focused on running debates.

We were unable, despite continuous pressure, to secure the participation of the two main presidential candidates in the final debate.  Ultimately an intervention cannot control the willingness of government officials and politicians to participate. Nonetheless, we continue to advocate for increased accessibility of senior government officials to citizens.

Despite intuition to the contrary, most viewers reported watching the debates on TV not listening to them on the radio. The live element including the audience and received some questions in real-time contributed to the power of the debates.

When we were able to share the data sufficiently in advance, the quality of the questions increased; they became more specific and based on particular challenges in the sector rather than general questions.


Research conducted by MIT for Twaweza shows that young people generally seem informed and interested in politics, however, politicians largely do not engage with issues important to young people. In response, Twaweza implemented the #WhatWouldYouthDo youth-focused national multimedia campaign, in partnership with the African Youth Development Link, in the run-up to the 2016 general elections in Uganda. The purpose was to awaken young people to their political power by encouraging them to base their decisions on issues of concern rather than party loyalty, financial inducement or personalities. The campaign sought to provide opportunities for young people to interact directly with local political candidates, alongside a creative multimedia campaign that could draw potentially disengaged youth into political dialogue. We worked with partner Buzz Events to ensure youth-relevant messaging, communications and events, and with African Youth Development Link to promote their Youth Manifesto, which was created through a consultative process involving thousands of young people across the country. The intervention also aimed to pressure political leaders and candidates to engage with and make reference to youth issues during the campaign.

Activities included:

  • Five political party debates broadcast live on TV and radio featuring the three major parties contesting the elections on the major topics of the Youth Manifesto
  • 1300 public service announcements (PSAs) featuring musicians promoting the messages of the youth manifesto and 700 DJ Mentions
  • 40 talk shows and interviews
  • 90 local candidate debates held at local councillor levels
  • Distribution of 10,000 DVDs containing a short film about the Youth Manifesto and PSAs featuring artists to communal watching spaces such as video bandas, salons and buses.

The monitoring components were built into the design. Alongside a host of reach and quality monitoring for individual components, researchers from MIT GOV/LAB conducted a scoping study that informed the design and assessed the effectiveness of the debates. The campaign achieved some significant results

36% of Ugandans were aware of the Youth Manifesto following the campaign

The debates were watched by five million people cumulatively and 91% of them were under 35

The political parties were reportedly following the debates closely and using them to craft youth-orientated messages

The debates directly reached over 1,500 young people affording them the opportunity to ask questions of the political parties seeking to represent them

Three questions arising from the youth debates were asked at the main presidential debate


  • There were multiple partners to coordinate in this intervention and that sometimes created synchronization and delivery challenges.
  • An experiment to reach youth with low literacy levels using automated phone services whereby citizens could listen to pre-recorded information about the different party manifestos and submit questions for the debates by recording them was not successful. Very few calls were received.


In the first Twaweza strategy, emphasis was placed on religious organisations as an important communications channel for citizen agency and general development messages. However, we struggled to make real in-roads into this space. But there is no doubt that religion has huge influence on people’s values and attitudes, even their actions. For example, in 2014. Sauti za Wananchi found that people trust religious leaders more than even their own relatives.

In 2014, Twaweza successfully entered into a partnership with the Christian Social Services Commission essentially tasked with managing service delivery for the Church (all denominations). So they oversee the work of church owned schools and health centres. The aim of our work with them was to distribute discussion guides to jumuiya groups (bible study / prayer groups) to inspire them to discuss and maybe even take action on issues in the health and education sectors. This distribution was complemented by radio content being aired on Christian radio stations in the areas (approximately) where the guides were distributed. The partnership was not completed until 2017 and faced some challenges:

  • Scale was large and so made it difficult to manage each segment well – we targeted 15,000 jumuiyas all across the country and 5 radio stations airing content three times per day, five days per week. It was impossible to quality assure all of this work.
  • Lack of focus – because the partnership took so long to complete and actually straddled strategies and managers, the objectives overall were not clear or changed a number of times throughout the project.
  • Insufficient links (timing and content) between the radio and discussion guide components.

Despite all of this, some of the monitoring findings are positive:

  • Six months after distribution, 44% of the Jumuiya leaders recalled having received the publications and 64% of those were able to recall the content of the publications.
  • Two out of three of those who recalled the publications said they were engaging and that they were used in the Jumuiya discussion groups.
  • One out of three Jumuiya leaders (33%) report having taken action based on the recommendations in the guides.
  • 36% of respondents said they shared the guides with friends, 34% took it home for reading
  • In addition, there was a lot of rich information contained in the individual reports from the different jumuiyas indicating that they discussed the guides and that these guides provided them with new ideas about their own roles in these two service sectors.


  • Trial the concept in a more limited geographical area so Twaweza can follow up better.
  • Define clearer objectives and calls to action
  • The sensitization of jumuiya leaders who really have to guide these discussions and ensure they move in the right direction is critical to the success of the intervention. Although this was done in the first phase of the partnership it was done by CSSC as a bit of an add on with no input from Twaweza, moving forward it will be more integral.
  • Ensure clearer links between radio guides for talk shows and guides distributed to jumuiyas including in terms of timing.


Increasingly in Tanzania, online platforms are gaining control of viewership, readership and listenership and pushing the boundaries of free expression more than traditional media. So Twaweza has invested in partnering with independent online media outlets to disseminate data, voices and insights and to support their development. With recent figures showing that there are as many as 23 million internet users in Tanzania, social media and online communication are becoming essential channels of communication.

We have been partnering with Jamii Forums since 2013. This partnership was originally purely instrumental whereby Jamii Forums presented an alternative dissemination channel for our data and positions.

Jamii Forums is Tanzania’s largest and most engaged community of online users. Part social media, part website Jamii Media provide the following (impressive) user data:

  • It has over 457,000+ registered users,
  • More than 300,000+ daily visits.
  • It has an 8min Average time on site per visitor.
  • It has an average of 12 pages/visit per day.
  • It has over 50 million page views each month

In recent years, our work with Jamii Forums has also taken on a different dimension as we have supported them to navigate a complex set of litigations against the directors and founders for refusing to reveal the names of whistle-blowers who released information on the site.

Twaweza has also recently entered into partnership agreements with Kwanza TV, an independent, investigative online TV station with 67,932+ followers, more than 775,000+ monthly and 27,000+ daily impressions on Twitter, 104,000 average views per video, and 10-minute average time on the site. The partnership will allow us to live stream events and disseminate our ideas and evidence.

Twaweza is also experimenting with working with individual social media influencers.

In reaching new audiences with our data and generating lively debate, we can claim some success: Jamii Forums has helped push Twaweza data out and has prompted extensive public discussion and debate. We can point to a huge extension of reach for Twaweza and greater engagement on the Jamii Forums platform compared to our own social media channels. Our engagement with them has led to millions of social media impressions and engagements. It is also worth noting that, during times when Twaweza has been the subject of hostile communications, some of our strident defenders have emerged on Jamii Forums. And finally, we have seen most strongly on Jamii Forums, the emergence of a popular discourse that can be summarized as ‘oppose research with research.’ Jamii Forums themselves have also benefitted from developing expertise as experts in communication for engagement of data online, and from our support during their recent difficulties.


  • For quick feedback and deeper engagement, social media is unsurpassed. But it is critical to use the right platform for the right messages and content.
  • There is a need for Twaweza to move beyond instrumental uses of online communication and incorporate this work more directly into the pursuit of accountability.
  • Individual influencers have allowed us to reach new audiences with more engagement, perhaps because the communication feels more personalized not institutional.


Ni Sisi is an idea that is at the core of Twaweza’s vision for development and our Theory of Change. It’s all of us who can make a difference in our own lives and societies.

In 2012, Twaweza decided to trial a different model of communications interventions – a large scale intense campaign that propagated our core message, Ni Sisi, across a range of traditional media platforms as well as some newer more innovative mediums.

Working with the market leader, J. Walter Thompson (JWT), we created a range of content including compelling TV and radio public service announcements and light-hearted radio pieces and set to work to paint Kenya and Tanzania orange. The campaign was implemented at a large scale across Kenya and Tanzania, and to a limited extent in Uganda.

In Kenya, Uganda and Tanzania, six powerful radio and TV adverts were aired thousands of times every day over three months on multiple media outlets. We also made use of creative outdoor advertising on walls, benches and billboards. In Tanzania and Uganda, many of our programmatic partnerships also focused on conveying the core Ni Sisi message.

No message was received by even 50 per cent of viewers (when they were asked to identify the message in an open-ended question) in Kenya and Tanzania.

High increase in people feeling that they themselves were responsible for tackling issues during post-campaign surveys. However, this occurred in exposed and unexposed groups. This might be due to other similar initiatives or messages.

However, in Tanzania, three issues, all covered by the campaign, saw a significant increase in respondents identifying themselves as primary change agents when compared to pre-campaign respondents and those unexposed to the campaign.

In Tanzania, 43% of citizens could spontaneously complete the tagline for the campaign.

Anecdotal feedback shows that the campaign was extremely popular and lead to a small spark of actions modelled on the adverts. For many years after the campaign, candidates coming to interview for positions at Twaweza would mention the campaign in connection with Twaweza’s work.


  • It is possible that the main ‘ni sisi’ message is received more subliminally so when asked you would identify the issue covered (pollution, corruption etc) as the main message of the adverts.
  • TV was the most powerful content while radio had a low impact, reach and recall. We speculate that people tend to engage with TV communally even more than other media.
  • Although we achieved significant reach and recognition, this came for a high send on traditional media buying. Value for money is hard to assess. But working with an agency expedited things that we might not have been able to process with as much speed and agility.


Although we achieved significant reach and recognition, this came for a high send on traditional media buying. Value for money is hard to assess. But working with an agency expedited things that we might not have been able to process with as much speed and agility.