Imagine the frustration. You arrive in a community with a questionnaire, a tablet, or even just a notebook. You introduce yourself, explain a programme intended to improve the lives of the people in the community and ask your first question. And then, nothing. Or worse, the carefully rehearsed non-answer. A polite but firm reluctance to engage. The clipboard, it turns out, has become a threat.

Lately, the stories I hear from field workers and programme implementers are telling a consistent and troubling tale: poor communities are becoming more guarded with their personal information. They do not want to share. They do not trust easily. And this is making it increasingly difficult to determine whether the programmes we design and implement are making a meaningful difference.

This is not a local quirk or a temporary problem. It is a deepening structural challenge, and understanding why it is happening is the first step toward finding an effective path forward.

The Fear Is Rational

The instinct in M&E circles is sometimes to frame community reluctance as a methodological nuisance, a data quality problem to be solved through better interviewer training or more compelling introductory scripts. But this framing misses something fundamental: in the Southern African context, the fear of sharing information is entirely rational.

Take South Africa’s Social Relief of Distress (SRD) grant. Research by the Institute for Economic Justice has found that intensified fraud crackdowns, including bank account monitoring, biometric checks, and algorithmic risk detection, have resulted in a shocking error rate. Howson et al. (2025) found that a very high proportion (71.7%) of grant exclusions for people living below the upper bound poverty line are erroneous. These exclusions are often due to incorrect bank verification checks. When a household has experienced, or knows someone who has experienced, losing a grant because of a transaction that looked suspicious to an automated system, their reluctance to share financial or household information with an outsider carrying a tablet is not irrational. It is wisdom.

This dynamic is reinforced by what Murray et al (2023) call the ‘chilling effect.’ Research in Zimbabwe and Uganda has found that the mere belief that information might be monitored or shared modifies people’s behaviour; not because they are hiding wrongdoing, but because they cannot predict the consequences of disclosure. In the absence of trust, silence becomes the safest strategy.

Then there is the broader pattern of what happens when recipients are repeatedly subjected to data collection without any visible benefit flowing back to them. Studies of social support users have found that when people experience data systems as unjust, they don’t just disengage from one interaction. They lose trust in the entire process (Theiss et al, 2025). That lost trust then extends to anyone perceived to be connected to that process, including NGO fieldworkers and community practitioners.

The M&E System Is Part of the Problem

It is uncomfortable to admit, but the monitoring and evaluation apparatus itself has contributed to the crisis of trust it now faces. Audit culture in development has produced a regime in which quantitative indicators, often very personal indicators, have become the primary language of accountability. Programmes must show numbers: beneficiaries reached, incomes changed, assets acquired.

As a result of audit culture, communities are providing data that seems to take from communities without giving back. We ask people to make themselves legible to funders and governments without any corresponding accountability running in the other direction. Communities that have been surveyed repeatedly across multiple programmes, often by different teams asking overlapping questions, develop what Cherop et al. (2026) have called ‘kuchoka’: research fatigue. Their 2025 study found that research fatigue prevalence was 56.3% in a repeatedly-surveyed community, with being asked personal questions identified as one of the strongest independent risk factors.

Stuck in the Middle: The M&E Practitioner’s Dilemma

So here is where I find myself. Funders require evidence. Quantitative data remains the dominant currency of accountability in development. Household surveys, standardised questionnaires, and logframe indicators are not going away. And yet the communities whose lives these tools are meant to illuminate are, with good reason, pulling back from them.

The pressure to produce numbers has not diminished. But the foundations on which those numbers rest, trust, willing participation and honest disclosure, are cracking.

What, practically, can be done?

A Different Direction: Toward Inclusive M&E

The literature does not offer a single silver-bullet solution, but it does point toward a set of principles worth taking seriously — especially for programmes operating on constrained budgets.

We need to be honest about what surveys can and cannot tell us. Quantitative data collected from communities with low trust is not neutral data. It is systematically biased toward the preferences and answers that feel safest to give. Triangulating survey data with qualitative methods, community observation, and administrative data gives a clearer picture of programme impact. The question ‘did this programme change lives?’ sometimes has to be answered through stories before it can be answered through statistics. At FinMark Trust, we are actively building a feedback loop of qualitative stories of impact coming from our programmes. Some of these stories are already available on our website.
We need to close the feedback loop. Smith Ramey et al. (2025) identify feedback mechanisms as one of the most cost-effective ways to sustain community participation in data collection over time. Willingness to participate in data collection increases when communities receive something back. This does not have to be elaborate: community practitioners can share feedback through a simple community meeting or even distribute a visual summary of what the data showed and what the programme intends to do differently. At FinMark Trust, we have taken steps to tangibly demonstrate our impact to our funders and partners. And, on some programmes, like our Community Digitalisation project, we are sharing stories of impact with the community through digital means, such as Facebook, but we do have an opportunity to extend this further.
Separating the support relationship from the data relationship is best practice. When in-community data collection is essential to programme evaluation, the trusted community practitioner should actively bridge the introduction to the data collection field worker— accompanying the beneficiary where possible, explicitly endorsing the new field worker, and explaining what information will and will not be shared. Often, we look to our community practitioner to also collect data. This is cost-efficient, but it can damage the trust relationship if the community support practitioner is also seen as a mechanism of surveillance. At FinMark Trust, we are actively looking for ways to reduce the burden of data collection for community practitioners. Our Community Digitalisation project is exploring WhatsApp surveys as a possible solution. Our Generating Better Livelihoods for Grant Recipients project is rolling out a management information system aimed at supporting our community practitioners in generating meaningful linkages for their project participants, but also ensuring that data is available to the project and monitoring and evaluation teams for their purposes.

A Closing Thought

The communities that development programmes are designed to serve have learned, often through hard experience, to be careful about who they let in and what they share. That carefulness is not an obstacle to good M&E. It is important information for our programmes to receive. It tells us that the relationship between development institutions and the people they claim to serve has been damaged, and that rebuilding it requires more than a better questionnaire design.

The clipboard will only become less of a threat when the community holding the door open can reasonably expect that what passes through it will not be used against them. Getting there is not only an ethical imperative. It is the only way we will ever reliably know if the work we are doing matters.

References

Cherop, F., Naanyu, V., Wachira, J. and Atwoli, L. (2026). “Kuchoka”: Investigation of Research Fatigue in Mosoriot, Kenya.

Dludla, S. (2026) Fraud crackdown on social grants risks excluding the vulnerable, research warns downloaded from IOL here: https://iol.co.za/

Howson, K., Baduza, S., Setambule, T. and Khumalo, T (2025) Systemic exclusion from a South African social assistance transfer: Drivers, impacts and who is most at risk. Institute for Economic Justice and Agence Française de Développment (AFD)

Murray, D., Fussey, P., Hove, K., Wakabi, W., Kimumwe, P., Saki, O. and Stevens, A. (2023) The Chilling Effects of Surveillance and Human Rights: Insights from Qualitative Research in Uganda and Zimbabwe in Journal of Human Rights Practice, Vol. 16, Issue 1, February 2024, pp 347-412.

Shore, C. & Wright, S. (2024). Audit Culture: How indicators and Rankings are Reshaping the World. Pluto Press.

Smith Ramey, J., Volk, F. and Millacci, F. (2025) Where Every ‘One’ Knows Your Name: the Transformative Power of Qualitative Approaches in Community-based Research in New Trends Qualitative Research, Vol. 21, issue 1.

Theiss, M. and Szelewa, D (2025). “Institutional Sources of Citizens’ Trust in the Welfare State: Literature Review.” in Social Policy and Society, Vol. 24, Issue 4, pp 637-655, Cambridge University Press.