As far as we are aware from our desk review, UNDP Pakistan is the first organisation to use PD to understand how to develop a women’s empowerment initiative – additionally, the tribal areas are the most difficult region in Pakistan for working on women’s empowerment. Therefore, PD is novel as an approach to women’s empowerment, to working in the tribal areas, and for the country office.

Traditional project design methods are top-down and prescriptive instead of disruptive. Traditional survey methodologies were simply not providing us with the necessary information on the lives of women in FATA. PD is less time consuming; cost-effective; offers ethnographic insights on sensitive issues; is locally led and participatory and therefore has the ability to foster more ‘trusted’ and sustainable solutions.

The final report of the positive deviance pilot is being edited to prepare a more concise, strategic, and focused document that looks at lessons learned and potential next steps. An infographic has been prepared as well and UNDP Pakistan’s innovation portfolio is now meeting with various projects and programmes to share strategic findings from the pilot and consider ways in which these could be scaled-up as more comprehensive project interventions. Further, innovations is also looking to use PD as complimentary approach to designing strategies for remote, conservative and extremely marginalized groups.

Further, the findings have been shared with partner UN Agencies working on issues of women’s empowerment to consider joint initiatives based on the pilot findings. The ambition is to iterate a small number of the key strategies identified as being used by the positive outliers to determine whether they can be successfully replicated in similar contexts.

Pakistan’s tribal areas suffer from being both remote and unstable – as such, it is a difficult area for UNDP to access. This resulted in the need for the pilot project to be implemented on the ground through a local partner – in this case an NGO that had strong experience working on the ground in the tribal areas and had well-established relationships with communities there.

Citizens: in this case, women and men in the community are the direct beneficiaries. In our project we could not test out the strategies but we foresee that the PD approach would lead to behaviour change within the communities.

Development practitioners: The findings offer insights on new way of gathering data and new insight on women empowerment programmes for development practitioners.

The key result of the positive deviance pilot was to generate behavioural insights and deeper information sets on psychometrics which provided us with a strong evidence-base. We now hope to move to further rounds of iteration and later scaling-up of tested and proven strategies which would allow for more comprehensive measurement of results.

From the approaches implemented by positive outliers, the key strategies we would wish to implement and further test are:

1. Establishing a network of employed women who can serve as role models
2. Demonstrating women are able to work and maintain honour
3. Creating channels of connection with employment spaces friendly to women
4. Using influential male members of the family to advocate on women’s behalf
5. Facilitating women’s employment in public service roles that earn community respect

Unable to access the tribal areas, UNDP needed to work through an established local implementing partner. While the partner gave us access, positive deviance was an approach not previously used by the UNDP Pakistan office or the partner. Maintaining the focus on testing this new approach, and ensuring the partner did not implement a traditional top-down community sensitization scheme was challenging.

Despite constant engagement, the final round of our pilot which would have been iteration (i.e. applying the strategies identified as being used by positive outliers in similar communities) was not fully completed. The partner focused on facilitating sessions with similar communities where the strategies were shared, but a rigorous iteration phase was not implemented. As such, this remains a next step in the process that we still wish to implement, in order to be able to more concretely determine results and scalability.

UNDP provided a safe space to fail – in traditional development projects, there is not often the space to fail and in order to encourage testing, learning, adaptation and growth, this is critical.

Constant communication and reflection on the innovative initiative as it is implemented is also crucial to ensure learning and course correction as necessary. We need to constantly reflect on what we are trying to achieve, and to ensure that all colleagues and any partners remain informed and on board.

Human resources seemed to be more important than financial resources – testing an innovative approach required time and focus to ensure we did not slip back into “old habits” of doing what we already knew how to do.

Last but not least – motivation, curiosity and a willingness to learn by those involved in the innovative approach. You need the right energy in your team.

To date, the innovation has not yet been replicated but the office has every intention to do so. As a starting point, we would like to complete rapid prototyping by testing the identified positive outlier strategies in the relevant UNDP projects (likely in early 2019). Thereafter, we will have a more concrete results base from which to scale-up to more comprehensive development interventions.

The PD approach has received significant interest in the office as a way of gathering insights on hidden behaviours that we are often not able to see. Many of the projects implemented by UNDP Pakistan focus on behaviour change – be it in relation to electoral processes, addressing environmental issues, or regarding individual income generating behaviours. As such, our experience to date is being shared with other projects with a view to replicating.

Atypical successful individuals (or groups) can give us insights into what could be possible for typical groups with additional guidance – therefore, positive deviance is an approach for identifying and observing these outliers. It is a methodology that values collective intelligence; the belief that intelligence is not dispersed from the leadership down but is instead distributed throughout the entire community. This concept is something that is often overlooked by development practitioners – getting a broader insight into the behaviours and psychometric characteristics of those people you are looking to engage with is extremely important. It will help us to improve our understanding of currently underserved groups, finding nuances and differences between people that may not be clear from their wider contexts alone.

To uncover hidden behaviours, it is necessary to build trust and do “respectful digging”. As such, if you are implementing PD directly or through a partner taking the time to build trust with those you are studying is critical.

There are various stages to completing a successful positive deviance initiative and it is important to dedicate adequate time and resources to each. The initial work of identifying positive outliers and studying their behaviours can be done relatively quickly. However, for us it was the iteration round that proved challenging; this requires additional resources, monitoring and a significant amount of time to see whether the behaviours can be adopted and applied by other individuals or communities. We were unable to complete it during our initial round but we through sharing our findings with other projects, we plan to test some of these strategies in iteration rounds over the coming year.

Finally, from experience we would recommend that the PD process is included at the project design stages as a compliment to baselines/part of agile development methodologies and for sensitive issues and hard to reach groups.