Five things we learned about engaging citizens on strategy delivery and Artificial Intelligence

Written by Dr Paola Pierri, Director of Research and Design, Democratic Society 

In 2022 Democratic Society carried out a rapid literature review and an international case study analysis to learn from best practices of community participation and engagement mechanisms for governments’ strategy delivery. This work was commissioned by the Scottish AI Alliance to provide initial insights and recommendations for ensuring full participation in the delivery of Scotland's Artificial Intelligence (AI) Strategy.  

The desk-based research collected 47 relevant case studies that could provide learning on what worked well and what did not work well in community participatory / engagement mechanisms (with a particular focus on strategy delivery and Artificial Intelligence). In terms of geographical spread the case studies feature a mix of UK and international examples.  

We wanted to learn about how to make engagement in strategy delivery meaningful and impactful for people and what techniques and methods were more successful to ensure that citizens and groups can sustain engagement for a long period of time (since strategy delivery can have a medium-to-long term time span).  

A report summarising all the main findings and recommendations can be read here.

Here we want to share the five main things we learned. 

First - nobody is talking about strategy delivery 

From our search emerged a scarcity of case studies that were specific and detailed about ‘strategy delivery’. This could possibly be due to the fact that participatory practice and engagement usually focus on the strategy and policy development phases. Ensuring full public participation during a strategy / policy delivery phase could be in fact perceived as a more difficult and challenging exercise, since delivery usually takes more time and requires sustained long-term engagement. Moreover, it might be easy to assume that engaging the public in the development phase might be enough, and to leave the delivery phase to experts instead, assuming communities and citizens might have less to contribute. Ensuring good documentation practices and the development of a learning culture throughout the delivery phase was one of the key recommendations for the Scottish Government’s delivery of its AI strategy. 

Second - One solution does not fit all  

There is no ‘one approach’ to the design and delivery of inclusive and effective community participatory and/or engagement mechanisms for government strategy delivery, as no one method or combination of methods can reflect the underlying and evolving interests and intentions of citizens on the matter at stake. A good communication plan and different mechanisms to integrate and act upon feedback loops are needed to check and adjust the participation strategy accordingly. 

 An effective participation and engagement strategy should include both, the continuous improvement of existing engagement mechanisms and platforms, as well as the seeking of new ones that are bespoke for the scope of the AI strategy delivery. A good mix of existing and new engagement mechanisms are expected to provide desired outcomes and mutual benefits to participants and institutions leading engagement activities. 

 Both top-down and bottom-up approaches should be included in the design of an inclusive and effective participatory strategy, as the first can provide citizens’ appraisal of predetermined policies that are put forward by governments; whilst the second can deliver a space for policy development and shaping of new tools and forms of governance; collective intelligence plays an active role in both cases. 

Third - Start with co-creating the design principles  

Identifying in a collective manner the key design principles that will inform the design of the participatory approach is an important step to ensure that values and ambition are set right from the start and that possible conflicts and divergence in views are also mapped and addressed early on in the process. In the report, we suggest some practical design principles that could inspire the work of those who are planning to use participatory methods for strategy  delivery.  

Fourth - AI needs its own methods 

The need for public institutions (and democratic infrastructures) that are resilient and responsive to the emerging challenges that society faces is a pressing one, and certain fields have been proving so far more challenging to democratic and participatory mechanisms than others. Digitalisation and the field of Artificial Intelligence are definitely among the areas where the need to rethink participatory and engagement mechanisms in new ways have been emerging. 

Until recently, citizens’ interest and experience in the field of AI has been framed only in negative terms: as a lack of interest, or an information deficit. When engaging with relevant, value-laden and live topics, like in the case of AI, disagreement and strong conflictual beliefs should neither be excluded, nor overlooked and participation should never be seen as a means for overcoming resistance to already designed policies. 

Fifth - This is not going to be an easy one to implement 

We have definitely learned that engagement in strategy delivery is by no means an easy task! It requires the expertise, the infrastructure, the resources and the time that are needed for such a long-term endeavour. Hopefully the research report and recommendations that it presents will be useful for the Scottish Government first, and for others (both practitioners and government officials) who will be embarking on a similar journey of designing community participatory and engagement mechanisms for strategy delivery.