Turn Data into policy

Turn data into policy

How to get started

Definition and Differentiation: A traditional policy vs. a data-driven policy

A traditional policy is generally defined as a rule, a regulation, a law, an administrative action or an initiative led by governments at all levels. The traditional policymaking cycle is a full-fledged process that involves assessment of previous policies, systems, and services; design and delivery of new policies and rules, validation and political approvals, implementation, monitoring and impact evaluation of these policies with an objective to achieve short-medium and long-term goals of a particular department/ministry. A data-driven policy is a set of rapid policy actions that lead to improvements in the functioning of policies and influences budget decisions leading to cost-saving and optimum utilization of resources in the right geography, on the right challenge(s) and for the right set of beneficiaries within appropriate timelines. With the Data to Policy Navigator, our aim is to support you with the journey of translating data to policy actions (and not formulating an entire policy from scratch). Data-driven policy actions help you with evidence-based decision making, supporting ‘proactive’ policy decisions rather than ‘responsive’ decision making. Backed up with data analytics and visualization, your policy actions won’t just help you save time and resources, but also help you build better political buy-in and demonstrate deeper impact.

Key term: Data-driven policy

Rapid policy actions with data analytics and visualization
Enables ‘proactive’ decision-making as against ‘reactive/responsive’ decision-making
Helps save time and resources
Mitigates risk of delays in obtaining political buy-in and approvals
Demonstrates quality and impact by harnessing the power of data

How to implement: what does turning data to policies mean?

As a policymaker, designing and delivering effective policies is your key role. What comes to mind when you think of non-traditional and more efficient ways of designing policies by turning existing data into policies? A few ideas you may be familiar with in more recent times are policy dashboards, algorithms and modification of public service delivery systems, platforms and so on. Most of these words are so technical that as a policymaker, it’s obvious for you to outsource these tasks. While the task of using the data and codifying it in the form of policies, using platforms, stacks, algorithms, workflows, AI and other digital tools continues to remain technical. The most critical task is making functional use of data in the overall policymaking process. In your policymaking journey, you’re either trying to build on or improve existing policies. To turn your ‘data to policy’ journey simpler, the navigator will help you recap and take the next steps towards turning data to policies, using an example. Here’s how you can get started:

Step 1: Quick Recap

In the previous section of the navigator, you learnedhow to define the problem. By now, you have a strong understanding of the pain points of your policy. You’ve defined the problem(s) in the policy ecosystem, you have your data ecosystem mapped, and you have clearly identified the gaps by now. Until this stage, you were focussing on building data partnerships and leading your team to support you in analysing and visualizing the data. As you move ahead into the actual policy design stage, you’re all set to start putting different pieces of the puzzle together to help you design your policy.

Figure 1: Your journey towards turning data to policy

As we begin blending your policy pain points with data, you’ll start to experience how you can leverage data and convert data-driven policy into a win-win opportunity for yourself as well as all stakeholders of the ecosystem.

Step 2: Combining Policy and Data

As a policymaker, you want to avoid reinventing the wheel and try working on a cost-effective policy design, unless you’re designing a policy for a completely new theme or sector. Having most of the ingredients in place to turn data into policy, it’s time to begin a collaborative policy design exercise and combine your policy with data. You may like to begin this process by involving multiple line ministries, your target beneficiaries/end-users and other relevant stakeholders in the policy design process to ensure policy coherence, collective impact and optimum utilization of existing resources and functions. Compared to traditional policy designs, your data-to-policy design comprises of programme actors as well as data actors. The task is to explore how data can help solve your policy challenges and how you can combine data insights to make your policies ‘proactive’ rather than ‘reactive’.

Turning Data to Policy: Combining policy and data solutions to address development challenges

Step 3: Policy Formulation and Validation

Policy formulation is directed towards the identification of shared objectives and alternative options for intervention in relation to the problem defined in the previous phase.

In this step, predictive and prescriptive analysis methods can be employed. For a given policy problem several possible policy measures might exist. In this step, the pros and cons of every measure are investigated and eventually one policy measure is chosen. Experimental iterations in the policy making process can be used to diagnose the effect of different scenarios. At the same time, these iterations are a good opportunity to gather data, validate and refine predictive models.

You now have your policy issues and pain points identified, some corresponding data (and other) resources mapped, and you have some policy solutions and measures identified. It’s time to start putting these pieces together to formulate your policy as follows:

Figure 3: Process of policy formulation

The pilot of Issy-les-Molyneaux (France) developed a mobility dashboard to visualise and identify the most important congestion points in a detailed way to support the policy makers in taking decisions and defining policies. As a first result, it became clear that the bulk of traffic originates from cars passing through, and not from inhabitants of Issy-les-Moulineaux. In collaboration with a local start-up an application was tested in congested areas to propose different paths, defined by the city on precise data (and not just on algorithms), and to communicate to drivers in real time.

‍

Once you’ve formulated your data-driven policy, you may decide to have policy validation done by socializing and embedding feedback loops for your policy intervention with political leadership, citizens and other stakeholders, using data visualization maps that display the outputs of the proposed policy interventions.

Some value addition approaches you may want to consider in your policy formulation:

Data for participatory policymaking
• Whole of Society approach
• Enhance trust between citizens and government using data
• Including all stakeholders in the policy design stage
• Example: New Zealand Government Guide
Data for anticipatory policymaking
• Design shock-proof, future-resilient policies using strategic foresight
• Example: Singapore, EU Parliament, UK, Netherlands, Finland and Korea
Using ethnography data for policymaking
• Generally involves smaller numbers of people
• Provides deeper insights from behavioral science to be fed into policies
• Example: How the UK is using ethnography and big data in their policies

Step 4: Policy Implementation

After the scoping, design and formulation phase of turning data to policy, you’re now ready to roll out your policy implementation. Governments have their own internal, customized guidelines and protocols for policy implementation. Unlike traditional policy implementation strategies, your data-driven policy intervention needs more concurrent planning and monitoring. Begin by designing a Policy Implementation Plan (PIP) and an implementation system/platform that enables you to track progress of your policy both quantitatively and qualitatively:

Start planning and executing preparedness workshops with all line ministries and stakeholders based on consensus built to co-design new policies and fill policy gaps with data.
Set data sharing and tracking protocols across all stakeholders.
Design and agree on governance frameworks across the policy implementation plan for all government and non-government stakeholders: who leads, who coordinates, who shares, who takes responsibility, and who takes action(s).
With the help of your IT/Data team, code the plan with algorithms that enables data tracking and allows you to roll out as well as monitor each policy action.
Your IT/Data team could help you by designing a platform that allows you to visualize the data to predict/anticipate key challenges and bottlenecks of the policy implementation and take proactive decisions and corrective measures to address them in advance.
Your IT/Data team could also help you automate the platform to convey alerts/risks/delays in achieving certain indicators to all relevant stakeholders for timely action.
As an advantage of implementing a data-driven policy, you could explore use of data to track the quality of the implementation of your policy intervention and can take proactive measures in addressing connected issues and challenges.

The implementation plan is necessary for policy implementation to be as effective as possible. While doing this the opportunities for data-based monitoring activities should be considered. When a data collection plan is included in the implementation plan, data collection infrastructures need to be designed together with data analysis for the implementation step.

In Mechelen (Belgium), a regional traffic model is being used to study the traffic in the city. The pilot has two objectives: The first objective is to provide traffic modelling for the city. Mechelen, together with the police zone Mechelen-Willebroek, has integrated an existing traffic model and aims to enrich its usefulness by adding data from ANPR (automatic number-plate recognition) cameras. The traffic model will be used as a predictive tool to monitor the impact of planned road works in the city. The second objective deals with the recently introduced “school streets” where traffic calming solutions such as road closures are in place. The aim is to measure and analyse traffic and congestion variation (both in the closed streets and the neighbouring streets) before and after implementing the measure. A recent policy decision introduced the concept of “school streets”, streets that are being closed at the beginning and at the end of the school day. The traffic model combined with local traffic count data measures and analyses the impact of school streets on traffic behaviour in and around the school streets. In this step, data can be useful to guarantee the full impact of the policy implementation and the achievement of the policy goals.

Since you’re now familiar with the step-by-step process of turning data into policy, let’s apply the steps to a particular policy challenge to find solutions to your policy challenges.

Scenario: Youth Employment Department

Imagine you have recently joined as the head of the Youth Employment Department. You realize that there are limited policy interventions on mapping skills of the youth to help them understand their career choices better - whether they are employable for jobs or possess entrepreneurship skills? You want to use data and take a policy decision that helps the youth make informed choices about their careers keeping in the mind the last decade where the skills needed in the market have considerable evolved.

You’re about to design policy intervention(s) on mapping skills of the youth to help them understand their career choices better - whether they’re employable for jobs or possess entrepreneurship skills.

Here’s how to apply the data to policy navigator steps to this policy problem:

‍Step 1: Define the problem

You have defined the problem and listed all pain points of the youth employment policy:

There’s a high unemployment rate at the national level. The existing policy isn’t helping to improve the numbers. There’s a high rate of dropouts from jobs after skill trainings and placements. You’re receiving complaints/feedback from employers/industry associations that they aren’t getting youth skilled enough to be retained in jobs.
After brainstorming with all relevant stakeholders, you realize that there’s a mismatch of skills and expectations at both ends - employers and youth. You also realize that the earlier policy never covered this problem, never focused on mapping skills of the youth and never had a mechanism to map expectations/feedback from employers.
You map the data ecosystem and learn there was some data captured by your department (and other related departments) on mapping skills, perceptions, expectations, experiences and feedback from both - youth and employers.
Some data on this challenge was also captured by foundations, academia and private sector working in this field. By now, you have identified both policy gaps (by defining the problem) and data gaps (by mapping your data ecosystem).

Step 2: Identify data sources

You begin to identify data sources by filtering all reliable sources of data that can help you map the skills mismatch at both ends.
After closely studying the data ecosystem map, you realize that few provinces across the country have more development partners helping with value additions and programmes on youth employment.
You have ensured establishing data partnerships with all stakeholders that help you access and collect data.

Step 3: Analyse the data

When looking at the data, you find:

Youth from rural areas tend to dropout easily and quickly from jobs: What could be your next policy intervention for this? Commissioning quick research/task force on identifying challenges/needs assessment or allocating more budget on soft skills trainings, awareness camps, and internships in rural areas?
A major percentage of employers stopped hiring from government skill training programmes: Is there a policy challenge or an implementation challenge that’s affecting the outcome and quality of skill training programmes? Do I have policies that help me track the quality of faculty, trainers and even the curriculums that are being used and applied to train the youth?

Step 4: Combining Policy and Data

Step 5: Policy Formulation and Validation

Based on the scenario above, you could plan policy formulation and validation as follows:

Your mandate is to bridge the skills mismatch and improve the rate of unemployment in the country. You aim at targeting X% rate improvement and Y number of youth making informed career choices. After setting the vision, mission and KPIs you could move backwards to design an action plan.

You ensure you’re looping in all stakeholders in the employment ecosystem, especially the ones who could help you with data collaboration and chart out terms of references and scope of work towards achieving the mission of X% rate of unemployment by bridging skills mismatch of Y number of youth. Together, with stakeholders, you design policy interventions to meet your targets:

Design a platform/dashboard that helps you filter data of the regions that need more efforts on quality skilling programmes to bridge the skills mismatch.
Add policy options, indicators and data-flow mechanisms that enable you to map the skills of the youth in the current context.
Add policy options and indicators that help you assess the feedback of target beneficiaries and stakeholders on the quality and impact of your current mobilization strategies, aptitude assessments, career counselling strategies, training curriculums, programme designs, quality of trainers, placement strategies, placements cells, etc.
Add indicators that help you track the employment ecosystem demands including the future skills needed (STEM, AI, Blockchain, UX/UI, etc.) for you to take policy decisions on new courses and training programmes that help improving the employment KPIs.
Add policy options for consistent enhancement of modules for mobilization, career counselling, skill trainings, etc. to meet changing needs of the employment ecosystem.
Add policy and data options to track and take corrective action for special value-added programmes for youth belonging to vulnerable, underprivileged sections of the society, youth with disabilities, etc.
Add policy options for reskilling, upskilling, multiskilling youth based on wide range of options available in the employment ecosystem.

Step 6: Policy Implementation

Based on the platform design, existing policy interventions and human resources delivering your policy on the field, design a policy implementation plan. Different countries have a different set of policy implementation strategies and protocols. Set goals – short-term, medium-term and long-term for your policy. When you roll out implementation and start getting interim data, you’ll be able to take larger policy decisions:

Whether you need to reallocate special budget for more scientific need-based aptitude tests, career guidance and counselling sessions, upgraded curriculums and enhanced trainings to match the skills mismatch between available jobs and employable youth.
Whether your department needs to anchor designing customized programmes and awareness camps to serve the unemployment youth’s expectations better so as to help them make informed choices.
Whether your department officials and field machinery need special trainings for enhancing quality in delivering services and functions.
Whether you need to make strategic investments on R&D to be able to do advocacy for cross-ministerial policy designs (e.g. life skills journey model of skilling a youth from school to college to technical skill qualifications).

What’s next? ‍

Now that you have designed and formulated your policy, you want to ensure its smooth implementation and monitoring as well as its success with impact. Read our next pieces on: