Evidence-based policymaking: Can technology reshape the legislative process?

By Rajesh Jaluka

There’s been a notable increase in demand for legislation to guide the future of generative AI technology. This renewed interest in the role government plays in technology raises the question: does technology influence legislation or vice versa?  

I recently read an article that caught my attention. Not because it called for the US government to regulate technological innovation but instead to leverage technology to shape the legislative process. US lawmakers are now pushing to establish a bipartisan commission that calls on experts to collect, review, analyze, and make recommendations to Congress using data to drive evidence-based policymaking.¹

As a proponent of truth in data, I applaud the lawmakers advocating for evidence-based policies. As a technologist who has worked with the public sector for nearly two decades, I’m passionate about the role technology can play in helping lawmakers deliver on their evidence-based promise.  

The promise of evidence-based policymaking

The new initiative would not only create more transparency, but also enable lawmakers to harness the power of data to gain deep insights for future policymaking and to prioritize the services government delivers to its citizens. But data can be challenging to find across different systems and organizations. It can be even harder to clean and normalize its format, all of which creates huge barriers for lawmakers hoping to leverage data for their policymaking.

Technology can change that. Here are just a few examples of how a modern, robust technology foundation can provide the necessary tools to leverage existing data for effective policymaking:

• Extraction: Data lives in diverse formats across different organizations and systems. Data integration technology has significantly matured to make it easier to extract from these diverse sources.
• Normalization: Data in these systems could be in varying structured and unstructured formats. Advances in machine learning can help normalize the structure and bring uniformity.
• Cleansing: Data quality vastly varies due to differences in validation rules or lack thereof. Various algorithms and statistical methods can cleanse the data by identifying missing values, duplicate records, and inconsistent formats, removing duplicates, enriching existing data, and converting to a standardized format or structure.
• Analysis: Data science has recently emerged as a multi-disciplinary field that integrates domain knowledge with statistics and computer science to provide deep insights not only from structured data but also from noisy and unstructured data.

Seven steps to evidence-based policymaking

Throughout my career, I’ve shown healthcare organizations, businesses, and governments how they can harness existing data to deliver meaningful outcomes for their constituents. Based on this experience, I’ve outlined the seven steps policymakers, CIOs, and business stakeholders can follow to implement evidence-based policymaking on behalf of citizens.

1. Policy assessment: Establish what is needed in terms of policy to address citizen priorities.
2. Data collection and analysis: Determine what relevant data is needed to inform the policy and put the data into a usable format.
3. Data examination: Explore and understand what data is collected, then identify trends, patterns, and visualizations.
4. Identify policy gaps: List which policy gaps exist and determine where there are challenges and opportunities.
5. Policy recommendation: Provide options to address citizen priorities and make sure to understand potential outcomes.
6. Policy roll-out and implementation: Ensure the execution of policy and track progress.
7. Design with an iterative approach in mind: The job is never completed. You should always be refining based on data that continues to shape policymaking.

A use case for evidence-based policymaking

The use of data can play a crucial role in shaping policies across various domains of the public sector. For example, healthcare policies aimed at enhancing the quality, safety, and delivery of care, or education policies that address issues of equity, curricula, and infrastructure. For the purposes of this article, let’s consider a policy that promotes investment in sustainable energy.

Many US states are currently exploring various approaches to promote sustainable energy investments. Ideas include offering subsidies to residents and tax credits to businesses that install renewable energy systems in public buildings. To provide a truly evidence-based recommendation, lawmakers need to consider all relevant data sources on the topic.

The sourced data does not have to be limited to government-owned data. Data from private companies and universities can also be used. For example, here are some sources of information for lawmakers to consider:

• Stanford’s DeepSolar provides insights into communities where the adoption of solar panels is lacking. One could tie this data with socio-economic data to target different incentives for different communities. 
• National Solar Radiation Database (NSRDB) provides historical data on solar radiation for any location in the United States. This data can be used to predict the potential solar energy available in that location. 
• MDPI’s paper offers a model that can be used to determine the value of integrating large-scale battery energy storage into current and future microgrids.
• The US Energy Information Administration reports that lighting accounts for 17% of all electricity consumed in US commercial buildings. Energy efficient bulbs can cut down the demand by up to 90%, thereby requiring fewer solar panels.²

By pulling a diverse set of information from a topic, lawmakers can design a more holistic energy policy that goes beyond solar panel installations and integrates energy-efficient lighting and microgrids.

Simulations in lieu of data

There are many sources of renewable energy beyond solar, such as wind, geothermal, and biomass. To ensure they’re recommending the best policy for constituents, lawmakers need to determine how each source will provide the optimum return on investment. This type of data may not be readily available, and lawmakers may need to do scenario simulations for various combinations.

Further, these simulations may need to factor in variables like socio-economic status, geographic terrain information, and solar radiation information. Policymakers may also need to conduct field trials of these scenarios to confirm the results of the simulations before encoding them into the policy.

As you can see, with the number of variables, it is difficult to create a sustainable energy policy that can anticipate all issues. This is the challenge policymakers are facing. It will need to be tailored by region and community. Therefore, starting with smaller ideas and using the results of the field trials to adjust will enable governments to maximize the benefits of these options.

For over two decades, modern technology companies have been using agile methods to incrementally deliver features and capabilities to their users. Similar principles could be applied to policymaking where policies are incrementally delivered and evidence of experienced and measured outcomes guides the next iteration.

The role of technology in government

I began this article with a question: does technology influence legislation or vice versa? While I believe this will be an ever-evolving relationship, one thing is clear: at the center of effective government are evidence-based policies backed by data—and technology will enable the use of data to improve the way government services are legislated, designed, and implemented for citizens.

Rajesh Jaluka is Chief Technology Officer for US Public and Federal markets at Kyndryl.

¹  House lawmakers introduce bipartisan evidence and data-based policymaking bill, FedScoop, May 2023
²  Trends in Lighting in Commercial Buildings, U.S. Energy Information Administration, May 2017