1 of 12

CMSC839C / INST878D -

Governing Algorithms and Algorithmic Governance

Class #11: Data-Driven Policymaking

2 of 12

Your 2nd Project Checkpoint

3 of 12

The role of the computer scientist in the policy table:

Data-driven policy enthusiasm

Detection

Measurement

Prediction

Simulation

Measurement

Ethics?

Causality?

The idea that government should base its decisions on data, evidence, and rational analysis is not new, of course. What’s new is the opportunity created by information technologies to crystallize problems and highlight effective solutions.

Computational Social Science (CSS), which brings together the power of computational methods and the analytical rigour of the social sciences, has the potential to revolutionise policymaking.

Large-scale data was formerly the province of the physical and life sciences, while social science relied mostly on qualitative data or survey data to understand human behaviour.

The data revolution from the 2010s onwards, where huge quantities of transactional data are generated by people’s online actions and interactions, means that for the first time, social scientists have access to large-scale, real-time transactional data on human behaviour.

4 of 12

The role of the computer scientist in the policy table:

Data-driven policy critique/dilemmas

#1 Scientists vs. Policymakers

Prajwal:

"Science cannot remain objective and outside of political influence once it starts making political decisions rather than advising on those decisions."

Tobias:

"Science isn’t normative and does not arrive at normative facts...Because science is not normative, scientists cannot govern in their capacity as scientists."

Sadia:

"I was surprised that the fourth reading, The big idea: should scientists run the country?, argued that scientists should not be calling the shots in a government, but that they should be off to the side"

5 of 12

The role of the computer scientist in the policy table:

Data-driven policy critique/dilemmas

#1 Scientists vs. Policymakers

“Democracy cannot dominate every domain – that would destroy expertise – and expertise cannot dominate every domain – that would destroy democracy.”

“In its obligation to embrace fallibility and uncertainty, science is antithetical to the current mode of politics in which admissions of doubt and error are regarded as weakness. Yet it is precisely because of those attributes that science is vulnerable to exploitation for political agendas.”

“The idea that scientists can speak truth to power in a value-free manner has emerged as a myth,” wrote the social scientist Sheila Jasanoff in 1990.

6 of 12

The role of the computer scientist in the policy table:

Data-driven policy critique/dilemmas

#2 Data-Driven Efficiency vs. Data Ethics

Miracle:

“Over reliance on data (specifically numerical and non-causal data) can create policies that lack human empathy and beneficial expertise…”

Rachel:

“Data-driven policy, while promising, isn’t universally accessible and may actually deepen inequities if we don’t consider ways to democratize access to these tools."

Hasara:

"Without an extensive discussion on interpretability and transparency in these models, the paper may overlook a key barrier: policymakers’ potential reluctance to rely on tools they cannot fully understand or explain to the public."

7 of 12

The role of the computer scientist in the policy table:

Data-driven policy critique/dilemmas

#3 Outsourcing Government [SV] vs. In-house development

Annie:

"I think that governments should be more than just a very large corporation, there is only so much that governments should take away from industry."

8 of 12

The role of the computer scientist in the policy table:

Data-driven policy critique/dilemmas

#4 Prediction vs. Causality

Anna:

"Within the social sciences, and computational social science especially, there's a quiet push to prove that social science is a "hard" science. Traditionally, this meant adapting social science techniques to fit the scientific method or proving social science theories using accepted "scientific" metrics like p-values. But there's also a push to adopt new computational techniques like machine learning, in order to be on the forefront of technology, even when those new techniques don't have a place within the existing epistemology. So now you have a tension within a tension among the people who want to demonstrate that social sciences can be rigorously scientific."

9 of 12

10 of 12

11 of 12

In-class activity: The RegLab

Choose a project from RegLab - realize what parts of the policy cycle it covers, and come up with risks & benefits in applying RegTech to improve Regulatory Governance in this case.

Detection

Measurement

Prediction

Simulation

Measurement

12 of 12

Next Week: RegTech Cases