More Like this

1. Finding, distinguishing, and understanding overlooked policy entrepreneurs

   https://doi.org/10.1007/s11077-023-09515-4  

   Arnold, Gwen ; Klasic, Meghan ; Wu, Changtong ; Schomburg, Madeline ; York, Abigail ( November 2023 , Policy Sciences)

   Scholars have spent decades arguing that policy entrepreneurs, change agents who work individually and in groups to influence the policy process, can be crucial in introducing policy innovation and spurring policy change. How to identify policy entrepreneurs empirically has received less attention. This oversight is consequential because scholars trying to understand when policy entrepreneurs emerge, and why, and what makes them more or less successful, need to be able to identify these change agents reliably and accurately. This paper explores the ways policy entrepreneurs are currently identified and highlights issues with current approaches. We introduce a new technique for eliciting and distinguishing policy entrepreneurs, coupling automated and manual analysis of local news media and a survey of policy entrepreneur candidates. We apply this technique to the empirical case of unconventional oil and gas drilling in Pennsylvania and derive some tentative results concerning factors which increase entrepreneurial efficacy.

    

   more » « less
2. Blunt force regulation and bureaucratic control: Understanding China's war on pollution

   https://doi.org/10.1111/gove.12485  

   van_der_Kamp, Denise_S ( March 2020 , Governance)

   In the past decade, China has increasingly resorted to top‐down, “blunt force” solutions to environmental problems; officials forcibly shutter or destroy factories to reduce pollution, at immense cost to local growth and employment. Why would a high capacity state like China resort to such a costly method of pollution control? I argue that blunt force enforcement is targeted at overcoming principal–agent problems in the bureaucracy. When institutions are too weak to hold bureaucrats accountable, political leaders increase oversight by drastically reducing the number of steps and resources required to produce a regulatory outcome—resulting in blunt force measures. I illustrate this logic through case studies from China's multiyear war on pollution. Using original data on blunt force measures, I then show how localities with lower bureaucratic compliance face greater blunt force pollution enforcement. These findings reveal that when bureaucratic accountability is weak, states pay a very high cost for cleaner air.

    

   more » « less
3. An investigation on requirement and objective allocation strategies using a principal–agent model

   https://doi.org/10.1002/sys.21511  

   Vermillion, Sean D. ; Malak, Richard J. ( October 2019 , Systems Engineering)

   In this paper, we investigate the dichotomy between system design delegation driven by requirement allocation and delegation driven by objective allocation. Specifically, we investigate this dichotomy through the lens of agency theory, which addresses cases where an agent makes decisions on behalf of another, that is, a principal. In current practice, design delegation largely involves requirement allocation as a means to inform agents of the desirable system characteristics. The value‐driven design paradigm proposes replacing requirements with objective, or trade‐off, functions to better guide agents toward optimal systems. We apply and adapt the principal–agent mathematical model to the design delegation problem to determine if a principal, that is, the delegator, should communicate using requirements or objectives with her agent. In this model, we assume the case of a single principal and single agent where the agent has certain domain knowledge the principal does not have and the agent accrues costs while solving a delegated design problem. Under the assumptions of the mathematical model, we show that the requirement allocation paradigm can yield greater value to the principal over objective despite limitations requirement allocation places on the principal to learn information from the agent. However, relaxing model assumptions can impact the value proposition of requirement allocation in favor of objective allocation. Therefore, a resolution to the requirement–objective dichotomy may be context dependent. The results and the analytical framework used to derive them provide a new, foundational perspective with which to investigate allocation strategies.

    

   more » « less
4. Undermatching Is a Consequence of Policy Compression

   https://doi.org/10.1523/JNEUROSCI.1003-22.2022  

   Bari, Bilal A. ; Gershman, Samuel J. ( December 2022 , The Journal of Neuroscience)

   The matching law describes the tendency of agents to match the ratio of choices allocated to the ratio of rewards received when choosing among multiple options (Herrnstein, 1961). Perfect matching, however, is infrequently observed. Instead, agents tend to undermatch or bias choices toward the poorer option. Overmatching, or the tendency to bias choices toward the richer option, is rarely observed. Despite the ubiquity of undermatching, it has received an inadequate normative justification. Here, we assume agents not only seek to maximize reward, but also seek to minimize cognitive cost, which we formalize as policy complexity (the mutual information between actions and states of the environment). Policy complexity measures the extent to which the policy of an agent is state dependent. Our theory states that capacity-constrained agents (i.e., agents that must compress their policies to reduce complexity) can only undermatch or perfectly match, but not overmatch, consistent with the empirical evidence. Moreover, using mouse behavioral data (male), we validate a novel prediction about which task conditions exaggerate undermatching. Finally, in patients with Parkinson's disease (male and female), we argue that a reduction in undermatching with higher dopamine levels is consistent with an increased policy complexity.

   SIGNIFICANCE STATEMENTThe matching law describes the tendency of agents to match the ratio of choices allocated to different options to the ratio of reward received. For example, if option a yields twice as much reward as option b, matching states that agents will choose option a twice as much. However, agents typically undermatch: they choose the poorer option more frequently than expected. Here, we assume that agents seek to simultaneously maximize reward and minimize the complexity of their action policies. We show that this theory explains when and why undermatching occurs. Neurally, we show that policy complexity, and by extension undermatching, is controlled by tonic dopamine, consistent with other evidence that dopamine plays an important role in cognitive resource allocation.

    

   more » « less
5. Model-free Learning with Heterogeneous Dynamical Systems: A Federated LQR Approach

   Wang, H. ; Toso, L.F. ; Mitra, A. ; Anderson, J ( August 2023 , arXivorg)

   We study a model-free federated linear quadratic regulator (LQR) problem where M agents with unknown, distinct yet similar dynamics collaboratively learn an optimal policy to minimize an average quadratic cost while keeping their data private. To exploit the similarity of the agents' dynamics, we propose to use federated learning (FL) to allow the agents to periodically communicate with a central server to train policies by leveraging a larger dataset from all the agents. With this setup, we seek to understand the following questions: (i) Is the learned common policy stabilizing for all agents? (ii) How close is the learned common policy to each agent's own optimal policy? (iii) Can each agent learn its own optimal policy faster by leveraging data from all agents? To answer these questions, we propose a federated and model-free algorithm named FedLQR. Our analysis overcomes numerous technical challenges, such as heterogeneity in the agents' dynamics, multiple local updates, and stability concerns. We show that FedLQR produces a common policy that, at each iteration, is stabilizing for all agents. We provide bounds on the distance between the common policy and each agent's local optimal policy. Furthermore, we prove that when learning each agent's optimal policy, FedLQR achieves a sample complexity reduction proportional to the number of agents M in a low-heterogeneity regime, compared to the single-agent setting. 

   more » « less