Data and Environmental Policy Response

This is a response to Aaron Harburg’s Data and Environmental Policy.

Environmental Law is predicated on the assumption that the government can effectively motivate its citizens to perform or abstain from certain activities.

All laws which criminalize behavior are coercive and part of the failing justice system.

The question is how incentives and punishments should be used to achieve and ensure the greatest effect.

That’s a good question.

There is evidence to suggest that excessive reliance on the threat of punishment alone is not only insufficient but counterproductive.

A link to the evidence would be cool.

Furthermore, without having clear metrics, it is unreliable to use any mixture of deterrent or incentive.

The good question here would be how to run the experiments and collect the data. I think this is the most important question. And I think the answer has something to do with decentralization, freedom, and transparency.

The US EPA and other state and local regulatory agencies do not have sufficient resources to thoroughly police all individuals and organizations to ensure total compliance.

This is a problem for all law enforcers trying to enforce laws against criminal behavior. It just doesn’t work and is unnecessary when there is a mechanism for people to make claims of injury (a good data source which is absent in our society).

Why do these occur? In large part, it is because the regulatory and economic circumstances converged to incentivize individuals and organizations to take the risk of non-compliance. The risk of non-compliance appeared less likely than the reward. When government regulations produce perverse incentives they need to be restructured if they are going to be effective. Does that mean harsher penalties? Perhaps not. B.F. Skinner discovered with a variety of organisms that the threat of punishment is less effective than using various schedules of positive reinforcement. Further studies in management have revealed there are ideal ratios of positive reinforcement and criticism. There is further evidence of this with the adoption of electric vehicles due to government incentives.

It’s good to think about punishments and incentives. But it is also good to think about root cause problems and solutions. Some of the root cause problems related to emissions has to do with the need for transportation. And the need for transportation is related to things like supply chain and remote work opportunities. So you can craft laws which facilitate decentralization for new intentional communities to emerge in a free society where data is collected and experiments can be run.

Any approach should involve meticulously collecting and analyzing quantifiable data to demonstrate with a reasonable degree of statistical probability if a given regulation or incentive is actually achieving its intended effect. In many respects, the relative success of the Clean Air Act was the insistence on quantifiable data.

I am curious to see a link to the quantifiable data and how it was collected.

Another aspect of the CAA was the tangibility of what they were trying to avoid. It is much easier to grasp why thick smoke is harmful to health than why invisible emissions are causing climate change. More so than other pollutants proposals to reduce greenhouse gas emissions must involve the development and adoption of technology. That necessarily means any Environmental Law must correctly produce the ratio of punishment to the incentive.

Good and smart people don’t have the opportunity to solve root cause problems associated with supply chain and remote work. We have a system of government which selects the wrong people for the job.

In the 1990s President Clinton initiated the Clean Car Program to spur green innovation within the automotive industry. This led to the EPA developing technology with the public sector in public partnerships through their Clean Automotive Technology program. Despite some of this technology being commercialized, the program was eventually terminated. The primary reason for this shutdown was because of greater regulatory power during the Obama administration. What was the point of using taxpayer dollars to help companies innovate when the government could just force them to increase fuel efficiency? I argue that kind of thinking is what lead to the disaster with VW. It also enabled the current administration to shut out valuable research for establishing fuel economy standards.

I don’t think you are wrong. But I also think your analysis is missing systems thinking and root cause analysis.

A better way to draft an effective policy is to approach the problem from a multi-disciplinary approach and have concrete metrics associated with each discipline. The problems that policy attempts to solve involve just as much complexity, if not more, than the vehicles that those policies are meant to regulate. Every vehicle is composed of thousands of parts from a long supply chain of plastics, metals, electronics, aerodynamics, material science, and more. An inaccurate model will lead to a vehicle that cannot drive or will quickly fall apart. Likewise, for the policy, there need to be network models that involve at minimum variables from disciplines of psychology, sociology, game theory, macroeconomics, supply-chain, and manufacturing. Yet, most of the policies that are introduced reflect a model of governance that still has not grappled with the vast potentials of data science and machine learning.

People in government just aren’t competent enough to do this. And so there must be government reforms which allow the competent people to run the experiments, discover better ways, and share the evidence.

At present many policies appear to be drafted from a simplistic view of wanting to achieve a simple metric, say increase fuel-economy, through bureaucratic enforcement and incentives. A metric that ignores the greater environmental harm that may be occurring as a consequence of attainment or nonattainment. For the policy of the future to be effective it must be as sophisticated as the reality it hopes to regulate and as precise as the goal it seeks to achieve.

It concerns me that incompetant people are drafting the policy of the future. I would prefer that they stop trying to draft policy and start trying to draft people, ideas, and experiments.

The startup world has demonstrated customer discovery is essential. Virtually any product is doomed to fail, no matter how innovative or sophisticated, without first talking to the end-user. A more sophisticated technical model combined with a willingness to actually speak with the people these policies aim to influence is where it all comes together.

Most startups fail. And the government needs to allow many experiments to happen so that most of them can fail safely while the good ones are identified and used to benefit the world.

Ultimately policy should be crafted using sophisticated models with tangible metrics that recognize the incentives across multiple domains through conversations with the people it aims to regulate and motivate. Without a robust balance of interpersonal communication and statistical analysis, we risk not only failing to achieve our goals but potentially making things worse.

Yes. And we ultimately need thousands of great problem solvers running experiments and collecting data while knowing that most are going to fail.

Disclaimer prior to 2023-03-25: being neurodivergent, I sometimes fail to communicate my thoughts clearly.
Disclaimer beginning 2023-03-25: posts are being generated with the help of GPT.