Scott Pruitt, the Environmental Protection Agency administrator, announced Monday his intent to withdraw the Clean Power Plan. Leaked reports suggested the EPA would scrap the CPP and ask for public input on how best to replace or change the rule. All of this will be based on a reassessment of the costs and benefits of one of President Obama’s landmark environmental rules.

The CPP is meant to combat climate change by decreasing carbon emissions from power plants. Obama’s estimates predicted the plan would hit coal generators hard but insisted that the benefits justified those costs. President Trump issued an executive order March 28 staying the implementation of the CPP and subjecting it to a re-evaluation of its costs and benefits. We expect the new analysis will show that politicians can’t separate politics from scientific assessments.

Under the Obama administration, the EPA estimated that the CPP would have climate and health benefits of between $20 billion and $34 billion. But, because the EPA used a high estimate for the social cost of carbon and estimated global climate benefits rather than just the national benefits, most people are expecting Trump’s EPA to estimate higher costs and lower benefits than Obama’s EPA, since Trump’s EPA will not rely on the same estimates of the cost of carbon.

Regardless of your political stance, this process should raise at least an eyebrow. Regulations are supposed to undergo thorough, scientific analyses. Ideally, those analyses are non-partisan. But when the costs and benefits of regulation are contingent on who is in the Oval Office, there’s an obvious problem. Regulations must be insulated (as much as possible) from these political pressures.

Simply put, agencies writing and implementing regulations shouldn’t be the same people who decide if those regulations are cost-effective.

The CPP is hardly the first time regulatory decisions have been determined by personal politics and not by scientific analysis. The politically controversial Dakota Access Pipeline, for example, was rejected by President Obama, but Trump quickly reversed that decision. Nothing about the pipeline changed between administrations, but the decision changed because it was made by someone with different political goals.

Big policy decisions aren’t always decided by the whim of the president — the Congressional Budget Office and Congressional Research Service both commonly provide nonpartisan and expert analysis of regulations and legislation. The CBO and CRS are insulated from the political pressures that are evident in decisions like DAPL, where the president was given a direct role in the decision to allow or prohibit the pipeline, and the CPP’s cost-benefit analysis, where the president can apparently direct the EPA’s to come to a conclusion that serves his political aims.

It turns out, it’s all about incentives. The CBO and CRS have reputations for doing nonpartisan and rigorous analysis. If they do shoddy work, they jeopardize their funding and thereby their existence. The CBO celebrates nonpartisan analysis as part of its culture. As Philip Joyce, a public policy professor at the University of Maryland, surmised after studying the CBO, “(Non-partisan) is the culture. It’s in the water.”

Now compare these incentives to those of the EPA. Scott Pruitt, the current EPA administrator, was appointed by Trump and confirmed almost perfectly along party lines. It is too romantic a view of politics to assume that people like Pruitt become interested solely in public welfare when taking office. Pruitt and others remain self-interested people with political opinions and ideas.

Since we can’t exclude the influence of politics on cost-benefit analyses of policies and regulations, we should instead subject them to intense scrutiny and re-evaluation by groups that have proven themselves to be less partisan than others. While the EPA has incentives to pass as many regulations as it can to expand the scope of the agency, the CBO, CRS and similar organizations have incentives to do good research.

Whatever comes of the Trump administration’s cost-benefit analysis of the CPP, it’s safe to assume, given the incentives of the analysts, that it won’t be non-partisan. Since effective regulation relies on impartial analysis, the EPA would do well to better insulate itself from the influences of politics and fall more in line with organizations like the CBO and CRS.