In my final year of graduate school, I was offered a fellowship to work in the U.S. Congress. During the interview process, I was asked to complete a peculiar assignment: design a policy to increase net retirement savings in the United States that won’t cost the federal government a dime. I pitched a modest plan to let Americans divert their tax refunds into retirement accounts—something not far off from the Obama administration’s now-defunct “myRA” plan. The proposal wasn’t particularly impressive and probably wouldn’t have improved the lives of many Americans, but it didn’t require any new federal spending–the crux of the assignment. I got the job.
At the time, I thought the test was just a kind of odd policy logic puzzle—a check on whether I could think on my feet. But after six years as a senior aide on Capitol Hill and more than a decade in Washington, D.C. economic policy circles, I’ve come to see that exercise in a very different light. It wasn’t an abstraction. It was a straightforward test of whether I could successfully function as a legislative aide in Congress, which is to say, could I create policies that would be approved by the hallowed budget scorekeepers at the Congressional Budget Office (CBO) and Joint Committee on Taxation (JCT).
As sociologists know well, institutions shape behavior and therefore outcomes, and the institution of legislative scorekeeping has quietly become one of the most powerful forces in American policymaking. On the surface, these offices provide technical estimates of how much a policy might cost. But in practice, they exert a kind of veto power over what ideas are considered viable in the first place. The result? A policymaking process distorted by technocratic constraints, unmoored from the values and needs of the people it’s supposed to serve.
When the Model Becomes the Master
To understand this dynamic, it helps to borrow a framework from sociologist Donald MacKenzie, University of Edinburgh, who once observed that financial models designed initially to observe markets (a “camera”) had begun to shape them (an “engine”). We’re seeing a similar phenomenon in Congress. Budget models were meant to offer neutral analysis. But over time, legislators began preemptively tailoring their proposals to fit those models, often to absurd effect. Congress, once the engine of democratic decision-making, now functions more like a mirror, reflecting only what the modelers say is possible.
This inversion has had serious consequences. Policies are watered down, delayed, or designed to “sunset” prematurely—such as the one-year expanded Child Tax Credit that expired on President Biden’s watch or this year’s expiring provisions of the 2017 Trump tax cuts—not because that’s what lawmakers want, but because it helps the bill clear the scorekeeping gauntlet. Entire investment categories, such as spending on long-term care for elders, are dead on arrival because their “scores” are politically unpalatable, even if their long-term benefits would far outweigh the price tag.
What’s more, scorekeepers, by convention, focus on the cost of a proposal over a 10-year budget window, routinely excluding benefits that would accrue in future years. Early childhood education, climate resilience, public health, these investments often pay dividends years down the line, but the CBO isn’t structured to count those gains. The costs, however, are immediate, visible, and heavily weighted. That asymmetry punishes forward-thinking policy.
Sociologists, with training in longitudinal analysis, cohort studies, causal inference, and more, should be helping to shape how Congress views the long-term effects of policy. This type of collaboration would ensure that legislators look beyond the cost of a policy to taxpayers to the real benefits to people and to our society. A policy that is “expensive” but dramatically reduces income inequality could well be a worthy trade. And of course, policies like taxing the rich have a “good score,” because they bring in substantial federal revenues. But as sociologists know, the benefits of eliminating extreme disparities of income and wealth in a society extend far beyond their effects on the U.S. Treasury.
Gatekeepers of Possibility
In the halls of Congress, this dynamic warps behavior. Legislative staffers are trained to write policy not for people, but for scorekeepers. Savvy aides spend weeks gaming out how to get a better CBO estimate, even cultivating relationships with scorekeepers to massage assumptions behind the scenes. During markup sessions, committee chairs will routinely reject amendments that “score,” regardless of merit. And members frequently refuse to co-sponsor bills unless there’s a favorable score attached, knowing full well that most proposals from rank-and-file lawmakers will never even get one.
In practice, that means entire swaths of policy never make it past the ideas phase, if staff bother to conjure them up at all—not because they’re unwise or unpopular, but because they’re deemed fiscally untenable. I was once asked to trim down a universal long-term care proposal to reduce its price tag, despite it being part of a broader Medicare for All bill already costing trillions. The exercise felt deeply unserious. But that’s how policymaking often unfolds: a performance for the models, not the people.
Rightsizing the Role of Models
This is not an argument for discarding economic models. We need them. But we also need to place them in proper perspective. Models are tools, not oracles. And policymaking must balance their projections with values, expertise, public opinion, and lived experience. Congressional offices track the opinions of constituents, recording whether they call or write in to support or oppose a particular policy. That should count for something. And when ten out of ten early childhood educators endorse a policy, that should carry weight, even if the CBO can’t quantify the return on investment in a spreadsheet. Causal estimates by sociologists of education demonstrating the benefits of a policy like universal pre-K or Head Start should count for something as Congress debates education policy. And community-engaged research by sociologists looking at how climate change harms frontline communities across generations should count for something as Congress debates how to address climate change. Fortunately, there’s a growing movement to reform the modeling regime. The National Academy of Sciences is investing in new tools to better incorporate climate risk into economic models, and the previous White House Council of Economic Advisers also prioritized this. New efforts, such as American University’s Institute for Macroeconomic and Policy Analysis, aim to develop models that reflect the actual contours of our economy, including market power, inequality, and financialization.
But better models alone won’t fix the problem. Congress must reclaim its role as the primary policymaker, not the CBO. That means making judgment calls, trusting experts, and listening to constituents even when the modelers’ estimates are incomplete or flawed. Legislators were elected to lead, not defer. And sometimes the right policy choice—saving the planet, expanding child care, housing our people—is so urgent, so obvious, that waiting for a model to catch up is a dereliction of duty.
Interestingly, Republicans in Congress are right now contemplating bucking decades of scorekeeping tradition to pass trillions of dollars in tax cuts for the wealthy and corporations under a “current policy baseline” framework. This is a technical term that means the Republicans don’t want to pay to extend the 2017 tax cuts slated to expire in December of this year. Although Democrats oppose this legislation (as do the vast majority of Americans), they could learn something from Republicans here. Sometimes, moving your agenda is more important than following arcane procedural rules.
Democracy on a Budget
Ultimately, this is a question of democracy. Constituents elect leaders to advance their interests, not to outsource decisions to technocrats armed with spreadsheets. Right now, we are living with the results of Congress’s self-imposed constraints: millions of Americans frustrated that they can’t afford the basics of their lives, from child care to health care to housing; we’ve underinvested in the future; and our policymaking process rewards small fixes over bold action.
We need better models, but more than that, we need more courage. We need lawmakers willing to shake off the invisible straitjacket, trust their mandate, and govern like the future depends on it—because it does.
Any opinions expressed in the articles in this publication are those of the author and not the American Sociological Association.