The IRS has problems that $80 billion won't solve

The recent Democratic bill on climate and taxes increases the funding for the Internal Revenue Service by an additional $80 billion over 10 years, with the purpose of collecting more revenue by enforcing existing tax laws. Bloomberg reports that the agency will use the money to "add auditors, improve customer service and modernize technology." I have general sympathy for this idea, but the more I read from its defenders the more wary I feel.

The general chorus is that the IRS doesn't have the people or technology needed to perform its mission. For instance, the nonpartisan Center on budget and Policy Priorities says that budget cuts have left the IRS "depleted," with the agency losing 19% of its budget and 22% of staff between 2010 and 2021. The agency has also been struggling to replace workers who leave.

A recent Washington Post article by Catherine Rampell details the backward state of information technology at the IRS, where "paper tax returns aren't scanned into computers; instead, IRS employees manually keystroke the numbers from each document into the system, digit by digit."

And the GAO has criticized IRS cybersecurity software and called an essential technology upgrade "troubling" given that it's not slated to finish until 2030. The agency also uses fax machines, an obsolete version of Windows, and the programming language of COBOL, which was designed in 1959 and standardized in 1968. While COBOL is still used in the private sector, such as in many banks, it is ill-suited for the kinds of innovations that have come to the American payments sector through Silicon Valley, starting with PayPal.

It's easy to say that the IRS has not had the staff or the money to do the necessary upgrades. But hold on: These software upgrades are supposed to save money by enhancing productivity, letting organizations do more work with fewer people. A reasonable person can be forgiven for asking whether an agency with a $13.7 billion budget really doesn't have enough to front some cash.

You might argue that IRS was too liquidity-constrained to shell out the cash up front, but is that argument believable? The improvements from better software usually pay off rather quickly, precisely because the software is labor-saving. The US has plenty of small to mid-sized businesses and non-profits with shrinking staffs and budgets. Yet most of those institutions have been able to upgrade to better software, often repeatedly. Unlike the IRS, many state tax agencies at least use scanners, and those are hardly the wealthiest or most nimble institutions in American society.

When I see that the IRS reduced staff by 22%, I imagine an alternate reality in which the IRS had replaced a good deal of its office staff with better information technology, as many American businesses started to do in the 1990s. In this parallel universe, the staff of the IRS is down and the productivity of the IRS is up, as has happened to so much white-collar office work. But that is not the world we live in.

The advocates for additional funding should better understand why not everyone in America is thrilled with the agency's new budget boost. It's not just a bunch of kooks who fear "an army" of weapon-toting IRS agents, or rich people who feel they shouldn't have to pay their fair share. It's normal people who think it's a bad idea to reward an agency that seems so dysfunctional.

For example, according to the GAO's exhaustive report, a major milestone in upgrading the agency's 60-year-old database has slipped nine years, from 2014 to 2030, despite the project being radically scaled back. The agency has also "revised the program's cost, schedule, and scope goals on numerous occasions, including seven times between 2016 and 2019." That does not inspire much confidence.

To be clear, I am on board with the basic idea that we ought to enforce existing tax law better rather than creating new tax hikes. Former treasury secretary Larry Summers and others have put forward compelling arguments for that conclusion. But are we so sure that another $80 billion is what will induce superior IRS performance?

I propose an alternative. Plenty of government agencies, or for that matter private institutions, have improved their performance without extra money, whether in America or abroad. So tell the IRS to start moving into the modern age, and after they have shown they can make better progress, give them the money they need to finish the job.