Policy-makers should ignore it and roll back the pandemic-era expansion.
Policy-makers should ignore it and roll back the pandemic-era expansion.
I t’s an ironclad rule of politics that one can find an economic-impact study supporting any government program, no matter how misguided the program is. That’s especially true when the program in question is supported by powerful interest groups and politicians.
The Affordable Connectivity Program (ACP) is no exception. It’s a welfare program that subsidizes households’ internet bills up to $30 per month. It was created during the pandemic on an “emergency” basis (no emergency ever existed) and then renamed and expanded. It currently covers about 23 million households but is scheduled to run out of funding in April. The Federal Communications Commission (FCC), which administers the program, stopped allowing new enrollment on February 7.
Congress should not renew the funding, and it should let the program expire. But the Biden administration, a bipartisan group of legislators, and various lobbying groups and telecoms want the funding to be renewed and the program to be extended.
One economic-impact study that purports to show significant economic benefits from the ACP is by Matthew Sprintson and Edward Oughton. The paper also looks at other broadband-subsidy programs started in the past few years, but we’ll focus on the ACP portion of the paper.
Sprintson is a high-school student at Thomas Jefferson High School for Science and Technology in Alexandria, Va. Oughton is a geography professor at George Mason University. Oughton has academic-research experience on internet issues, but he’s not an economist. The lack of economic credentials for either author, and lack of a college education for one author, ought not disqualify their paper out of hand. Let’s judge their work for what it says.
The literature review establishes that broadband internet access is good for economic growth, but it does not do so in a way that addresses the specific situation of the United States or the ACP. For example, it notes that “increased broadband penetration leads to positive impacts on GDP after more than half of the population has gained access to the Internet.” According to the FCC’s most recent Broadband Deployment Report, in 2019 (before the pandemic and before the ACP took effect), 95.6 percent of Americans had access to fixed terrestrial broadband internet. (If you include 4G LTE mobile-internet access, that proportion rises to 99.9 percent.)
The paper cites over 100 articles, but it inflates the number of references by, for example, providing three citations to back up the claim that “broadband infrastructure investment expands opportunities for entrepreneurs and encourages business creation.” Backing up even obvious statements with citations is conventional in academic writing, but this example is representative of the quality of many of the observations made in the paper. The literature review dodges the relevant question: whether existing research shows that it is a good use of government resources to distribute subsidies to households to expand or improve internet access that is already almost universal.
Sprintson and Oughton have the honesty to include that “for regions that already have significantly comprehensive broadband infrastructure, there are generally diminishing returns to scale, with increased infrastructure spending leading to diminishing returns on investment.” But they don’t address whether the U.S. might be such a place.
Instead, they proceed to estimate outlandish economic benefits from the ACP. They say the $14.2 billion spent on the ACP so far has increased GDP by up to $55.2 billion. They posit a Keynesian multiplier of 3.89 for the program.
A Keynesian multiplier is calculated by dividing an increase in GDP from government spending by the amount of government spending. Government spending is a component of GDP, and then there are additional knock-on effects from government spending that could increase GDP further. In this example, the idea is that the ACP subsidy frees up more money for households to spend. When they spend that extra money, that counts as income for someone else, who then has extra money to spend as well, etc.
The Keynesian multiplier is often little more than magical thinking by advocates for big government. If it were really true that $1 in ACP spending creates $3.89 in economic growth, which is what the paper’s Keynesian multiplier of 3.89 would mean, then Congress inadvertently discovered a way to hack the economy by creating an internet-subsidy program during the pandemic. (For perspective, a 2019 study from the Department of Agriculture estimates the multiplier from SNAP, the federal food-subsidy program for low-income households, to be 1.5.) Why stop at $14.2 billion? Congress should put a sizeable portion of the federal budget into the ACP if it’s getting that kind of return.
Of course, it isn’t. Sprintson and Oughton don’t even do Keynesian economics well on Keynesians’ own terms. Keynesians at least understand that if people receive extra income, they save a chunk of it. They postulated something called the “marginal propensity to consume,” a proportion that reflects how much people spend if they receive extra income. For example, an MPC of 0.6 would mean that for an extra dollar of income, people would spend 60 cents. A paper published by economists at the Federal Reserve Bank of Chicago estimated that the MPC for recipients of Covid stimulus payments in April 2020 was 0.46.
Sprintson and Oughton do not include an MPC in their analysis. They assume that households spend all of the extra money they receive as a result of the ACP subsidy. Not even Keynesians believe that.
They also assume that households will receive an income boost of $360 each. To get that, they multiply the monthly subsidy of $30 by twelve, which makes sense at first glance. But they work under the assumption that because the average monthly broadband price is $61, every participating household will effectively receive a full $30 in extra income.
That is not how the program works, and not everybody pays $61 per month currently. The program’s eligibility is targeted at low-income households. Low-income households were already eligible for reduced-price service through internet-service providers, often for as low as $10 per month, before the ACP came into existence. They could also receive a subsidy of up to $9.25 per month through the FCC’s Lifeline program, which has covered broadband since 2016.
Internet-service providers have aggressively marketed ACP enrollment to low-income customers. For example, AT&T advertises $30 per month service as “free” with ACP enrollment. A hypothetical low-income household paying $10 per month that then enrolls in a $30 plan that is entirely covered by the ACP subsidy does not have $30 extra to spend. It has seen a $10 monthly expense disappear and therefore only has $10 extra to spend monthly compared to before the subsidy.
The paper also fails to note that a decent chunk of the subsidy is captured by internet-service providers, especially for low-income internet plans. This likely explains part of the reason why internet-service providers support the ACP and want Congress to give it more funding.
Sprintson and Oughton could argue this is still good for people receiving the subsidy. The $30 “free” plans AT&T advertises for low-income households now provide faster service than the $10 plans they advertised before the ACP. But by this logic, why stop at $30 per month? If the government provided everybody with $200 per month in subsidies, everybody could afford premium internet service.
And Sprintson and Oughton’s literature review is about expanding internet access, not increasing the speed of internet to someone who already has access. There is no way the economic boost from a higher internet speed is anywhere near as large as the boost from having internet for the first time.
The paper pats itself on the back for being “the first macroeconomic assessment of the broadband infrastructure programs within the Bipartisan Infrastructure Law.” One hopes future assessments will be more rigorous. In the meantime, policy-makers should disregard this bogus study and roll back pandemic-era welfare expansion.