Stock

Marc Joffe

Like mom and apple pie, the public library seems so intrinsically good that it should be beyond criticism. But like any institution that consumes millions of tax dollars, public libraries should not be free from scrutiny. And the facts are that neighborhood libraries have largely outlived their usefulness and no longer provide value for the public money spent on them.

Consider the situation in Northern California, for example. In this fiscal year, four Bay Area counties (Alameda, Contra Costa, San Mateo, and Santa Clara) are collectively spending $270 million to operate their library systems, with some cities chipping in extra to finance extended operating hours. Contra Costa County is spending $20 million of state and county funds to build a new library in Bay Point, and El Cerrito voters may see a sales tax measure on the November ballot, part of which will go to building a new library as part of a transit‐​oriented development near a Bay Area Rapid Transit station.

The public library’s historical functions of lending physical books and enabling patrons to view reference materials are being made obsolete by digital technology. An increasing proportion of adults are consuming e‑books and audiobooks in addition to or instead of printed books, with younger adults more likely to use these alternative formats.

In response, libraries have tried to reposition themselves as “third places”: alternatives to homes and offices where people can relax, learn, and socialize. But the private sector offers numerous third places of its own, with coffeehouses being the most common.

In Walnut Creek, California, the public library has responded by adding its own coffee shop, but just a few minutes away, residents and visitors can relax and enjoy free Wi‐​Fi at the Capital One Café at no cost to taxpayers and without being required to buy a cup of joe.

While no third place used by the public can be guaranteed to be safe and clean, private operators have a stronger incentive to provide an attractive environment because they otherwise risk going out of business.

They also face fewer legal restraints in enforcing public decorum. A 1991 federal court decision prohibited a New Jersey public library from “barring patrons who are not reading, studying or using library materials, who harass or annoy others through noisy activities or by staring, or whose ‘bodily hygiene is so offensive’ that it is a nuisance to others.”

As the Bay Area News Group reported in February, the Antioch library had to be temporarily closed after multiple incidents “including a couple having sex openly in the bathroom, a wanted criminal using a library computer who was later removed by Antioch police, a racist letter left on the service desk and intoxicated library patrons acting aggressively.” The Contra Costa Public Library, which operates the Antioch facility, reopened it four days later after negotiating an emergency contract for private armed security and arranging for a patrol car to monitor the exterior.

Library advocates argue that their public terminals offer essential internet access to those in need. But some patrons use free internet access at the local library to view pornographic content, sometimes to the distress of other terminal users, including children. And low‐​income individuals are eligible for the Federal Communications Commission’s Lifeline program, which provides a free smartphone with internet access.

Like local post offices, neighborhood libraries once served an important community function but are now becoming increasingly irrelevant. And, as with post offices, libraries continue to receive funding because they enjoy support from a relatively small but vocal segment of the population, while the rest of us are usually too reluctant to question their utility.

A version of this article originally appeared in the Contra Costa Herald.

Romina Boccia and Dominik Lett

Congress recently approved a $95 billion foreign aid package to send aid to Ukraine, Israel, and the Indo‐​Pacific region. Including this new “one‐​time” emergency spending in fiscal projections distorts the Congressional Budget Office (CBO) baseline. With interest rates persistently high and long‐​term fiscal forecasts dire, Congress needs to budget honestly and transparently now more than ever. Congress should revise the baseline to exclude emergencies.

Emergencies Distort the Baseline

Each year, the independent, nonpartisan CBO produces a budget baseline used as the measuring stick to evaluate different fiscal policies’ short‐ and long‐​term effects. CBO’s reports and cost estimates regularly influence the legislative process, confronting legislators with the tradeoffs involved in budgetary decisions.

One flaw of the baseline—comprised of projections of future spending and revenues—is counting temporary emergency funds alongside regular discretionary appropriations. When estimating future discretionary spending, CBO assumes both regular and emergency appropriations grow side‐​by‐​side each year with inflation. Due to this practice, continuing discretionary spending at the previous year’s level would counterintuitively appear as a deficit reduction relative to the budget baseline.

The picture is further muddled by the differing treatments of discretionary spending, mandatory spending, and revenues. When CBO scores individual bills, discretionary costs are estimated in relation to current law. Meanwhile, the costs of mandatory spending and revenue changes are estimated using the baseline.

Regardless, CBO’s budget baseline directly impacts congressional scorekeeping. When Congress produces its budget resolution—the overall framework within which Congress considers legislative bills that affect spending and revenues—it relies heavily on CBO’s baseline. Artificially boosting projected spending by including emergencies alongside regular appropriations has meaningful downstream effects on congressional fiscal decisionmaking.

Treating temporary emergency provisions as permanent and growing expenditures creates an insidious ratcheting effect, baking in a bias toward higher spending. Spending‐​prone legislators can point to the artificially elevated baseline to justify further spending increases. Moreover, it undermines the original purpose of emergency spending—a budgetary safety valve for necessary, sudden, urgent, unforeseen, and nonpermanent situations.

The latest foreign aid package illustrates just how broken the emergency spending process is. Not only will it add $95 billion to the deficit and bypass spending limits, but it will also add $40 billion in interest costs over the next 10 years. In total, the fiscal cost of Ukraine aid could top $240 billion. Adding insult to injury, Congress mandates that CBO distort the baseline by treating this funding as a permanent, growing expenditure in the baseline despite the vast majority of new supplemental budget authority only being provided for fiscal year 2024.

Congressional desensitization to emergency deficit spending is a major reason why emergency spending over the last three decades ballooned to $14 trillion (including interest) and why the public debt is sitting at a record $27 trillion. And that’s before considering the pernicious long‐​term effects of building in an expectation for higher spending based on temporary emergency provisions billed as one‐​time costs.

Sensible Fiscal Reforms

The Stop the Baseline Bloat Act (H.R.8068), recently introduced by Reps. Glenn Grothman (R‑WI) and Ed Case (D‑HI), strikes emergency spending from the budget baseline. This change would make CBO’s baseline less biased toward higher spending. Congressional practice of relying on emergency appropriations to fund ongoing issues poses a significant challenge to fiscal responsibility and undermines the integrity of the budget process. Emergency appropriations are intended for unforeseen and urgent circumstances, not as a stopgap for chronic budgetary shortfalls or to circumvent the regular appropriations process. Additionally, if Congress is intent on relying on emergency spending on a regular basis, CBO should provide an alternate estimate of budget projections that includes emergency spending. Read our “Six Reforms to Enhance Transparency and Fiscal Accountability through the Congressional Budget Office” blog for other CBO reforms that would enhance transparency and accountability in the budget process.

The United States cannot afford to continue recklessly high levels of deficit spending given our current fiscal trajectory. A baseline change that accurately reflects the nature of emergency spending is a critical step in this direction. Curbing this and other abuses of emergency spending should be a bipartisan priority.

Travis Fisher and Alex Nowrasteh

In a recent piece titled “Climate Change and Globalization,” Charles Kenny argues that globalization can help reduce the negative effects of climate change and even help reduce CO2 emissions. The essay was one of twelve in the Cato series Defending Globalization: Society and Culture. Although we agree with the call for liberalized global trade, we disagree with many of the essay’s recommendations regarding climate policy and multilateral agreements.

First, let’s recognize some common ground. We greatly respect Dr. Kenny and value the open exchange of ideas. One of the authors of this blog post (Alex) read Kenny’s book The Plague Cycle and thought it was masterful. We also agree that free trade means: higher incomes, easier adaptation to a changing climate, and reduced costs of CO2‐​mitigating technologies (through the removal of tariffs that inflate the cost). All of these changes lessen the negative impacts of climate change.

However, Dr. Kenny goes further than free‐​market economists would suggest when it comes to mitigating the CO2 externality. He endorses the goal of “net zero” CO2 emissions and argues for a global cap‐​and‐​trade system to achieve it. Reading Dr. Kenny’s piece, we came away with the impression that CO2 is an unalloyed bad; in contrast, we see it as the result of the productive use of hydrocarbon energy, which represents over 80 percent of global primary energy use. We must avoid calls to swiftly move away from hydrocarbons simply because they (likely) introduce a negative externality—especially when the cost of such an energy transition isn’t well considered.

Source.

“Net Zero” CO2 Is Not a Desirable Goal

Under a Pigouvian approach to internalizing a negative externality, economists advocate for “pricing” (read: taxing) the externality rather than eliminating it at all costs. In other words, the textbook policy prescription—assuming we know the social cost of CO2 (SCC)—is to tax CO2 emissions at the SCC and then let emission levels fall where they may. Capping the quantity of CO2 emissions would only be an economically efficient policy in the unlikely scenario that the level of the cap happens to equal the quantity of emissions produced under an efficiently‐​priced SCC.

Based on real‐​world observations, the efficient level of CO2 emissions under a Pigouvian approach is not zero. Given the myriad productive uses of hydrocarbon‐​based energy and the demonstrated preference for the large and growing global consumption of it, the economically efficient level of CO2 emissions could be very high. Unfortunately, we may never know because the SCC—although indispensable in concept—is nearly impossible to nail down with any degree of scientific certainty. We do know, however, that low and moderate CO2 taxes like the one implemented in British Columbia have not significantly reduced CO2 emissions.

Kenny writes:

But efforts to put legally binding caps on carbon dioxide emissions or to create a global market for the right to emit carbon dioxide (a “cap and trade” regime) have floundered. That’s unfortunate: markets really are likely the most efficient way to reduce global emissions, and the failure to develop one has surely slowed progress toward a low‐​carbon global economy.

The level of CO2 tax implied by a “net zero” cap‐​and‐​trade system is inefficiently high under a Pigouvian framework. Specifically, a net zero policy is only economically efficient if the SCC matches the cost of CO2 removal. Presently, CO2 removal technologies cost multiple times more than even the EPA’s inflated estimate of the SCC. In other words, reasonable estimates of the CO2 externality do not justify the “net zero” emissions future envisioned by Dr. Kenny.

Industrial Policies and Global Treaties Are No Silver Bullet

Kenny rightly concludes that the economic cost of ending global trade, migration, tourism, business travel, and reducing “food miles” are not worth it because they would negligibly reduce emissions at enormous economic cost. The same analysis should be applied to subsidies for green energy projects. A good place to start would be to estimate the cost‐​effectiveness of CO2 reductions from the Inflation Reduction Act (IRA). How much will the IRA reduce CO2 emissions after spending $3 trillion by 2050, and is that reduction worth it?

He writes:

The US experience with the Inflation Reduction Act provides evidence of another key role that globalization is playing in the transition to a zero‐​carbon economy: providing the workforce. The act may create more than half a million jobs, concentrated in sectors that are already facing severe labor shortages, including agriculture, manufacturing, and construction.

Here, Kenny commits the “broken window fallacy.” Economic activity created via subsidies does not indicate an increase in economic activity overall. Likewise, jobs created by subsidies do not indicate an increase in jobs overall. There is a net cost to society when subsidies and other industrial policies distort markets. Whether the alleged environmental benefits of the IRA are worth the $3 trillion price tag is a question for policymakers to wrestle with, but framing the IRA as a net job creator runs afoul of economic reasoning.

Dr. Kenny’s article also pushes for selective liberalization of climate‐​friendly trade and migration to reduce GHG emissions:

Given that climate change is a global issue, this suggests there could be an important role for global agreements. In the case of the Montreal Protocol protecting the ozone layer, that involved legally binding restrictions on chlorofluorocarbon use as well as payments from richer to poorer countries to help them adopt new technologies. A recent extension of the protocol involved limits on the use of hydrofluorocarbons, another refrigerant that is a powerful greenhouse gas.

Pros and cons of decarbonization aside, global agreements like the Montreal Protocol are problematic because they lump together free trade and environmental protection elements with policies aimed at the global redistribution of wealth from rich to poor countries. Further, as a matter of political process, before we give multilateral climate agreements like the Paris Agreement the weight of a binding international treaty, they require Senate ratification (and the US Senate has not ratified the Paris Agreement).

Dr. Kenny’s proposals go further than internalizing the global CO2 externality and get into the realm of industrial policy, such as green skills partnerships and multilateral development banks (MDBs). He writes:

If we could guarantee low tariffs on green goods through a World Trade Organization agreement, agree to green skills partnerships that would help workers move to where they are needed to construct renewable energy infrastructure, and use multilateral development banks to help flood the zone with financing for zero‐​carbon investments, global links would help get us to global sustainability even faster.

We heartily agree that freeing the movement of capital, goods, and workers to move internationally to where they are most highly valued is important for myriad economic, ethical, and environmental reasons. But artificially increasing government investment in green technology using MDBs capitalized by taxpayers is particularly expensive (and of dubious value if the IRA and other laws are any guide). This approach simply ignores the problems with industrial policy that we’ve cataloged over the years.

Conclusion

Climate policy is more difficult and complicated than other areas of public policy, but the principles remain the same: individual liberty, limited government, free markets, and peace. Dr. Kenny’s article breaks with libertarian principles (and textbook insights from Pigouvian economics) by endorsing the goal of “net zero” CO2 emissions and embracing a set of global interventionist policies to achieve it.

Adam N. Michel

Americans in ten states currently pay top tax rates above 49 percent, which is likely above the revenue‐​maximizing tax rate. Under President Biden’s proposed tax increases, almost three‐​quarters of the states would face top combined federal, state, and local tax rates above 49 percent. 

Tax rates affect people’s decisions about how much to work and invest. The effects are often greatest for the highest‐​income individuals because they are able to change their behavior the most.

Under the progressive tax rate system, additional income earned over certain income thresholds is subject to higher tax rates, ranging from 10 percent to 37 percent. The highest‐​paid Americans are charged a federal marginal income tax rate of 37 percent on each dollar earned above $731,200 (married) in 2024. The average top rate increases to 46 percent after factoring in state, local, and Medicare taxes.

When the government takes close to half of each additional dollar earned, it changes an individual’s decisions about how much to work and what type of work to do. There are at least four ways high tax rates decrease economic activity.

Hours worked. High taxes reduce the return to additional work, deterring people from putting in the long hours often associated with high‐​paying jobs. A doctor may choose to see fewer patients in a day, or a lawyer might bill fewer hours.
Labor force participation. Retirement‐​age workers and second earners (spouses) can opt not to work at all if their after‐​tax income doesn’t justify the effort. High labor taxes can also incentivize young people to delay labor force participation.
Human capital. The incentive to invest in additional years of education and training decreases when the resulting reward (high pay) is reduced by high taxes. A young doctor may not choose to spend additional years developing a technical specialty in surgery or research if he only gets to keep 40 or 50 cents of every additional dollar he will earn.
Entrepreneurship. High taxes diminish the returns to starting or expanding businesses. Business start‐​ups are risky (most fail), and the high earnings of founders and executives compensate for long hours, specialized skills, and the risk of failure. High wage taxes reduce entrepreneurship and business growth.

A fifth effect of high tax rates is avoidance. Individuals can move to regions with lower taxes, maximize credits and deductions, and shift income toward types of activities that are taxed at lower rates.

The academic literature summarizes these effects—often just the first two direct labor force effects and avoidance—by estimating an “elasticity of taxable income.”[1] The bigger the elasticity of taxable income, the more economic damage (deadweight loss) is caused by the tax.

Using a simple but flawed formula, academics translate estimated elasticities into a revenue‐​maximizing tax rate.[2] This is the top of the Laffer Curve where the economic damage from the tax increase is so large the higher tax rate raises no additional revenue. 

Estimating a revenue‐​maximizing rate can lead researchers to incorrectly imply that the revenue‐​maximizing tax rate is economically or socially “optimal.” Even below the revenue‐​maximizing point, tax increases explicitly trade off fewer work hours, less entrepreneurship, and a smaller economy for exponentially decreasing revenue gains.

The Revenue Maximizing Rate

The academic literature includes a range of estimates of the elasticity of taxable income. A widely cited 2011 article by Peter Diamond and Emmanuel Saez claimed that the mid‐​range elasticity estimate is 0.25, implying a revenue‐​maximizing top tax rate of 73 percent. Their relatively low elasticity (and high top tax rate) has been widely disputed, especially as applied to the highest‐​income earners.

A recent paper by Joshua Rauh and Ryan Shyu in the American Economic Journal: Economic Policy finds an elasticity of 2.5 to 3.2 for California taxpayers facing some of the highest marginal tax rates in the United States. Alan Reynolds’s 2019 review for the Cato Journal concluded that the elasticity for high‐​income taxpayers is at least 0.8 and likely higher. The average elasticity reported in eleven studies surveyed by Aparna Mathur, Sita Slavov, and Michael Strain is about 0.7.[3] In 2004, Saez found a similar 0.7 mid‐​point estimate for the top 1 percent.

Using the same method as Diamond and Saez (and accepting its deficiencies), a 0.7 elasticity of taxable income implies an upper bound revenue‐​maximizing rate of 49 percent. This is the same result reported in a 2020 Economic Journal article using a different model that incorporates human capital accumulation. Using other plausible parameters, the top rate could be as low as 25 percent and is most likely between 35 percent and 40 percent.[4]

Under current law, the top marginal federal income tax rate is almost 41 percent after accounting for Medicare taxes. In 2024, state income taxes range from 2.5 percent (Arizona and North Dakota) to 14.4 percent (California).[5] Nine states have no individual income tax. The Tax Foundation reports local income taxes in 2023 for the largest local taxing jurisdiction, which range from 0.085 percent (Des Moines, Iowa) to 4 percent (Portland, Oregon) across 13 states.

Figure 1 shows that across the 50 states and DC, combined federal‐​state‐​local top marginal income tax rates range from 41 percent in states with no income tax to more than 55 percent in California and New York City. Under current law, combined top income tax rates exceed 49 percent in 10 states.

Figure 2 shows that under President Biden’s proposed 39.6 percent top income tax rate and 2.1 percent Medicare surtax, combined income tax rates would exceed 49 percent in 36 states and DC. California’s highest rate in the nation would be 59 percent. Tax rates paid in California, New York, New Jersey, and Oregon would be the highest in the developed world—higher than Japan and Denmark’s 56 percent top rates and higher than every other country in the OECD.

While not explicitly called for in the president’s budget, many prominent Democrats have also proposed removing the Social Security payroll tax cap (wages subject to the current 12.4 percent tax are capped at $168,600 in 2024). Uncapping the Social Security tax would be a large marginal tax rate increase on high‐​income earners. Combined with Biden’s other tax increases, an uncapped Social Security tax would subject high‐​income taxpayers in every state to tax rates above 49 percent, with top income tax rates ranging from 57 percent to 71 percent.

Conclusion

Taxes on wages make individuals worse off by directly reducing their take‐​home pay and disincentivizing additional market work. High tax rates on the most productive Americans also have negative economic consequences for the millions of Americans who benefit from the goods, services, and jobs supported by high‐​paying jobs. These occupations often demand long hours, years of additional schooling, continuing education, and risk‐​taking, which are directly disincentivized by high marginal tax rates.

When considering revenue‐​maximizing tax rates, it’s worth ending with Milton Freidman’s observation that “if a tax cut increases government revenues, you haven’t cut taxes enough.” In other words, the optimal tax rate is far below the top of the Laffer Curve. Policymakers should keep tax rates as low as possible for all Americans, regardless of income level.

[1] If the elasticity of taxable income is 0.7, then a 10 percent change in the tax rate causes reported income to fall by 7 percent.

[2] Revenue maximizing tax rate, t*= 1/(1+p x e), where p is a Pareto parameter describing the income distribution and e is the elasticity of taxable income. Diamond and Saez assume p is 1.5, which likely biases up their estimate of t*.

[3] Average calculated by Alan Reynolds.

[4] 25 percent rate assumes an elasticity of 2.5 and a Pareto parameter of 1.2.

[5] California rate includes a 1.1 percent state payroll tax.

Paul Matzko

After years of legislative false starts, a TikTok ban bill has passed Congress. Notionally, the bill uses the threat of a ban to force a sale of the Chinese‐​owned app—thus allowing members of Congress to claim that it is not a ban—although the expert consensus is that China would rather stop US operations altogether than allow a sale to an immediate competitor.

The legislation has been remarkably rushed; the bill was introduced in March and passed in April. During that time, no general and open hearings on the bill have been held and no unambiguous evidence of Chinese tampering with US user data has been shared with the public (despite calls from senators to do so).

As Jennifer Huddleston and I have argued previously, Congress has opted for the most extreme regulatory option on the table instead of adopting intermediate measures that could have addressed concerns about data surveillance and algorithmic manipulation.

But that doesn’t mean TikTok is doomed. The company has already signaled that it will challenge the law in court on First Amendment grounds. Given that a TikTok ban would impair the free speech of 170 million Americans, they could have a plausible case. Previously, a federal district judge blocked Montana’s ban of TikTok both on free speech grounds and because proponents offered little evidence of a national security risk beyond a “pervasive undertone of anti‐​Chinese sentiment.” The same is true of the federal bill, perhaps even more so.

And rushing the law through Congress might have weakened the legal case for a ban by not waiting for a review by the Committee on Foreign Investment in the United States (CFIUS). The CFIUS has blocked foreign ownership of apps in the past on national security grounds, most notably with the acquisition of the Grindr dating app in 2019. A CFIUS recommendation for TikTok’s divestment from its Chinese owners would have given the law at least the imprimatur of due process. (Although the constitutionality of CFIUS’s recently expanded authority has yet to be upheld in court. The only time a lawsuit has ever been filed over a CFIUS review, the government was forced to settle on unfavorable terms.)

Instead, supporters have defended the law by appealing to precedents from Cold War‐​era broadcasting, when foreign ownership of radio and television stations was limited. However, these precedents are unlikely to apply to a TikTok ban because they are rooted in the so‐​called “scarcity doctrine,” which was a legal argument based on the natural limits of the electromagnetic spectrum. It was used to justify government licensure and regulation of broadcast speech, a sharp contrast to prohibitions on government interference in print media. But the internet doesn’t rely on scarce spectrum and so the courts have generally treated online speech like print speech rather than like broadcast speech, meaning a far higher degree of protection under the First Amendment.

The clearest legal precedent from the Cold War era actually militates against a TikTok ban. In the 1960s, when Congress restricted American postal access to Chinese communist literature, the US Supreme Court unanimously upheld the right of Americans to consume the speech of their choice, even when it was overt propaganda from the Chinese Communist Party.

American consumers have the constitutional right to choose the speech platform of their choice, even if it is a choice that unnerves the national security apparatus and congressional China hawks.

Andrew Gillen

Immediately after the Supreme Court overturned his last big student loan forgiveness plan, President Biden announced a new effort that would rely on a different law, an effort that is now nearing completion. The administration has released the draft regulations that would waive accrued interest, forgive debt for borrowers that have repaid for 20 years (25 years if there is debt for graduate school), automatically enroll students in programs that would result in loan forgiveness, and waive debt for students that attended low‐​quality institutions. The administration hopes to issue final regulations in a few months with a goal of forgiving debt this fall.

The administration estimates that 27.6 million borrowers (around 64 percent of all borrowers) would have some or all of their debt forgiven under the proposed regulations. Around 9 percent of all outstanding student loan debt would be forgiven, at a cost of $147.5 billion.

This plan suffers from several severe flaws.

The proposed regulations are unfair.

One of the big selling points for going to college is that college graduates earn more, which means that forgiving student loans tends to be regressive. The main exception to this would be dropouts who take on debt but do not get the expected earnings boost. The new regulations do not specifically target such students. In fact, the Penn Wharton Budget Model estimates that the typical household income of beneficiaries for the most regressive provision is $313,000. This clearly is not a well‐​targeted plan to help those in need. Indeed, the administration admits that “President Biden has vowed to use every tool available to cancel student debt for as many borrowers as possible, as quickly as possible.”

Even if forgiving loans was not regressive, it would still be unfair. Why should those who never went to college (some of whom avoided it because they didn’t want to take on debt), or those who paid off their loans, or parents who worked extra hours to help their children avoid debt now have to pay for other people’s student loans?

The proposed regulations are bad policy.

Two common mistakes of student loan programs are using them to subsidize education and using an inappropriate interest rate. The proposed regulations exacerbate both errors.

Regarding subsidies, whether and how much we subsidize college is a completely separate question from how student loans should operate. As Susan Dynarski and Daniel Kreisman wrote for the Brookings Institution,

The government should seek neither to make nor to lose money from student loans. Student loans correct a capital market failure … Federal student loans therefore solve a liquidity problem, not a pricing problem. Student loans are appropriate neither for raising revenue nor for subsidizing college.

Whether you want to subsidize colleges a lot, a little, or not at all, student loans aren’t the way to do it. As I summarized in a study from 2020,

Student loans are a terrible method of subsidizing higher education. Providing subsidies through loans … is poorly targeted, distributing the subsidy only to those who borrow. Yet around half of all community college graduates do not take out loans. Using loans to subsidize higher education also gives the biggest subsidies to those who borrow the most (relatively wealthy graduate students).

Unfortunately, student loans are heavily subsidized. The Congressional Budget Office estimates that the subsidy rate for the student loan portfolio is 21.6 percent, meaning that the government will lose 21.6 cents for every dollar that is lent. The proposed regulations would increase this subsidy even further.

Regarding interest rates, Nicholas Barr argues that the interest rate on student loans should be close to the interest rate the government pays to borrow. Yet the proposed regulations would waive interest for millions of borrowers including those in income‐​driven repayment plans. Yet Barr points out that “In a system with income‐​contingent repayments and forgiveness after 25 years, an interest subsidy achieves no single desirable objective.” By retroactively setting the interest rate to zero for many borrowers, the regulations would move us even further from a sound policy.

Biden does not have statutory authority to impose these regulations.

The administration’s last attempt at loan forgiveness was thrown out by the Supreme Court for violating the major questions doctrine, which requires clear congressional authorization for executive actions of major economic or political issues.

Student loan forgiveness on the scale proposed in these regulations clearly meets the major economic or political threshold. The overturned loan forgiveness plan was estimated to cost around $500 billion. The Committee for a Responsible Federal Budget estimates this one would cost $250–750 billion once you include the economic hardship provisions (which are still under development). Even the $147 billion cost that the administration estimates (which excludes the economic hardship provisions) would qualify as major since it dwarfs the scale of annual aid provided by main federal financial aid programs like Pell Grants ($27 billion) and student loans ($83 billion).

Nor is there clear statutory authorization. As Mark Kantrowitz lays out, the entire legal case for the regulations rests on a provision that allows the Secretary of Education to “enforce, pay, compromise, waive, or release any right, title, claim, lien, or demand, however acquired, including any equity or any right of redemption.” Yet there are four problems with the claim that this provides legal authorization for the new regulations.

First, the most plausible reading of this authority is that it applies to loan forgiveness plans that have been authorized by laws Congress has passed, such as the Public Sector Loan Forgiveness (PSLF) program. Congress passed a law setting up PSLF, with all sorts of eligibility criteria and requirements to waive debt for some borrowers. Why would they have done that if the secretary could have just waived all that debt under existing law?

Second, the waiver authority is granted for a different loan program. In particular, it applies to the Federal Family Education Loan Program (FFELP) which stopped making loans in 2010, not the Direct Loan (DL) program under which all new and most old debt falls. The administration claims that a provision stipulating that DL loans “have the same terms, conditions, and benefits” as FFELP loans means the waiver authority applies to DL too. But as Kantrowitz notes, “a waiver of a ‘right, title, claim, lien, or demand’ is not a term, condition or benefit of a loan.”

Third, the secretary’s waiver authority is only authorized on a case‐​by‐​case basis. For example, PSLF requires students to document their eligibility and income, which are then evaluated separately for each borrower. The proposed regulations ignore the requirement that waivers be granted on a case‐​by‐​case basis and instead try to waive part or all of the debt for almost two‐​thirds of borrowers en masse.

Fourth, the regulations ignore the requirement that the federal government try to collect the debts it is owed. In particular, the Federal Claims Collection Standards lay out various conditions under which debts to the federal government can be waived. How do the new regulations deal with these requirements? By simply stating they can be ignored. Is everyone now allowed to just ignore regulations they don’t want to comply with, or just the Biden administration?

In sum, the proposed regulations are unfair, badly designed, and lack legal authorization. They would further distort an already badly designed student loan system. Unfortunately, the courts are the only way to stop the regulations from taking effect. Fortunately, the previous failed student loan forgiveness plan has provided a roadmap to defeat this one in court too—namely, Missouri (which has standing to sue due to quasi‐​public student loan servicing entity) sues and the courts find the plan violates the major questions doctrine (which requires clear congressional authorization for executive actions of major economic or political issues).

David Inserra

In a ruling issued on April 17, the European Data Protection Board (EDPB) issued a decision that could unravel the very viability of social media company business models. The technical, inside‐​baseball nature of the case involving advertising and data privacy means it has flown under the radar but nonetheless presents a dangerous precedent that uniquely targets American tech companies. 

So first, what is going on in this case? In the EU, the sprawling General Data Protection Regulation (GDPR) places many demands on companies in how they manage the data of users, including requiring users to consent to the use of their data. As a result, Meta decided it would offer users two ways to use its platforms:

Users have the free, standard method, in which their user data is collected to create personalized and tailored ads.

In exchange for a small monthly fee, in essence a subscription, users can refuse to have their data collected to personalize ads.

The new EDPB decision, however, considered and rejected this fee‐​based alternative model, demanding that “large platforms” offer an option in which platforms provide their services for free but do not use customer data to target ads.

This goes to the heart of how social media and many other online services are funded. Users and policymakers often forget that the reason many online platforms are free is because they run ads. When in a now‐​famous 2018 congressional hearing, Senator Orrin Hatch confusedly asked Mark Zuckerberg how Facebook could “sustain a business model in which users don’t pay for your service?” Zuckerberg simply replied, “Senator, we run ads.” This awkward interaction sparked a host of articles and memes mocking Congress’s failure to understand the way these companies operate.

via GIPHY

This decision by the EDPB deserves the same treatment because it either completely fails to understand how online businesses work or does understand but simply does not care about the harm it will inflict on US businesses. By targeting its decision only at “large platforms,” the EDPB shows that it has no problem allowing the use of personal data for targeted advertising when done by smaller companies, conveniently protecting European companies.

Let’s be clear about what this decision does. The decision states: 

The offering of (only) a paid alternative to the service which includes processing for behavioural advertising purposes should not be the default way forward for controllers. When developing the alternative to the version of the service with behavioural advertising, large online platforms should consider providing data subjects with an ‘equivalent alternative’ that does not entail the payment of a fee. If controllers choose to charge a fee for access to the ‘equivalent alternative’, controllers should consider also offering a further alternative, free of charge, without behavioural advertising, e.g. with a form of advertising involving the processing of less (or no) personal data.

The EDPB is demanding that beyond the normal ads‐​driven model or the subscription model, Meta and other large platforms must provide a service at no charge but also not have personalized ads. The problem is that the whole reason advertisers pay high premiums for ads is so they can focus their ads on the types of users that will buy their products, support their candidate, or otherwise engage with them.

If a company is selling a first‐​person shooter video game, its target demographic is generally young men who, through their social media behaviors, have shown that they like playing video games. Similarly, if a company is selling professional women’s attire, its audience is working‐​aged women who have expressed interest in buying professional clothing online.

But if the company selling the video game did not have access to the data on who likes video games, that means it must advertise as much to women looking to buy a pantsuit as it does video gamers. Conversely, the women’s clothing company will be advertising its dresses and skirts to work from home video gamers. 

If that sounds like a bad way to advertise, both for the advertiser and the consumer, you would be right. But that’s exactly what the new EDPB decision would require.

This means that advertisers will not find ads to be worth as much. Large platforms will need to drastically lower the costs of ads to keep selling them. But if they change nothing else, that means platforms will make a lot less money and may even lose money. And even if a platform thought it could take the hit of losing money in Europe, then every other country in the world could demand that its users get the same treatment. That isn’t sustainable. 

So realistically, what are the options here? Well, if large platforms can’t make as much money per advertisement, then they might need to increase the number of ads users see. This would be frustrating for EU users as they would have to wade through countless, untargeted ads in their social media feeds. Not a great option and one that regulators might arbitrarily complain is not “fair” to EU users.

Or platforms will simply cease operating in the EU. Together with ever‐​increasing regulations under the new Digital Markets Act and Digital Services Act, this decision must raise the question of whether or not the EU market is worth it. While EU regulators may think their market is too important and too valuable to leave, this decision may be the straw that breaks the camel’s back.

Social media users across the EU may soon find out the hard way that requiring large tech companies to provide their services for free will not work the way the EDPB hopes it will.

Michael Chapman

Uri Berliner, a 25‐​year‐​long journalist at National Public Radio (NPR), recently resigned from his job after being suspended because he wrote about the entrenched left‐​wing/​woke mindset at NPR, calling it a “progressive silo.” Berliner’s tell‐​all confirms what media critics have long known, and reveals why taxpayer funding of the news media is not a good idea. 

It would not matter whether NPR was liberally biased or conservatively biased. The bottom line is that if politicians (bureaucrats) control the funding of the news, then the news likely will be politicized.

NPR, a non‐​profit media organization, was founded in 1970 by the Corporation for Public Broadcasting (CPB), which had been established by Congress in 1967 under the Public Broadcasting Act. NPR receives funding from private donors, corporate sponsors, foundations, and federal, state, and local government. The federal funding comes from the CPB, which provides grants to local radio stations that, in turn, pay to use content created by NPR, such as the news programs All Things Considered and Morning Edition.

According to NPR, “federal funding is essential to public radio’s service to the American public, and the programming fees from the local stations “comprise a significant portion of NPR’s largest source of revenue.” (NPR’s emphasis.) “The loss of federal funding would undermine the stations’ ability to pay NPR for programming, thereby weakening the institution,” the news outlets’s website insists.

The loss of federal revenue would result in “less journalism,” added NPR. Yet Berliner’s revelations suggest our tax dollars are sometimes paying for propaganda and not straightforward journalism.

At NPR, “the rise of advocacy took off with Donald Trump,” said Berliner. “[W]hat began as tough, straightforward coverage of a belligerent, truth‐​impaired president veered toward efforts to damage or topple Trump’s presidency.”

The outlet also “turned a blind eye” to the Hunter Biden laptop story. A colleague “said it was good we weren’t following the laptop story because it could help Trump,” said Berliner. He added that politics also “intruded into NPR’s Covid coverage, most notably in reporting on the origin of the pandemic. One of the most dismal aspects of Covid journalism is how quickly it defaulted to ideological story lines.”

“In DC, where NPR is headquartered and many of us live, I found 87 registered Democrats working in editorial positions and zero Republicans. None,” Berliner writes. “[P]olitics were blotting out the curiosity and independence that ought to have been driving our work.”

In addition to the political climate, senior management pushed racial and gender identity policies. For instance, NPR journalists were required to ask every person they interviewed “their race, gender, and ethnicity (among other questions), and had to enter it in a centralized tracking system,” said Berliner. “We were given unconscious bias training sessions. A growing DEI staff offered regular meetings imploring us to ‘start talking about race.’”

Despite those policies, “NPR’s news audience in recent years has become less diverse, not more so,” said Berliner. “[N]ow, the audience is cramped into a smaller, progressive silo. … Our news audience doesn’t come close to reflecting America. It’s overwhelmingly white and progressive, and clustered around coastal cities and college towns.”

If NPR were private, receiving no taxpayer funds, like The Nation or NBC News, its coverage and “less diverse” audience would likely raise little concern. NPR could be as woke as it wants or as conservative as it wants. The bottom line is that there is no reason why taxpayers should be forced to fund news organizations.

In the Cato Handbook for Policymakers (9th Edition), scholars at the Cato Institute write, “In a society that constitutionally limits the powers of government and maximizes individual liberty, there is no justification for the forcible transfer of money from taxpayers to artists, scholars, and broadcasters. … Moreover, the power to subsidize art, scholarship, and broadcasting cannot be found within the powers enumerated and delegated to the federal government under the Constitution.”

The same chapter says “Congress should eliminate the National Endowment for the Arts; eliminate the National Endowment for the Humanities; and defund the Corporation for Public Broadcasting.” (CPB’s FY2024 operation budget is $535 million.)

For radio and TV, “the selection process is inherently political,” reads the Cato Handbook. “Why are taxpayers in a free society compelled to support news coverage, particularly when it is inclined in a statist direction?”

As Uri Berliner wrote, “[W]hat’s notable is the extent to which people at every level of NPR have comfortably coalesced around the progressive worldview. And this, I believe, is the most damaging development at NPR: the absence of viewpoint diversity.”

If individuals want to support a “progressive” NPR through charitable contributions, that’s fine. But don’t force taxpayers to help foot the bill.

Paul Best

North Carolina officials announced a long sought‐​after victory in early 2022: VinFast, a Vietnamese electric vehicle startup, will invest $4 billion to build a 2,000-acre automotive factory in Chatham County, creating 7,500 jobs for the community as it helps to restore the state to its former industrial glory.

But skepticism grew about the plan as more details were released. The state intends to use eminent domain to purchase and then demolish over two dozen homes, five local businesses, and the century‐​old Merry Oaks Baptist Church. Officials announced a bevy of grants and tax breaks for VinFast in the form of a $1.2 billion incentive package, courtesy of taxpayers. VinFast went public by merging with a special purpose acquisition company last August, opening at $10 a share and quickly running up to $93 before falling hard back to earth. Shares closed at about $2.40 on April 22.

And while the electric vehicle market was still buzzing in early 2022, optimism for the industry’s outlook had dipped sharply by the time VinFast actually broke ground in summer 2023. It doesn’t help that VinFast has burned through $5.8 billion over the last three years as it vies to gain traction in the cutthroat industry.

In the first edition of Free Society, the Cato Institute’s new quarterly magazine, we dove into VinFast’s plans for the factory, why North Carolina officials handed out a $1.2 billion incentive package, and the concerns of many Chatham County residents who feel like they’ve been forced to watch all these developments from the sidelines.

One of those residents, 76‐​year‐​old Lena Stone, built a house in the 1970s that she still lives in today, surrounded by kids and grandkids just a short drive away. The quiet life she’d spent decades building was upended in August 2022, when she was informed at a community meeting that her home was in the way of infrastructure improvements and other construction necessary for VinFast’s factory.

“It’s more like, ‘This is how it’s going to be,’ versus giving us options, so it was kind of a shock,” said Karley Michelle, Stone’s granddaughter. “She’s really had a hard time with it because she’s 76, and she’s been there ever since she was in her 20s.”

Since publishing this story, local news outlets reported that construction has stalled as VinFast seeks permit approval for a new factory design. Two law firms also just filed a class action lawsuit against VinFast on behalf of shareholders who allege that the company overstated the strength of its business and made other misleading statements. A VinFast spokesperson said that the allegations in the lawsuit are “not accurate.”

Read the full story in the first edition of Free Society. You can also listen to this episode of Cato Daily Podcast on VinFast’s plans and the fallout in Chatham County.

David Inserra

In recent weeks, a simmering conflict over the censorial practices of the Brazilian courts erupted into open conflict. Reporting from various sources shows how government forces— most notably Supreme Court Justice and President of the Superior Electoral Court Alexandre de Moraes—investigated, secretly censored, and arrested Brazilians accused of spreading “fake news” and “anti‐​democratic” misinformation, often with little due process. Social media companies were silenced and punished if they resisted.

In response, X, formerly Twitter, declared that it would not comply with illegal and “draconian” demands before retreating under threat of punishment. While the government has defended itself by claiming it is protecting democracy, its increasingly authoritarian efforts to stamp out what it deems to be misinformation show how censorship is a threat, not an ally, of democracy.

The fuse was lit when journalist Michael Schellenberger released internal Twitter documents from the past several years that described a host of often secret demands made by Brazilian authorities. These included,

Demands by the Brazilian Congress in 2020 to release the private, direct messages of certain users, which Twitter refused to comply with for being against Brazilian law.

Police and prosecutors threatened to arrest a Twitter employee for failing to provide private Twitter user‐​data without a court order, something that other tech companies apparently provide even though it is forbidden by Brazilian law.

A court order demanding Twitter unmask the identity of anonymous Twitter accounts critical of a political figure who was currently under investigation for corruption and had millions of Brazilian reais (several hundred thousand US dollars) seized by authorities. 

A series of court orders in 2021 and 2022 that demanded Twitter reveal the identities of and demonetize accounts supportive of then‐​President Bolsonaro, including Bolsonaro himself, for content or hashtags that criticized the Superior Electoral Court and election procedures as part of a court‐​led investigation into misinformation. Twitter’s lawyers referred to these demands as “mass and indiscriminate disclosure of private users data” that are a “fishing expedition” and “a violation of privacy and other constitutional rights.” 

Other legal demands, often by Moraes himself, to suspend accounts, including those of sitting members of Congress for electoral misinformation. Twitter often pushed back, but in some cases had to comply to avoid significant fines if they did not comply within extremely short timeframes, even as short as one hour. Given that these are often secret orders, social media companies are effectively prohibited by the court from telling users why their content is being suppressed.

Other reporting by Glenn Greenwald noted that this wasn’t some fever dream of the political right in Brazil, and reporters at the New York Times questioned if the militant defense of democracy embraced by the Brazilian high courts was now posing a threat to democracy.

The Times cited an instance where a Whats App chat of eight businessmen in Brazil leaked, in which two suggested that they would prefer any outcome, even a coup, to Lula becoming president. In response, Moraes ordered the homes of all eight businessmen raided, their bank accounts frozen, various records subpoenaed, and certain social media accounts banned.

In another case, Moraes jailed without trial five individuals for their social media posts that he claimed attacked Brazilian institutions. Moraes also suspended the encrypted communication app, Telegram, for refusing to take down and provide information on prominent Bolsonaro activists for spreading misinformation.

And so, after years of growing censorship demands, Elon Musk said X would no longer abide by Moraes’s efforts to silence and expose dissent through often secret court orders and people have no timely or meaningful ability to appeal. Musk condemned the actions of Moraes and said that he would soon publish all the demands made by the judge.

Not keen on having his censorial demands ignored and exposed, Moraes struck back, expanding existing disinformation investigations into criminal ones against Musk himself for obstructing justice, being a member of a criminal organization, and incitement. Such crimes carry sentences of more than ten years. Perhaps due to this threat or concerns for the safety of X employees in Brazil, the company retreated and announced it would comply with the demands of the courts.

While the Brazilian courts already had substantial power in Brazil, many of these recent actions claimed even further powers to the unelected judiciary. Courts are often thought of as being neutral defenders of rights and due process, but Brazil’s courts—and especially Moraes—have claimed a far broader mandate, free from limits of a US First Amendment.

Attacks on the court, government institutions, or the “truth,” not with violence but with words, are taken as personal affronts to the honor of the judiciary that are then investigated, prosecuted, and judged by the courts themselves. Members of the judiciary take it upon themselves to “defend democracy,” as well as coordinate with executive branch officials to suppress the misinformation.

And this doesn’t even begin to cover the censorial actions coming from the Lula government or attempts by his allies to pass legislation that was inspired by the EU’s Digital Services Act (DSA). Such legislation would give significant power to the government to regulate tech companies and hold them liable for “fake news,” effectively unraveling Brazil’s Marco Civil, its core internet law that is comparable to Section 230 of the US Communications Act. When technology companies used their expressive rights to highlight the harms of the legislation, the government demanded the companies justify their opposition to the courts and hit Google and Telegram with threats of large hourly fines and blocking their platforms in Brazil unless they rescinded their statements opposing the bill.

To be clear, electoral violence such as that seen in the storming of Brazil’s capital buildings by supporters of Bolsonaro on January 8, 2023, like any form of violence, is to be condemned and relevant laws fully prosecuted in the courts. But the “crimes” described above by various news sources are mostly non‐​violent, non‐​threatening speech, and much of it took place before the violence of January 8.

A crusading and unaccountable court that silences, jails, and shuts down individuals, journalists, politicians, and companies for mere “misinformation,” often in secret or with little due process, is a clear threat to liberty and democracy.

Whether it is the current regimes in Turkey and Venezuela, the Russian Tsar, the USSR, and now Putin’s autocracy in Russia, Weimar and Nazi Germany, or countless other cases throughout history, the widespread suppression of speech under the guise of protecting democracy, society, or the government goes hand in hand with rising authoritarianism, government abuse, and human suffering.

While it’s unclear how this current dispute will end, let’s hope that dragging this censorship out into the sunlight will help restore free expression to Brazil.