Government legislators pursue spending increases because they believe that spending solves problems in society and benefits their constituents. However, they use methods of fiscal illusion to try to hide the costs of spending. One method is debt. By borrowing, legislators push the costs down the road for future legislators and taxpayers to deal with.

Spendthrift legislators can get away with borrowing more and taxing less in the short run, but in the long run taxes will rise to cover the higher interest costs on the accumulating debt. Relative to state gross domestic product, state and local taxes tend to be higher in states that have higher state and local debt, as shown in the chart below. The tax and debt data are from the Census Bureau for 2021.

The chart includes a fitted regression line showing the strong positive correlation between taxes and debt. Perhaps states that borrow more end up raising taxes to cover the rising interest costs. Or perhaps states that favor higher spending tend to pursue it with both taxes and debt. There are some outliers—Kentucky has high debt and average taxes, whereas Maine and Vermont have high taxes and average debt.

People living in the top-right states should be asking questions. Residents of Vermont, Maine, and Hawaii should be asking why their tax burdens are almost twice as high as the burdens in the lowest-tax states. And residents of Kentucky, West Virginia, New York, and Hawaii should be asking why their states have racked up debt burdens three or four times higher than the burdens in the lowest-debt states. And all of us should be asking why the federal government is vastly more irresponsible with debt than the worst-run states.

The federal “tax gap” is the amount of taxes owed but not paid, essentially the amount of cheating on federal taxes. All tax systems have tax gaps, and there are pros and cons of using tougher enforcement to reduce the gap. The size of the federal tax gap is not a major problem, as it has been stable over time relative to the size of the economy and is smaller than the average gap in Europe.

The IRS released a new estimate of the tax gap yesterday, which it billed as a “significant jump from previous estimates.” But when compared to the size of the economy, the new estimate is not a significant jump. Despite much political rhetoric to the contrary, tax cheating is not a growing problem for the economy. 

The gross tax gap in 2021 was $688 billion, according to the IRS. After late payments and enforcement actions, the net tax gap was $625 billion. Of the net total, $475 billion stemmed from individual income taxes, $37 billion from corporate income taxes, $112 billion from payroll taxes, and $1 billion from estate taxes.

The new report includes gross tax gap estimates for prior years. The dollar values of the tax gap have increased over time, but the gap is similar to previous estimates when compared to U.S. gross domestic product, as shown in the chart. I’ve used the GDP of the middle year for the multiyear gap estimates.

The flip side of the gross tax gap is the “voluntary compliance rate,” which is the tax paid on time divided by the estimated full amount owed. The IRS report shows that the voluntary compliance rate has hovered between 82 and 85 percent since 2001, and it was 84.9 percent in 2021.

No one likes tax cheating, especially when perpetrated by the rich and powerful. But there are civil liberties trade‐​offs when the government tries to close the gap by overly vigorous enforcement and heavy-handed regulations. The optimal tax gap is not zero because that would impose huge compliance costs on taxpayers and because the IRS makes frequent errors.

That said, there is a win‐​win approach to reducing the tax gap. Simplifying the tax code and cutting tax rates would reduce the ability and incentive for taxpayers to cheat. It would also cut compliance costs and boost civil liberties. Congress should put tax simplification back on the agenda.

I discuss IRS funding and the tax gap in recent congressional testimony here.

Federal spending and deficits are at dangerously high levels, and interest costs on government borrowing are soaring. If spending is not restrained, we may face an economic crisis. Congress should cut spending, and one good reform target is the huge and fast-growing Medicaid program.

Medicaid provides health coverage for lower-income individuals. It is administered by the states but the federal government provides about two-thirds of the funding. Federal Medicaid costs have more than doubled over the past decade from $265 billion in 2013 to $616 billion in 2023.

The federal government funds Medicaid with open-ended matching payments to the states. For administrative costs, the federal match averages 62 percent, and for medical costs the match is based on the Federal Medical Assistance Percentage. The FMAP is determined by each state’s per capita income—states with lower incomes receive higher matches and states with higher incomes receive lower matches. State FMAPs range from 50 to 83 percent.

The national average FMAP in coming years is expected to be 60 percent. That means for every $10 million that states expand their Medicaid programs, they pay $4 million and they send a bill to Washington for $6 million.

However, the federal match is even higher on Medicaid enrollees under the 2010 Affordable Care Act. The Act provides a 90 percent match to states that expand coverage to all adults under age 65 up to 138 percent of the poverty level. That creates a large incentive for states to join the ACA expansion and 40 states have done so.

The cost of the ACA expansion exceeded projections. A 2016 study found that enrollment in 24 early adopting states was double than originally projected. New York projected 76,000 ACA enrollees by 2015 but actual enrollment was 285,000. California projected 910,000 enrollees by 2016 but actual enrollment was 3.8 million.

Federal subsidies for the states create many distortions, which high federal matching rates exacerbate. A modest first step to reform Medicaid would be to reduce the ACA matching rate. The ACA expansion is now costing federal taxpayers about $120 billion a year.

Bad Incentives of Federal Funding

If a state with an FMAP of 60 percent expands its Medicaid by $10 million, it would cover $4 million and the federal government would cover $6 million. This creates an incentive for states to expand benefits beyond what they would provide if state taxpayers paid the full cost. Medicaid expansion allows state policymakers to spend on their constituents while only being accountable to their constituents for a fraction of the costs.

Also, as Michael Cannon noted, if a state spends, say, $4 million on an activity such as policing that does not receive matching federal aid, it buys $4 million of police protection. But if a state spends $4 million on Medicaid, it buys $10 million in health care spending. So federal aid distorts the legislative priorities of state policymakers, and public funds are misallocated.

Furthermore, Medicaid’s high federal match combined with the program’s size and complexity undermines state-level monitoring for waste and fraud. For every $10 million that state auditors may save from reducing waste or fraud in a state with a 60 percent match, the state government would gain just $4 million. As a result, states likely underinvest in program integrity activities such as examining billing claims and recipient eligibility.

Medicaid has long been on the Government Accountability Office’s high-risk list. The program’s improper payment rate was 16 percent in 2022, which amounted to $81 billion of erroneous or fraudulent payments. Some legislators think the rate is substantially higher.

State governments themselves abuse Medicaid with schemes to inflate matching dollars. To boost their federal match, they classify non-Medicaid spending as Medicaid spending, and they create accounting schemes such as overpaying health care facilities run by local governments and imposing taxes on health providers.

Medicaid’s financial structure creates irresponsibility all around. Budget expert James Capretta noted, “Medicaid’s current federal-state design also undermines political accountability. Neither the federal government nor the states are fully in charge. As a result, each side has tended to blame the other for the program’s shortcomings, and neither believes it has sufficient power to unilaterally impose effective reforms.”

Reduce ACA Matching Rate to FMAP

To generate federal budget savings and improve state incentives, Congress should cut the 90 percent ACA match to the state FMAP rates. The Congressional Budget Office estimates that such a reform would reduce federal deficits by $604 billion over 10 years.

A more thorough reform would be to convert federal Medicaid payments to per-capita block grants, while loosening regulations on the states to foster innovation. With block grants, the states would be encouraged to find more efficient benefit structures and to reduce fraud and waste. That was the successful approach of federal welfare reform in 1996. A further advantage of block grants is that they could help equalize current disparities between the states in federal Medicaid payments.

Michael Cannon has proposed Medicaid reforms in his new book Recovery, and further reform ideas are here and here.

The Employee Retention Tax Credit (ERTC) is a refundable payroll tax credit—equivalent to a cash payment of up to $26,000 per employee—for businesses and nonprofits with significant pandemic-related revenue declines or shuttered by government orders. The payments were intended to reward companies for keeping employees during the COVID-19 pandemic.

Since passage, the fiscal cost of the credit has ballooned to more than $230 billion, three times the original estimate. Over three-quarters of that cost is likely a windfall to business owners and tax preparers. 

The ERTC was created in the CARES Act and subsequently expanded as part of the 2021 appropriations, expanded again in the American Rescue Plan Act of 2021, and partially clawed back in the Infrastructure Investment and Jobs Act.

Since then, there has been in-depth reporting, a congressional hearing, and successive IRS guidance highlighting the many problems with the program. In September, the IRS announced it would stop processing new ERTC claims and increase audits following more than a year of repeated warnings to taxpayers against claiming inflated credits. The ERTC was the first entry in the IRS’s annual Dirty Dozen list of tax scams.

In a recent story following the IRS’s crackdown, the Wall Street Journal detailed the rise and fall of third-party ERTC processing firms that have engaged in extensive marketing campaigns, aggressive (possibly fraudulent) credit claiming, and exploitative contingency arrangements in which some third-party processors receive up to 25 percent of the credit.

It’s easy to find stories of businesses needing ERTC funds to stay open. These are the stories Congress reacted to when struggling businesses in their districts were waiting on delayed tax refunds.

However, the majority of the ERTC spending looks to have been a windfall to businesses that did not need the funds to retain employees through the pandemic. What doesn’t make the headlines are the thousands of businesses for which the payments are “icing on the cake,” as one accountant put it, “cash they can claim because they qualify,” as opposed to because they needed the cash to retain employees.

The Joint Committee on Taxation initially estimated that the ERTC would cost about $77 billion after the successive legislative changes. Using early filing statistics through the 2021 tax season, a U.S. Treasury economist estimated the credit cost $99 billion (a highly uncertain figure given processing delays). Of that figure, about half of the total credit (about $50 billion) was estimated to come from amended returns—whereby taxpayers claim the credit retroactively after they filed their original return.

As of September 2023, the IRS had paid out more than $230 billion in credits, the steep increase largely due to a flood of amended returns. The Congressional Budget Office pointed to the unexpected surge in credits as one factor for this year’s lower-than-expected tax revenues.

The large number of third-party ERTC processors submitting amended returns is economically damning. For the subsidy to have the intended effect of keeping employees on the payroll who would have otherwise been laid off, the taxpayer must know about the credit and change their behavior during the pandemic closures. There is no employment incentive for taxpayers who submit claims on amended returns upon advice from third-party consultants years after the pandemic ended.

If the credit did not change behavior, it is simply a windfall.

If we generously assume that all of the JCT’s initially estimated $77 billion went to keep employees on the payroll during the pandemic, we can conclude that the remaining $153 billion of ERTC spending was likely a windfall to business owners and third-party processors who saw an opportunity to exploit the program’s complexity and lenient rules.

Extrapolating from the other estimate’s $50 billion in claims not submitted on amended returns through the end of the 2021 filling season (and assuming subsequent payments were made to amended returns), as much as $181 billion or 78 percent of the $230 billion in ERTC spending to date has been a windfall to businesses. By one estimate, nearly half of any windfall to business owners is captured by individuals with incomes above $250,000.

Some may argue that the ERTC funds were, in some cases, used for productive ends, such as to pay out employee bonuses or expand businesses to meet post-pandemic demand. While this is invariably true, it is an argument for lower taxes for all businesses and not just for those who could exploit the arcana of a poorly designed and haphazardly administered tax credit.

Since it created the ERTC, Congress has added dozens of new tax credit schemes worth trillions of dollars to incentivize everything from computer chip manufacturing to hydrogen fuels. We cannot predict how each new credit will unravel, but Congress should expect similar types of grift, fraud, ballooning costs, and other unintended consequences.

The core lesson policymakers should learn from the mistakes of the ERTC is that the tax code and the IRS are not well-equipped to distribute targeted subsidies. As a matter of economic planning, industrial policy by tax credit is just as economically flawed as industrial policy by any other means. Tax credits for every social and economic ill is a recipe for fiscal ruin and economic malaise.

On April 10, 2023, Congress terminated the three-year-long COVID-19 national emergency—one of the most expensive emergency declarations ever at more than $7 trillion in spending, as reported by the Committee for a Responsible Federal Budget and the Heritage Foundation. Whenever such emergency declarations occur, the president is required by law to submit spending reports to Congress. With respect to the COVID-19 spending, President Biden apparently has not done so.

Based on press releases, letters to the administration, and congressional floor speeches, the administration has reportedly failed to follow the law. Failure to report emergency expenditures undermines transparency and accountability. This needs to change.

The National Emergencies Act (NEA), enacted in 1976, exists to limit the executive’s emergency powers. It requires the president to submit a spending summary for each ongoing national emergency to Congress every six months. The president must also send a final total ninety days after an emergency declaration ends. NEA expenditure reports are supposed to encompass any spending that is “directly attributable” to a national emergency declaration.

A major impediment to holding the executive accountable for ignoring emergency spending reporting requirements is that these reports are not easy to trace. NEA spending reports are not typically published online and the few that are lack detail as shown by this compilation of Treasury Department emergency expenditure reports. Greater transparency is essential in policing executive emergency spending abuses. Congress should revise NEA expenditure reporting rules to ensure they are as comprehensive as practicable and easily accessible to both legislators and the public.

Follow the Money

Emergency spending reporting requirements are intended to shed light on how taxpayer dollars are used during emergencies. Yet, the current system is convoluted. No centralized repository exists for NEA expenditure reports, and the executive can choose to whom in Congress to submit them, often leading to inaccessible records. Full-text versions of these reports are usually confined to internal congressional committee records.

For these reasons, the public usually only hears about major executive emergency abuses, such as President Trump’s border wall funding and President Biden’s student loan cancellations. The public is left in the dark about more insidious executive emergency spending abuse. This obscures the full scope of emergency spending.

On the heels of one of the most wasteful emergency spending splurges in U.S. history, legislators and the public should demand greater accountability from the executive by requiring more detailed and transparent emergency spending reporting.

Congress could amend the Congressionally Mandated Reports Act to make NEA reports available online. This would make the reports easy to locate for researchers and legislators, aiding Congress and the public in holding the executive accountable. Moreover, Congress could further define the format and content of NEA expenditure reports to add additional accounting details (something already done for some types of disaster assistance).

Senator Roger Marshall (R‑KS) and Representative Paul Gosar (R‑AZ) recently introduced the National Emergency Expenditure Reporting Transparency Act in the Senate and House, respectively. It would amend the Federal Funding Accountability and Transparency Act to include future NEA expenditure reports. Doing so would allow concerned legislators and members of the public to easily search for NEA-related funds.

The Bigger Issue

Emergency spending abuse goes hand in hand with expansive emergency powers. For decades, Congress has delegated away an array of emergency powers to the president. The progressive Brennan Center identifies more than 135 statutory powers that may become available when a president declares a national emergency. Thanks in part due to never-ending NEA declarations, Americans now live in a constant state of emergency.

As Cato’s Gene Healy explains, “emergency rule is a bipartisan temptation.” In 2018, President Trump invoked emergency powers to divert nearly $4 billion in military construction funds for a border wall against Congress’ wishes. As Trump himself noted, “I didn’t need to do this, but I’d rather do it much faster.”

President Biden similarly used emergency powers late last year in an attempt to cancel $430 billion in student loan debt. The Supreme Court shot down Biden’s scheme but didn’t stop Trump’s border funding.

Congress should rein in emergency spending powers instead of relying on the Supreme Court to police an overeager executive. The ARTICLE ONE Act would be a great start. The bill overhauls the NEA framework by having presidential emergency declarations automatically expire unless Congress affirmatively extends them.

As it stands, the emergency “state of exception” has become the norm. Congress should amend the NEA, the Congressionally Mandated Reports Act, and/​or the Federal Funding Accountability and Transparency Act to require easily accessible and detailed accounts of emergency expenditures. One key step toward limiting abuse of emergency spending powers is for legislators and the public to better understand the full size and scope of emergency spending.

Congress should hold the executive accountable. They can start by requiring stricter reporting about executive emergency expenditures.

The federal government’s debt is massive and growing rapidly. How massive? Federal debt held by the public of $26 trillion is about eight times larger than the combined debt of all state and local governments of $3.3 trillion.

Not only is state and local debt much smaller, but the states have more justification for accumulating debt than the federal government. That is because a larger share of state-local spending is for capital investment, which is partly financed by debt. State-local debt is matched by large holdings of return-producing assets such as highways.

By contrast, federal debt generally funds consumption—such as entitlement programs—not capital investment. And while federal borrowing during crises and recessions is reasonable, the debt should be paid back when the economy is growing.

The chart below compares federal debt as a percent of U.S. gross domestic product to the state-local debt of each state as a percent of state GDP. Federal debt of 98 percent in 2023 compares to the simple average of state-local debt of 13 percent. (The state ratios are from Census data for 2021, but total state-local debt has changed little since then).

Also consider that most state-local debt consists of “revenue bonds,” which will be paid back from revenues for facilities such as utilities, colleges, and hospitals. “General obligation” bonds backed by taxes account for a minority of state-local debt. Thus, not only is state‐​local debt much smaller than federal debt, but less than half of it will be paid back by taxes.

For these reasons, federal debt is a far larger threat to taxpayers, young people, and the economy than state-local debt. Whether for reasons of political culture or lack of constitutional or statutory rules, federal policymakers have been much more fiscally irresponsible than their state-local peers.

That said, there is substantial variation between the states. States such as Wyoming, Idaho, and North Carolina show that modern governments can be run using very little debt. Policymakers in Kentucky, Hawaii, and New York should learn fiscal lessons from these low-debt states—and so should the federal government.

Note: I have argued that state and local governments should keep debt to an absolute minimum. They should use taxes or other current charges to fund infrastructure, or they should privatize it. Krit Chanwong contributed to this post.

Two recent developments reinforce the case against new, large government-run rail projects. These initiatives are usually a bad deal for taxpayers given their high and unpredictable costs, long construction cycles, and disappointing ridership.

First, Prime Minister Rishi Sunak radically downsized Britain’s ambitious HS2 high-speed rail project intended to connect Central London with Manchester in the north of England. Estimated costs had ballooned to £106 billion (or $129 billion at the recent exchange rate).

HS2’s northern terminus will now be at Birmingham, limiting the project which was originally expected to include 335 miles of track, to just 140 miles. And the first six miles between London Euston and Old Oak Common in the London suburbs will only be built if private funding can be found.

Sunak announced his decision at the Conservative Party conference on October 4, telling attendees:

HS2 is the ultimate example of the old consensus. The result is a project whose costs have more than doubled, which has been repeatedly delayed and it is not scheduled to reach here in Manchester for almost two decades… and for which the economic case has massively weakened with the changes to business travel post-Covid. I say, to those who backed the project in the first place, the facts have changed. And the right thing to do when the facts change, is to have the courage to change direction. So I am ending this long-running saga. I am cancelling the rest of the HS2 project.

Much the same can be said of California’s high-speed rail project. As I discussed here previously, estimated project costs have roughly tripled since it was originally sold to voters in 2008 and the completion date will be at least a decade later than the original expectation of 2020. And, with California population not growing in accordance with projections, the high-speed rail authority has reduced its ridership forecasts.

Perhaps California Governor Gavin Newsom can take a page from Prime Minister Sunak by truncating his own high-speed rail project. Newsom indicated such an intention in his 2019 State of the State speech, but rapidly backtracked after facing criticism from project advocates.

Meanwhile, in Northern California residents have been absorbing the latest bad news about one of its own rail-based white elephants. The six-mile extension of the Bay Area Rapid Transit (BART) system through downtown San Jose received a new price tag and delivery date.

The construction cost estimate was raised from $9.3 billion to $12.2 billion (or roughly $2 billion per mile), and revenue service is now not expected until 2036.

The new completion date is one year after Californians will be prohibited from buying cars with internal combustion engines. As I wrote previously, the extension will replace very few car trips and a large proportion of those theoretically would have been in electric vehicles. As a result, the BART project is an extremely inefficient way of battling climate change.

Local leaders have generally supported the BART extension despite its poor cost/​benefit profile, but perhaps now someone in power will call for a reappraisal. As with high-speed rail, the project has strong support from advocates and special interests. It remains to be seen whether any California official will stand up to the train lobby and take the side of the state’s beleaguered taxpayers.