The US government has long been revenue short with respect to its spending needs – an ageing population; high inequality; increases in the costs of the goods the government provides, like healthcare and education; and the need for more defence spending demand greater government expenditure (Hunt 2017). In recent weeks, these needs have been heightened by the COVID-19 crisis. National debt was 79% of GDP at the end of 2019; it is expected to grow to 96% by 2022 due to relief measures already legislated (Borodovsky 2020) and will likely rise further as the crisis unfolds. Given that interest rates are at historic lows and in the absence of inflationary pressures, preoccupation with growing debt levels would be misplaced (Baldwin and Weder di Mauro 2020). But it is as useful as ever to identify ways for the US to raise more revenue in a progressive way.
Our recent work estimates that a substantial investment in the Internal Revenue Service’s ability to policy tax evasion will raise over $1 trillion in a decade (Sarin and Summers 2019). Since our study was released, former IRS Commissioner Charles Rossotti has weighed in on the potential of compliance efforts (Rossotti 2020). Drawing on his extensive experience at the helm of the IRS, he provides detailed estimates that suggest that better policing of tax evasion can generate even more than we expected – over $1.6 trillion in a decade.
The purpose of this column is to provide the current state of play with respect to the revenue potential of a robust investment in compliance. A careful review of the evidence makes clear that the IRS can expect to generate over $100 billion annually, and perhaps more than $200 billion, from better enforcement of existing tax laws. Compliance measures alone are insufficient to address the government’s large and growing revenue shortfall. But curbing tax evasion will raise revenue in a progressive way and has the advantage of making the tax code more efficient. It also has the virtue of increasing the revenue potential from any future rate increases or reforms that prove necessary.
Background on the tax gap
Today, the IRS fails to collect nearly 15% of federal taxes owed. Adjusting the most recent estimates for inflation and income growth, left unaddressed, it will fail to collect $7.5 trillion in the coming decade. And yet despite the magnitude of the tax gap, the IRS’s resources to police tax evasion are at historic lows. Since 2011, its budget (in real terms) has decreased by 15%, and its enforcement funds have decreased by 25% (Figure 1). Less enforcement capacity directly impacts the magnitude of the tax gap: there is a 1:1 relationship between the resources that the IRS has at its disposal to police tax evasion and the revenue that it is able to collect from those who underpay (Figure 2).
Figure 1 Decrease in real IRS budget since 2011
Figure 2 Change in examination coverage and additional tax liability imposed post-audits since 2011
We argue that overhauling the IRS’s enforcement efforts by (1) substantially increasing audit rates, with a focus on high-income individuals; (2) improving third-party information reporting; and (3) investing in better technology could raise $1.1 trillion in a decade, net of costs.
Commissioner Rossotti estimates that compliance investment can raise even more – over $1.6 trillion in a decade, net of costs. And unlike our more naïve approach, these estimates account for taxpayer response and the need to phase-in reforms: annual additional tax collections increase from $80 billion in the first year to over $260 billion by the end of the decade.
While it is difficult to be confident about precisely how much revenue will be collected, it is reasonable to expect $1 trillion over a decade at minimum, and likely closer to $1.5 trillion if the IRS is afforded the resources it needs to policy evasion.
Approaches to shrink the tax gap
Table 1 Approaches to shrink the tax gap
Increasing examination rates
Audit rates across all filing categories are at historical lows (Figure 3), but the decline is most pronounced for high-income individuals and the corporations they own. In fact, audit rates for top earners have fallen so drastically that today a person in the top 1% (making $500,000 or more annually) is as likely to face IRS scrutiny as a person on the Earned Income Tax Credit (EITC) (Figure 4).
Figure 3 Share of returns audited by filing category
Figure 4 Audit rates for those earning over $500,000 annually vs. EITC recipients
This is puzzling as both a matter of equity and efficiency. High-income individuals earn more, and thus the revenue consequences of their misreporting are more significant. Further, they earn income in more opaque categories – like rental income, capital gains, and proprietorship income – with higher rates of noncompliance. Relatedly, the wealthy own pass-through businesses, where net income is meant to be reported on individual returns. Yet these structures are opaque (nearly 15% of partnership income cannot be traced back to its ultimate owner) and under-examined (99.6% are not audited, see Cooper et al. 2016).
We estimate that increasing audit rates, especially for high-income earners, could generate over $700 billion in the coming decade directly. This perhaps understates the revenue potential, because increasing audit rates as we propose to 20% (or more) for millionaires will also have an indirect effect on those not examined. Further, Commissioner Rossotti points out that returns could be even larger if within the high-earner category, examination efforts are concentrated on those who accrue business income and focus on pass-through returns.
Improved information reporting
When a salaried employee files her tax return, income is correctly reported – and taxes fully paid – 99% of the time, because tax payments are automated by withholding requirements and income easily verified by simultaneous reports filed by her employer. Misreporting rates are 55 times higher for proprietorship income or rental and royalty income, free from third-party reporting requirements that would enable the IRS to verify the veracity of these filings.
Increasing information reporting is regarded as a highly effective way to increase tax compliance (Herndon 2019, McTigue 2019). But past efforts on this dimension have fallen short. The most recent effort to expand information reporting was a provision of the Affordable Care Act that required that businesses provide details on all purchases of goods and services over $600 annually. Revenue raised was limited, and the burden on small businesses was so significant that changes were rolled back.
In our work, we attempt estimation of the revenue potential of overhauling information reporting, proposing that we increase reporting requirements for categories of income that are most often misreported – like partnership, S-corporation, and proprietorship income – which we estimate could increase tax collection by nearly $1.5 trillion. This is a large sum, and it is unclear – especially given historical experience – how effective information reporting will be for these income sources. We naively haircut this estimate to just 15% of the $1.5 trillion total, suggesting a sizeable gain of $300 billion from increasing reporting requirements.
Commissioner Rossotti lays out a more concrete proposal for information reporting. For individuals who accrue substantial business income, returns can be verified with information their banks provide about deposits received and disbursements made. He estimates that the individual income gap can be shrunk by more than three times our estimate (over $960 billion in a decade) when such changes are coupled with technological advancements that allow the IRS to better identify evaders. Imposing similar cross-party reporting requirements on medium size pass-throughs is estimated to generate another $700 billion in revenue.
Overhauling IRS technology
As the share of electronic filings increased from less than 1% in 1986 to more than 90% today, the IRS’s technological capacity has not kept pace (Weinberger 2019). By the end of 2017, over 60% of IRS hardware was past its useable life (TIGTA 2017). The IRS’s systems are the oldest in the federal government – so antiquated that it is challenging to hire programmers able to work with them (Cohen 2017).
Technological deficiencies have real consequences for the efficacy of tax collection: even for existing information reports, the IRS is currently able to pursue just over 20% of the discrepancies that it identifies, and this overstates its success rate because its ability to match filings to information reports is imperfect.
Improved technology would enable the IRS to better target limited enforcement resources. And for increases in information reporting to reach their full potential, it will be imperative to efficiently identify discrepancies with cross-party reports and audit the most suspicious returns first. The IRS’s ability to do so currently is limited, but recent advances suggest substantial untapped potential: a pilot programme launched to apply big data approaches to detect filing errors and avoid the issuance of invalid refunds saved the IRS $4.4 billion in 2017, and cost only $90 billion – a 50:1 return (GAO 2018). Assume that additional investments, including an expansion of this programme, will have much lower returns, perhaps 10:1. Even this means spending $10 billion on investment in a decade would generate an additional $100 billion in tax revenue.
Our work and Commissioner Rossotti’s independent efforts leave us convinced that tax compliance is an issue with substantial significance for the federal budget. We believe that the revenue potential from a dedicated focus on tax compliance almost certainly exceeds $100 billion per year and may even reach $200 billion. Given the magnitude involved in the government’s response to COVID-19 and the challenges already preoccupying policymakers before the pandemic, this is likely not enough of a revenue increase. But it clearly the place for tax reform efforts to start.
Beyond its revenue potential, the case for compliance efforts is strengthened by the fact that largest portion of additional taxes collected will come from the top. High earners disproportionately accrue income in opaque categories like capital gains and proprietorship income, where misreporting rates range from 17% to 55%. Moreover, by targeting audit resources and the scope of reporting requirements, it is possible to pursue enhanced compliance programmes which would fall almost entirely on those above given income thresholds, without large revenue sacrifices. A final benefit of increased compliance is an improvement in economic efficiency and growth. Economic activities will take their most economically efficient form, rather than their most tax efficient one.