(2) The Economic Costs of Wage and Tax Fraud

Introduction

The profits reaped by contractors and developers engaging in illegal labor practices come at a substantial cost to the workers and taxpayers of Massachusetts. This section assesses the dollar value of the direct costs of wage and tax fraud in the Commonwealth’s construction sector for 2019. We approach this in two ways. First, DUA audits identify both the number of workers misclassified and the gross payroll of workers who are directly affected. By applying tax and contribution rates required of Massachusetts construction employers, we calculate the losses to the state and its taxpayers directly resulting from the violations discovered by the DUA. However, since DUA audits represent just a fraction of the instances of payroll fraud, this approach can only provide a lower-bound projection of the economic costs.

Estimating the full costs of wage and tax fraud is challenging and, in the end, inexact. This is not unexpected, as firms and individuals involved in fraud often go to great lengths to conceal how much money is exchanged in the underground economy. This includes, but is not limited to, the use of cash-only payments and check-cashing operations instead of more established (and regulated) banking institutions. Because so much occurs in the shadows, we cannot directly assess the amount of cash that exchanges hands in order to estimate the full costs of payroll fraud. Instead, the approach we use—based on a model developed in a 2019 report commissioned by the Attorney General’s Office of the District of Columbia, which sought to resolve this very problem—relies on assumptions about the number of workers affected by fraud and their annual earnings.¹³³ We believe that the results of this approach are a thoughtful step forward in quantifying the costs of wage and tax fraud in the Massachusetts construction industry, however we acknowledge the inexactness of our outcomes and have accordingly been transparent about our assumptions and methods (see Appendix D).

Unemployment Insurance Audits

Table 7 presents the projected economic costs of worker misclassification in the Massachusetts construction industry for 2019 based solely on the findings of DUA audits. As a starting point of the analysis, DUA records reveal that the average worker identified as being misclassified earned an average of $9,459 on random audits and $10,866 in the sample that includes both random and targeted audits. These amounts imply that affected workers are typically seasonal, temporary, or otherwise hired for short-term projects.

Table 7. Estimated Costs of Worker Misclassification using DUA Audits, Massachusetts Construction Industry, 2019

The results in the bottom half of Table 7 reveal that worker misclassification in the Massachusetts construction industry allowed offending employers to illegally reduce their labor costs between $24.4 million and $32.6 million in 2019.¹³⁴ Workers bear a substantial portion of the burden, as it is projected that they were not paid between $2.7 million and $3.6 million in overtime pay (i.e., the “half” in “time-and-a-half”). In addition, misclassification allowed construction employers to offload between $8.4 million to $11.2 million of FICA tax obligations onto the backs of workers; this is because under the eyes of the law, workers operating as independent contractors or in an off-the-books arrangement are considered “self-employed” and therefore responsible for both the employee and employer share of the FICA tax. Misclassification also denies workers their legal rights to UI coverage, workers’ compensation insurance, and other benefits; this is the result of contractors’ evasion of contributions to these social programs, including $8.1 million to $10.8 million uncollected by the state’s unemployment insurance system and an additional $5.2 million to $7.0 million in workers’ compensation insurance premiums not paid.¹³⁵

Empirical Method: The Full Costs of Misclassification

The cost projections presented in Table 7 are substantial and calculated using direct evidence of worker misclassification from the Department of Unemployment Assistance. But these results are lower-bound cost estimates, as it is reminded that DUA audit results under-represent the volume of payroll fraud occurring in the Massachusetts construction industry. Further, while the costs above are based on an assumption of $9,000-$10,000 earnings per worker at each employer, it should be recognized that construction workers typically work for multiple contractors in a given year; as a result, the annual income of affected workers is likely much higher than the numbers provided. In order to develop a more complete set of cost projections—which encompass all instances of wage and tax fraud outlined in this study and workers’ annual earnings—we rely on an empirical method advanced by Dale Belman and Aaron Sojourner in 2019 and refined in a 2020 study by the Institute for Construction Economic Research. For a full review of this approach, its limitations, and its application to Massachusetts, see Appendix D.¹³⁶

In building cost projections of the full extent of worker misclassification, the lack of direct evidence of many instances of wage and tax fraud requires us to make assumptions about the total number of workers affected, their hours worked, and their annual earnings had they been employed legitimately; the use of assumptions introduces a nontrivial margin of error into the cost estimates. This issue compels us to present four different scenarios; these are provided in Table 8. The first column examines the economic costs assuming the minimum number of workers (22,146) offered by the indirect method described earlier in the paper and a conservative assumption about workers’ earnings level ($35,200, or the 10th percentile of legal earnings in construction occupations in the state).¹³⁷ Conversely, the last column explores the outcome with the maximum number of workers (36,719) affected and a less conservative income assumption ($44,960; the 25th percentile of legal earnings).¹³⁸

The results of this methodology presented in Table 8 reflect that wage and tax fraud likely allowed Massachusetts construction employers to reduce their labor costs by well over $100 million in 2019. Workers bear a substantial burden of these illegal labor practices. The results suggest that contractors evaded at least $19.3 million in overtime and premium pay in 2019 (i.e., the “half” in “time-and-a-half”). Employers also offloaded at least $59.6 million in Social Security and Medicare tax obligations onto the backs of workers. Add in the fact that these workers are not typically eligible for unemployment insurance benefits and are not covered by a workers’ compensation insurance policy, and it is evident that contractors’ and developers’ decisions to operate illegally degrades living conditions for workers and their families.

Table 8. Projected Economic Costs of Payroll Fraud, Massachusetts Construction Industry, 2019 (in $ millions)

Payroll fraud also harms taxpayers, representing an unwelcome public subsidy of illegal and unethical behavior. The results of Table 8 suggest that contractors’ evasion of their legal responsibilities led to a shortfall in the Massachusetts unemployment insurance fund of at least $24.5 million in 2019. Fraudulent employment practices also allowed contractors to avoid paying a minimum of $37.0 million in workers’ compensation insurance premiums. Combined with the evasion of overtime pay and the offloading of FICA responsibilities, the results of Table 8 reflect that wage and tax fraud allowed construction employers to illegally reduce their labor costs by an estimated $140.4 million using conservative assumptions rising to more than $200 million using less conservative models.

The results offered in Table 8 represent the direct costs associated with payroll fraud in Massachusetts’ construction industry, however there are also indirect economic costs. Although workers have a legal responsibility to report their full earnings to the IRS and Massachusetts DOR, employers’ failure to withhold taxes and provide required employment documentation open the door to affected workers to underreport or not report their income.¹³⁹ This leads to substantial income tax shortfalls for federal, state and local governments. The estimated income tax loss in Massachusetts caused by payroll fraud in the construction industry in 2019 is between $6.7 million and $41.3 million; this range is broad for reasons outlined in Appendix D. Further, these projections are based on very conservative assumptions; as such, the authors believe that these estimates likely understate the income tax loss to the Commonwealth.

While Table 8 represents an attempt to project the direct economic costs of wage and tax fraud, there are reasons to conclude these values may understate the full costs of worker misclassification in the Massachusetts construction industry. First, Part 1 of this study highlighted that contractors engaging in this form of illegality are also more likely to forego safety expenditures and ignore OSHA regulations. Such cost savings, however, are not captured in Table 8 as we lack credible estimates of the per-worker cost of safe and responsible contracting. Second, these results assume zero direct wage theft, or the explicit nonpayment of promised compensation to employees discussed in Part 1. We know this is an egregiously conservative assumption, however we lack credible quantitative estimates for its extent in the Massachusetts construction sector. However, if we had assumed that workers in fraudulent employment relationships lost 1% of their earnings to wage theft, the aggregate loss to workers—and cost savings to employers—would range from $7-$8 million (most conservative income and worker assumptions) to $14-$16 million (least conservative assumptions).¹⁴⁰ Finally, there are many other indirect costs— such as increased demand for public services, downward pressure on legal wages in the industry, lost profits for law-abiding businesses, lost tax revenue from labor brokers not reporting their income to the DOR, and decreased funding for apprenticeship training—that are also not incorporated in the analysis due to data limitations.¹⁴¹

1099-MISC Filings

The enormous projected cost to taxpayers offered above is further supported by a re-examination of 1099-MISC data provided by the DOR. Despite accounting for just a fraction of all 1099-MISC forms issued in the industry, the results indicate that $165.9 million in non-employee compensation was not reported on personal income tax forms between 2016 and 2019; this equates to 32% of the dollar value of all 1099-MISCs issued in the construction sector and reported to the DOR.¹⁴²

Although some of this income underreporting is attributable to non-fraudulent reasons, it is revealing that the “missing” money appears to be disproportionately lost to subsectors of construction identified earlier as the most likely to be engaging in wage and tax fraud (e.g., framing, painting, roofing, flooring); see Appendix B for a full review of 1099-MISC data by contractor type.^143,144 Considering that income underreporting rates are expected to be even higher among workers operating on a cash-only basis (i.e., those not issued a 1099-MISC), these results highlight both the magnitude of the dollars lost and the importance of employers issuing W-2s and engaging in income tax withholding.¹⁴⁵

Workers’ Compensation Trust Fund

As discussed earlier, contractors engaging in wage and tax fraud often fail to maintain a valid workers’ compensation insurance policy. These decisions impose substantial financial and psychological costs on injured workers and their families, as they must deal with enormous medical bills and a loss of income that should have been covered by a workers’ compensation policy. It is a practical impossibility to know the full extent of this problem, however data from the Department of Industrial Accidents offers some perspective. For this study, the DIA provided us with de-identified data on Workers’ Compensation Trust Fund cases from 2016 to 2020; the authors focused on the 300 cases with complete data on cash settlements in the file with at least $1,000 paid out. While the data do not allow for the definitive identification of workers’ occupation or industry on a case-by-case basis, our review of the data suggests that the state fund distributed between an estimated $5 million and $8 million to injured construction workers given the failure of their employing contractors to have a valid workers’ compensation insurance policy.^146,147

While the numbers above represent the financial impact to the Commonwealth, a review of the case files reflect a far more heartbreaking human cost. Injuries include blindness, amputations, broken spines, skull fractures, and innumerous broken bones. A substantial number of workplace accidents occur when workers fall from ladders, roofs and building tops; accidents involving table saws and nail guns also happen with unfortunate regularity. To make matters worse, some contractors’ self-interest and callousness goes beyond their failure to maintain a valid workers’ compensation policy: some cases involve the employer putting a worker in an unsafe situation and then failing to call an ambulance after the worker has suffered a devastating injury