By Hannah Halbert and Zach Schiller

Good morning, Chairman Blessing, Ranking Member Clyde and members of the committee. My name is Zach Schiller and I am research director of Policy Matters Ohio, a nonprofit, nonpartisan organization with the mission of creating a more prosperous, equitable, sustainable and inclusive Ohio. Thank you for the opportunity to testify.

House Bill 382 is an improvement over previous unemployment compensation (UC) solvency bills, but it still depends too heavily on cuts in worker benefits. In particular, the bill includes another 9 years of freezing maximum benefits and a reduction in the maximum number of weeks of benefits from 26 to 24 for most claimants. It makes modest increases in employer UC payroll taxes, raising the amount of wages subject to UC payroll taxes to $11,000 and adding a 0.1 percent increase to the minimum safe level (MSL) tax. Even with these changes, it does not produce solvency, even under the weaker solvency standard it adopts, by 2030. The bill leaves the Ohio Department of Job & Family Services to implement vague language on key issues such as dependency benefits and how many weeks of benefits will be available. Serious questions also exist in relation to the bonding of the debt called for in the proposed constitutional amendment. While the sponsor says employers and employees are splitting the cost of the plan almost 50-50, the analysis that produced that figure is incomplete and does not clearly support that conclusion.

We need a strong, solvent UC system. Solvency will benefit both employers and employees. As introduced, House Bill 382 does not accomplish these goals, though it includes some elements that could move us toward them. A year ago, Policy Matters Ohio proposed an employee tax as a key element in a solvency solution—and in improving access to our UC system, which remains inadequate for many low-wage workers. We are encouraged that House Bill 382 includes an employee contribution and consequently reduces the benefit cuts in previous bills. However, the employee premiums in the bill contain problematic provisions.

A real solvency solution must be based on the real reasons our fund went broke and remains weak. As we have previously testified, inadequate employer taxes are the major reason for that. Average employer taxes in Ohio are lower than those across the country, and state unemployment compensation taxes overall are a tiny share of employer costs (17 cents out of a total compensation bill of more than $33 an hour, according to the latest national survey by the U.S. Bureau of Labor Statistics). Ohio employer taxes have been lower than the national average for 19 out of the last 21 years. For the year ended last June, Ohio employers paid an average of 54 cents on every $100 in wages, less than in every neighboring state, and well below the national average of 63 cents. Data from the U.S. Department of Labor show than on average, Ohio employers paid $251 per employee last year in state unemployment taxes, vs. a national average of $337.

At the same time, benefit levels in Ohio are well in line with those nationally. Over the 12 months ended in June, our average benefit per week was $359.02, just 3 percent higher than the national average. But jobless Ohioans receive benefits for a shorter period of time on average than their counterparts across the nation. As a result, the overall amount paid per claimant is a little below what it is for the country. Our system for years has allowed fewer unemployed to qualify than in other states, partly because of our stringent earnings standard. This year, we require employees to average $247 a week over at least 20 weeks of work. This means that a minimum-wage worker employed 29 hours a week all year long would not be eligible for benefits. In the 12 months ended in June 2017, only 21 of every 100 jobless Ohioans were paid UC benefits, compared to 27 of every 100 unemployed workers in the U.S. Improving eligibility so that more low-wage workers can qualify would be a key improvement to Ohio’s system.

These conclusions about eligibility and tax and benefit levels need to be kept in mind as you examine House Bill 382. Here are some of our conclusions so far:

Not a real solvency package. HB 382 does not bring the UC system to solvency by 2030, even with the lower minimum safe level required under the bill. Assuming one moderate recession between now and then, the fund would be well over $1 billion below the revised solvency target. Solvency matters in part because it protects employers from federal tax increases triggered after the state has to borrow from the federal government to pay UC claims. It is important to consider that adjusting the UC financing system to make the system more solvent benefits Ohio employers.

Weeks of benefits: Cutting the maximum number of weeks of benefits to 24 would leave Ohio out of sync with the vast majority of states. Ohio has been paying up to 26 weeks of benefits for almost 70 years. This year, only 9 states offer less than a maximum of 26 weeks. The bill allows the full 26 weeks if “The individual was separated from the individual’s most recent employment because the individual’s ability to perform the work depended on weather conditions.” This is so vague it’s impossible to know how many or which workers would be affected. It calls on ODJFS to establish guidelines for determining what this means. The rules could be written to be expansive or narrow. More importantly, Ohioans who meet the already stringent conditions to qualify for unemployment benefits should be able to do so regardless of the weather.

Benefit freeze: The bill would extend the freeze on benefit maximums through 2028, or another 9 years beyond 2019, when the two-year freeze beginning next January 1 is set to end. That would freeze the current maximum for those without dependents at $443. ODJFS projects that would rise to $662 in 2029, when the freeze ends. So maximum benefits by that time would be about a third lower than they would be otherwise. This is a critical problem with the bill.

Dependency: The bill says that the ODJFS director can reduce the maximum weekly benefit “if the director finds that additional sources of household income reduce or eliminate the individual’s need to receive up to the maximum weekly benefit for that dependency class.” As it is now, claimants must report certain information about income to the agency in order to qualify for dependency benefits. Adding the new requirement would take that much farther, and besides being cumbersome, would place considerable power in the hands of the agency. This is an additional administrative burden that is particularly ill-timed when ODJFS is under cost pressure and has been taking steps that make access to benefits more difficult. It could seemingly allow ODJFS to look at whether someone is getting supplemental nutrition aid or other public benefits; if they’ve won the lottery; and whether their spouse is receiving a pension, along with any number of other sources of income.

Employee premium: The bill sets up an employee tax that says employees should pay 10 percent of what their employer does. Using such a tax instead of benefit cuts, as Policy Matters previously proposed, is a good idea. Employee contributions make up half the financing of Social Security (retirement, survivor’s and disability) and Medicare programs. It makes sense to ask employees to pay a modest amount while they are employed to avoid major reductions in benefits when claimants are unemployed and in financial crisis. We have advocated for an employee premium primarily for this reason. In our initial plan we included an employee contribution because it made the system truly solvent, avoided cuts to benefits, and allowed the state to expand benefits to low-wage workers. However, the form of the tax in HB 382 does none of this, and some of its particulars are far from ideal.

Federal law places no limits on the design of state employee contributions used for UC financing. This flexibility is confirmed by the variety exhibited in the three states currently using employee UC contributions. For example, states need not use either the federal taxable wage base of $7000 or Ohio’s existing $9000 taxable wage base when imposing a state UC payroll tax. Indeed, Pennsylvania imposes its UC employee contribution on total wages (that is, with no taxable wage base), while both Alaska and New Jersey use the same state taxable wage bases for employee contributions as those used for their employer UC taxes (Alaska ($39,700) and New Jersey ($32,600), as of 2016). Unlike HB 382, none of the three states has a threshold for how much employees have to work or earn before they start paying the tax

The table below summarizes recent information on employee payroll taxes in these three states.

The flexibility concerning UC employee taxes extends to having these taxes adjusted according a state’s UC financing needs. For example, in Alaska—which uses a unique UC tax rate-setting mechanism directly based upon benefit costs—the employee tax rate is set annually to produce 27 percent of the state’s average benefit cost rate while falling no lower than 0.5 percent or rising no higher than 1.0 percent of taxable wages. The Pennsylvania employee rate is likewise adjusted annually based upon the state’s trust fund solvency. In theory, an Ohio UC employee contribution could take effect only when triggered by trust fund levels or some other indicator.

Last year, the maximum employer tax rate in Ohio was $783 per employee; the average was $251, and the minimum was $27. So if HB 382 had been in effect last year, and worked the way you would think, the most an employee would have paid is $78, the average would have been $25, and the lowest would have been $3. Rep. Schuring has produced other figures, with rates ranging from $5.50 to $112.20, based on the proposed $11,000 taxable wage base. While this indicates that the amounts would be relatively low, it also underscores that the language outlining the employee tax is poorly drafted. It says that the employee premium is to be “ten percent of the premium paid by the employee’s employer based on the employee’s wages under section 4141.25 of the Revised Code.” Presumably, what is meant here is that employees will pay 10 percent of the average amount their employer paid per employee. But we need to better understand exactly how it would work and the rates that it would produce. Until we do, it is impossible to be sure of whether this is close to a 50-50 deal for employers and employees.

Several other aspects of the employee tax in the bill also need to be changed. Making the tax depend on how much in tax one’s employer pays is not the best way to go about it. Neither employers nor employees would choose to involuntarily separate workers from their jobs, and evidence is weak that higher employer taxes have any effect on the extent to which lay workers off. It’s not as if workers who are employed at high-tax employers will necessarily be better-paid. A better solution is the system in Pennsylvania, where employee tax is based on gross wages.

The bill calls for the tax to be withheld from employee paychecks only after they have worked enough and made enough to qualify for benefits if they were laid off. This means that taxes will only be withheld after each employee has worked 20 weeks. The bill also has a special provision covering employees who work enough to qualify for benefits, but do so at multiple employers, and never hit the 20-week threshold at one employer, so never pay the employee tax. It says that the agency will withhold from those claimants’ benefits the amount of taxes that they should have paid. This means that at the very moment they most need the money, claimants will see their benefits reduced.

An alternative solution would be to tax employees from the first dollar of wages they earn. The agency could keep track, based on employer filings, of how much tax is paid, could then refund tax money to those claimants who applied for benefits and were denied. ODJFS needs to be properly funded so it can assume this responsibility.

The bill also calls for the employee tax to be put into the mutualized account. That’s the account that is supposed to take care of charges that no individual employer is responsible for, such as when an employer goes out of business but has employees who receive benefits. Because of the way the current law is written, various funds that have nothing to do with the functioning of this account are added into it, including, bizarrely enough, the additional federal unemployment taxes that Ohio employers paid when they were in debt to the U.S. government. As a result, no employer has paid any mutualized tax since 2012, one reason why the fund is insufficiently financed. We should be getting rid of these flaws in the tax code, not adding to them. If we divert the employee tax into the mutualized account, it will cut employer taxes, weakening the solvency of the fund. The employee tax should go into the overall trust fund, not into the mutualized account.

Another major concern is how the tax would cover employees who work for nonprofits, governments and others that reimburse the agency for unemployment benefits they pay instead of paying the tax each quarter. There are 1.1 million Ohioans who work for these employers, more than 600,000 of whom are public employees. The bill would have them pay the employee tax, which makes sense, but it wouldn’t be withheld out of the regular earnings. Rather, the year after the reimbursing employer had paid the agency for benefits their laid-off employers got, workers at that employer would pay 10 percent of the total.

For instance, assume that a county has 500 employees and has to lay off 100 in a calendar year. These 100 employees each draw $400 a week for 20 weeks in the prior year. This comes to $800,000, which the county reimburses to the trust fund. As the bill is drafted, it appears as if the remaining 400 employees would pay a total of $80,000 in UC coinsurance the next year, or $200 per employee. This shows that workers could get hit with a larger amount than employees of private companies that pay regular UC taxes. A better solution would be to do what New Jersey and Pennsylvania do, namely, treat employees at reimbursing employers the same way as others, so they have the tax withheld from their regular pay.

Is this a 50-50 deal? A Sept. 29 memorandum from the Legislative Service Commission pegs the bill as close to a 50-50 deal, based on numbers from ODJFS. But in fact we don’t really know; some factors, like the clauses on dependency and the weather-related benefits, haven’t been included in the estimates. The estimates used in the memo go out to 2030. Since the freeze on maximums ends in 2028, the employer share grows significantly after that; if you estimated the shares only through 2028, employees would pay a larger share. Moreover, it appears that the LSC memo does not include the federal tax savings for employers of avoiding further federal borrowing that will take place under current law. ODJFS estimates that if nothing is done, the fund will become insolvent in 2020 if there is a modest recession, or in 2021 if there isn’t one. With or without a recession, the fund is projected to fall deeper into the red each year after that. So that means we likely would begin borrowing again from the federal government, and by 2023 or 2024, employers would have to make additional federal tax payments (or pay off money borrowed privately). These are additional taxes that employers will avoid (or at least will be reduced) if a solvency plan is approved. These lower employer taxes do not appear to be included in the estimates of who is paying for the plan. As a result of these factors, the memorandum does not capture the full impact of the bill, and it cannot be confidently said it is a 50-50 deal.

Bonding: The bill and separate HJR 4 call for a constitutional amendment allowing Ohio to issue private bonds to pay unemployment compensation debt. A number of states have done this, and it can at times allow states to save money vs. what they pay the federal government in interest. However, it is not a real solution to the UC debt; it merely transfers the debt to private investors, making it less transparent, and making profits for the private firms that handle the deal. In addition, there are significant questions about what has been proposed. Under HJR 4, obligations under the amendment “are general obligations of the state. The full faith and credit, revenue and taxing power of the state shall be pledged to the payment of debt service on those outstanding provisions as it becomes due, and bond retirement funds provisions shall be made for payment of that debt service.” None of the 8 states that issued UC bonds since the last recession did so as general obligation debt; instead, they issued revenue bonds that were backed by employer taxes added on to the existing UC employer tax. Backing such bonds with the full faith and credit of the state for the first time makes the state itself responsible in a way that it is not now for financing unemployment compensation. It also raises a question of what this will mean to the state’s debt ratings. Why does the State of Ohio uniquely want to make itself responsible for these obligations?

A few other points about HJR 4 raise questions. It limits the flexibility of the state by requiring that the bonds be tax-free. This would keep the state from using bond money to directly pay benefits, as some states have done. It also calls for the General Assembly to determine the maturity dates of the debt; an argument could be made that this should be done by financial experts.

The taxable wage base and the MSL rate. The bill raises the taxable wage base to $11,000. While that’s an improvement, it does not bring it anywhere near the national average of $13,782 or what it would be if it had been adjusted according to wage levels or prices. Raising the MSL rate does not deal with any of the flaws we’ve previously cited in the UC tax system (such as a formula for calculating that rate that lowers it at just the wrong moment, or the misguided accounting that attributes MSL taxes to the mutualized account and individual employer accounts) that undercut the effectiveness of the Ohio’s UC tax system.

Other issues. The bill raises a number of other issues that we will not go into in this testimony.

Summary. As detailed, HB 382 is an improvement over previous bills. It is a positive step to reduce benefit cuts and include an employee contribution, albeit one with significant flaws. But while the plan is a little closer to the true solvency bill that Ohio needs, it doesn’t fix the flaws in our employer tax and doesn’t get us to solvency. It cuts benefits too much, and still leaves us wondering how much employers and employees would contribute. It includes too many vague provisions, leaves too much up to ODJFS and does nothing to remedy Ohio’s overly stringent eligibility standard. The bonding proposal raises its own set of questions, beginning with why it would uniquely have Ohio issue general obligation bonds. The whole package, in short, needs a lot of work.

In fashioning a solvency solution, the General Assembly should keep in mind the main reason why we have a solvency problem in the first place. Employer taxes have been set too low. This underfunding has benefited employers during stable times, even while benefits for claimants have not been extravagant. In fact, many low-wage workers have been unable to participate in the system, helping to keep employer taxes low. Their continuing sacrifice should be considered in evaluating the overall fairness of any proposal.

Thank you for this opportunity to testify. I will be happy to answer any questions you may have.