Minimum Wages and Payroll Taxes

There’s been a lot of talk lately about the minimum wage, now that New York and California have increased it to $15 (over the next few years).

This is a literal, not a rhetorical or loaded question: what stands in the way of keeping the minimum wage where it is, but (where mathematically possible) decreasing payroll taxes on minimum wage workers by the equivalent of the proposed increase in the minimum wage? I encountered the proposal decades ago in an op-ed piece in The Washington Post by James Glassman, but have heard almost nothing about it since.

I have no idea whether or not the minimum wage gives rise to a significant unemployment effect for low wage workers, but if there is one, you’d avoid it by this method–while engineering a “pay raise” for the very people who’d benefit from an increase in the minimum wage. Yes, you’d lose some payroll tax revenue, but if we’re talking about minimum wage workers, so what? How much revenue do they generate? And why would there be an imperative to take it from them anyway? They’re the wage earners who can least afford it.

One objection I can think of is procedural: while minimum wages are set at the state level, payroll taxes involve federal fiscal policy. But I wonder how conclusive that is. Anyway, the explanation can’t possibly be that unless we tax minimum wage workers, fiscal catastrophe will ensue. Can it?

I don’t have a worked-out view on this; just puzzled why it’s not part of the discussion or on the table as far as policy options.

13 thoughts on “Minimum Wages and Payroll Taxes

  1. I like this idea mainly because it avoids distorting markets (and hence, all else being equal, reducing the number of available jobs or hours of work available) and coercing employers.

    Here is one potential downside: this is another step toward making a system of general social insurance into a system of public welfare/charity. One of the central political means to the modern welfare state was moving from “poor laws” and the like to schemes of social insurance in which everyone participates on a fair basis. I say these systems are already at least halfway toward being public charity programs (in a functional sense and relative to other ways people do things like have money for retirement) and that the incentives will push them more and more in this direction. I suspect that we will have to have a more straight-forward public justification for what is, at least functionally, public charity anyway. Maybe the sooner the better?


  2. Could the payroll tax be decreased sufficiently to make up the difference in the increase of the minimum wage? That’s a sincere question; I have no idea what the numbers are.

    Here’s a few objections I can imagine people making even if the numbers worked out. First, the revenue isn’t much when you look at individuals, but it’s a ton when you sum up all the people working minimum wage jobs. If the gain in income among minimum-wage earners doesn’t alleviate their challenges and obstacles to the same degree as the social services funded by the tax revenue, then we end up with no improvement. Second, even granting the acceptability of progressive taxation, why should people who are gainfully employed full-time not contribute via taxes, just like everyone else? The minimum wage increase lifts these full-time workers out of poverty and enables them to contribute; your proposal might lift them out of poverty, but at the cost of making them free riders or something close to it — something that we might accept for the truly poor, but surely shouldn’t accept for the non-poor.

    These objections certainly aren’t conclusive, and I have no idea whether they even have much to be said for them. But they’re what instantly comes to mind.

    The question about the unemployment effect is an important one, and not one that I can answer. Economists of various stripes seem to be about as dogmatic in their views on that question as religious fundamentalists, and to disagree as much, too, so no appeal to authority is going to be sensible. It certainly matters, but so too does the question of how much we should value employment if the employed are paid wages that aren’t sufficient to enable them to live above the poverty level. If the complaint among businesses turns out to be that the law prevents them from hiring more people at wages so low that their employees can barely afford to eat (awful food) and pay rent (in slum housing), then it doesn’t seem like a very strong complaint, particularly when income disparity is so great.

    In any case, despite my saying some very anti-libertarian things here, I don’t have any firm views on this question, so I look forward to reading some more contributions.


  3. 1. Decreasing the payroll tax would increase state budget deficits (which are particularly bad in California and New York) at least in the short to medium run.

    2. Your proposal would be seen as a pro-rich, anti-poor bill which saves mega-corporations like Wal-Mart and MacDonald’s money instead of their poorest workers.

    3. The effects of changing the minimum wage are far more apparent than changing the payroll tax rate to the average voter. I’m not sure that the average voter even knows what payroll taxes are (even though they show up on their paychecks).

    Liked by 1 person

    • I believe it would satisfy those who want the hike to $15 very little. California’s minimum is now $10. Subtracting 6.2% for SS and 1.45% for Medicare leaves take-home pay of $9.235. Raising that to $10 is only $0.765 once versus $0.9235 five times in the proposal ($1 per year through 2021).

      I don’t understand fahertym’s first two comments. The payroll tax goes to the federal government, not the states. Also, many Wal-mart and McDonald’s employees are low-paid (“their poorest workers”).

      Liked by 1 person

      • I partially (or mostly) misunderstood the nature of the payroll tax. I thought it was only a tax directly on the employer for hiring employees, though that seems to only be a small percentage. Most of it is for social security and medicare as you say. I thought Irfan was saying that the government should eliminate the tax paid by employers for hiring employees, and then companies would naturally pass on some of those savings in the form of higher wages due to having lower input costs. This would probably be seen as anti-poor, pro-rich by the electorate.


      • Merlin,

        Thanks for doing the numbers. You’re right–the proposal wouldn’t satisfy those who want the highest hike from the higher or highest existing minimum wage baselines. But couldn’t it satisfy those who want a smaller hike from a lower baseline? If so, the proposal could be retooled to apply only there.


    • Just a clarification relevant to all three comments above: the “payroll tax” in Glassman’s sense refers to the dedicated payroll tax that funds Social Security, Medicare, and SSI (disability), not federal or state income taxes per se. I think colloquially, the phrase “payroll tax” has come to have two different meanings, one referring generically to all of the with-holdings from one’s paycheck, the other referring to the narrower sense Glassman has in mind. But the post was about the narrower sense. Low income workers are already effectively exempt from paying federal income taxes (vs. payroll taxes) because they fall under the income threshold to pay. Presumably, if your income is “low or moderate,” you’re eligible for an Earned Income Tax Credit.

      Re (1): exemption from the (dedicated) payroll tax doesn’t affect state budgets.
      Re (2): the proposal might be seen that way, but it might also be seen as making its beneficiaries “free riders,” and I think those two perceptions cancel each other out. But ultimately, I think the perception is wrong and irrelevant.
      Re (3): It’s an empirical question how apparent payroll taxes are. When I worked low and minimum wage jobs, people groused quite a lot about payroll taxes, and even produced conspiracy theories to the effect that the taxes went straight into our employers’ pockets. But I think the math and economics should trump the perceptions.

      I agree with your penultimate paragraph (on unemployment effects). Comments on the rest (applicable to Michael’s comment):

      Even if the payroll tax couldn’t be decreased sufficiently to make up the whole difference of keeping the minimum wage exactly what it is, you could combine a tax exemption with a smaller increase to the minimum wage to get the same effect. So if unemployment effects are proportional to an increase in the minimum wage, you’d have less of one with a tax exemption. But I don’t know the actual numbers, either.

      On tax revenue: the social services funded by the payroll tax are Social Security, Medicare, and disability. All three are potentially relevant to minimum wage workers. A minimum wage worker could need disability, and could be supporting someone who needs Social Security or Medicare. In some cases, she could be someone on Social Security or Medicare (e.g., someone over 65 at a minimum wage job). So decrease in revenue (if not offset some other way) could affect these cohorts.

      I don’t have precise data in front of me, but I find it implausible to think that most minimum wage workers would be adversely affected by decreased revenue streams to Social Security, Medicare, and disability. Even if some were, we’d be left with the fact those who weren’t might be left better off with more disposable income in hand right now. (It’s worth remembering that many minimum wage workers double up on jobs. So having two tax exemptions means having two minimum wage paychecks of disposable income. I can easily imagine circumstances in which a worker would be rational to prefer that revenue in hand to some incremental improvement to SSI.)

      Another option is that we do as we do in the case of the EITC: we redistribute the tax burden upward (to higher income workers). Of course, what this means is that once a minimum wage worker exits minimum wage work, they get whacked with the tax burden they thought they “escaped.”

      The preceding is the reason why I’m skeptical about the relevance of a free riding problem here. I don’t see that the proposal creates any special problem of free riding, or even any genuine free riders at all.

      First of all, the tax code is full of exemptions and deductions already. Some people see that as “free riding” (which would make for a huge number of “free riders” in this country), but you could plausibly see them as special exceptions or provisos to general tax rules, so that the exemptions/deductions are there to tailor the (very crudely drawn) rules to people’s special circumstances in a fair way (cf. Aristotelian equity, NE V.10). The Glassman proposal would just be another exemplification of this latter “equity”- or “fairness”-based approach, and not free riding any more than any other standard deduction is.

      Second, it’s worth remembering that no one literally pays no taxes. (Equivalently: everyone pays some taxes.) Even minimum wage workers exempted from Social Security payroll taxes would have to pay (e.g.) state sales taxes (which are highly regressive). If they rent, they indirectly pay property taxes through higher-than-otherwise rents (because property taxes are passed on to renters by landlords). Even if you qualify for a full Earned Income Tax Credit, you have to apply for it. You can’t just sit there and expect to get it automatically: the time you spend on making the application is unremunerated labor, whether it generates revenue or not. Not all services generate revenue: everybody is liable for jury duty, and has to take some action in response to a summons, even if it’s to ask to be excused (which isn’t always granted). Finally, as I said, once minimum wage workers escape minimum wage, they get hit with payroll taxes, so any “free riding” is temporary unless we’re talking about the tiny minority of workers who remain on minimum wage for their whole lives. (Minimum wage workers are themselves about 4-5% of workers.)

      So “free riding” is a tricky concept. If there is any “free riding” here, it simply consists in entirely general facts applicable to huge swatches of people: there is no exact correspondence between what you pay in taxes and what you consume in government services, nor is there any presumption that there should be one. Hence: Some people pay more than they consume. Many people consume more than they pay. Some pay for things they never consume. Etc. That’s just a feature of any system of progressive taxation. It’s a general reflection of the fact that a government is not a fee for service organization.

      Consider the case of John, who consumes $1 million in services over his lifetime, but pays only $500,000 in taxes. Now suppose that part of the shortfall arises from the fact that John was a minimum wage worker for 5 years, therefore exempt from dedicated payroll taxes. Then he was eligible for the EITC for ten years. Then he became richer and paid more in taxes, but didn’t live long enough to cover all this costs. When he dies, do we pronounce him a free rider? In one sense, perhaps, a “yes” follows from the sheer fact of the half million dollar shortfall. But in another sense, the answer is “no”: free riding is a normatively loaded, not a purely accounting-based concept. If he paid his fair share, he’s not a free rider. It depends on his circumstances. We could spin lots of other scenarios out in which middle class Joan (not a minimum wage worker) produces the same sort of revenue shortfall over the course of her lifetime, but for different reasons (different tax exemptions, let’s say), she still isn’t a free rider. For better or worse, none of us is literally expected to have paid back every penny of government services we ever consumed.

      On Michael’s potential downside: I don’t think there is a sharp distinction to be drawn between a system of social welfare and a system social insurance. Both are redistributive. Both involve “free riders” in the accounting sense. Neither need involve “free riders” in the normatively loaded sense I have in mind. Those seem to me the fundamental similarities.

      Both “welfare” and “social insurance” are, in my view, extremely misleading misnomers. “Welfare” is supposed to be a subvention you get from the government as a matter of right free and clear for doing nothing, but that is not what welfare actually is. Strictly speaking, welfare is a loan that you’re supposed to pay back after you get off welfare. Meanwhile, “social insurance” is supposed to be something you “pay for as you,” so that you “get what you put into the system.” But the truth is, there is no presumption that you literally get what you pay for under a social insurance scheme. So I don’t see a downside, because I don’t see any real difference in “moving from one thing to the other.” I just see movements, for better or worse, within a single redistributive system.


      • I’m with you on the free riding issue. It infuriates me how often people who discuss taxation pretend that income and payroll taxes (in the broad sense) are the only taxes that people pay. Still, it might be a relevant consideration from a rhetorical / popular opinion point of view, given that it would likely be seen by so many as allowing minimum wage earners to be free riders.

        As for the question of whether most minimum wage earners would be adversely affected by cuts to Medicare, Social Security, and disability, I think it’s sufficient to note that most of these people will eventually be in a position to collect social security and may very well need Medicare coverage. But even if I am not adversely affected in a direct way as a minimum wage earner, the adverse effects to other people living in my communities should matter, one might think. If nothing else, if old and disabled people aren’t getting support from government programs, somebody else has to foot the bill, and that’s likely to be their children, who are likely to be minimum wage earners in many cases. Is the amount of money they take home rather than pay into social security going to even out to the amount of money that their parents receive from social security? That doesn’t seem likely.

        In any case, it looks like the math won’t work anyway. But I don’t think the effects on employment are clear at all, and if they aren’t very severe, then it isn’t clear how much we should be worried about this. Suppose we could demonstrate that unemployment will rise given a minimum wage increase. If it nonetheless remained the case that everyone employed got paid a decent wage and that we had the tax revenue to take care of the people who can’t find work, wouldn’t that be better than an arrangement in which many people are working hard but being paid poverty wages while people who can’t find work aren’t getting anything at all? That’s not an entirely rhetorical question. The answer seems clear even if the antecedent is fulfilled — which I’m not taking for granted — but I figure somebody must have something plausible to say in favor of a negative answer given how many people seem to oppose any minimum wage or government run unemployment insurance.


        • I basically agree with your first and third paragraphs. We’re agreeing on the free rider issue, but I’m curious what others have to say about it. I just profess myself puzzled why minimum wage workers exempted from payroll taxes would be regarded as free riders in a world where everyone else takes all kinds of tax deductions. When self-employed people take deductions for business expenses, is that free riding? Why not say, “Yeah, so you had business expenses. So what? You think we’re going to let you free ride by giving you a deduction for it?” Same with the home mortgage interest deduction. I don’t see the principled distinction between the one thing and the other.

          I don’t really agree with this, though:

          But even if I am not adversely affected in a direct way as a minimum wage earner, the adverse effects to other people living in my communities should matter, one might think. If nothing else, if old and disabled people aren’t getting support from government programs, somebody else has to foot the bill, and that’s likely to be their children, who are likely to be minimum wage earners in many cases.

          I think a minimum wage worker would be well within her rights in saying:

          I have a strong, fundamental obligation of self-support. That obligation is a lot stronger than any obligation I have to somebody’s (or everybody’s) grandparents. So if the tax exemption means the difference between my paying the rent or not, or having a small cushion or not, or getting a bus pass or not, or getting gloves or new shoes this winter or not, that takes precedence to everybody’s grandma. They had a lifetime to save some money for retirement. I’m just starting out and trying to get by. First things first. Anyway, fact is, the minute I get out of this minimum wage hell hole, I’ll lose the exemption. So that’s when I’ll start worrying about other people, but not before then.

          We also shouldn’t exaggerate the neediness of Social Security or Medicare recipients as a group. Neither Social Security nor Medicare are means-tested programs (the way Medicaid is), so it seems a stretch to demand sympathy from low income workers for middle to high income beneficiaries. As for the low income beneficiaries, it seems fair to demand the relevant sympathy for them from higher income workers. I think minimum wage workers have a certain entitlement to saying, “Could you leave us alone to make some money for ourselves, at least for awhile?” The Glassman proposal reflects that ethical perspective, which I agree with. I’m assuming he offered it at a time (1996) when the numbers worked out better, but I think the proposal retains its ethical attractiveness even if the numbers don’t any longer work out the same way. (I don’t mean “let’s implement it in defiance of the math,” I mean “even if we can’t implement it, it makes a certain legitimate ethical point.”) Yes, the welfare of others should matter, and be a part of one’s own, but I think that other-concern shrinks to a relatively small increment when we’re talking about the lowest wage earners in the land (well, nearly the lowest, since some workers get sub minimum wages).

          I don’t have a view on the unemployment effects, but I find it odd that so much of the libertarian discussion should focus on it–indeed that so much discussion should focus on it at all. The real ethical-political issue is whether interference with an otherwise mutually consenting labor contract counts as an injustice or a rights violation, not the employment effects of doing so. There are actually problems on both sides here.

          The libertarian position ought to be “yes”: if I contract with you to receive 1 cent a day to dig ditches in the noonday heat, well, that’s what I get, even if it means that I’ll die after a week of work. I mean, I agreed to the work for that wage, so what’s the problem? And a third party’s stepping in and giving me $1 a day would be a coercive interference with our contract. Plus unemployment effects.

          But that isn’t what libertarians actually tend to say. This is what they say. But even if you want to say all that, it seems to me that if you want to call yourself a “libertarian” at all, you need to clarify what you think, in principle, about coercive (initiatory) interferences with labor contracts. Are they ever justified? After all, one standard libertarian view is that unconscionability is an illegitimate concept that violates libertarian principles. Is that what libertarians ultimately want to say–there are no unconscionable contracts, hence sub-minimum wages are not unconscionable and should be upheld, regardless of how low they go? At this point, it seems a dodge to invoke economics and say, “Well, employers don’t set wages at whim. They do it in response to market forces….” Yeah, yeah, fine. But what I want to know is: as a matter of principle, are you willing to repudiate the doctrine of unconscionability or not?

          The left-liberal position ought to be “no,” but this turns out to be less clear than one would think. Is the left-liberal position that sub-minimum wage labor contracts are unconscionable? I’ve never been clear on the answer. People will cite evidence against unemployment effects and insist on the need for “living wages,” but ultimately, I think one needs a general principle for when it is legitimate to interfere with an otherwise voluntary contract, and when not. Or else one needs an account of how low wages invalidate the voluntariness of a contract. And left-liberals seem (to me) reluctant to do either thing. Maybe the professional literature is structured differently than the popular debate, so I suppose I’m describing the latter, but I don’t think I’m misdescribing it. (Worth noting that unconscionability was the original justification for the minimum wage.)


          • Well, I’m not going to defend the libertarian view, and I’ll agree that left-liberals are often rather less clear on these issues than one might expect. I’m sympathetic to them insofar as it does seem clear to me that people might agree to work for wages much lower than what their work is worth and what they need; I don’t think we should be in much doubt that absent legal protections for workers, some employers will exploit some workers. But to my mind, at least, it isn’t entirely clear that minimum wage laws do the trick, or that they’re the best way of doing the trick. It’s even less clear to me whether raising the minimum wage in the cases we’re seeing now is the best course, and even less clear whether it would be best to raise the federal minimum wage, or to raise it to as high as $15. Dunno.

            As for your response to my second point, though, what you say there doesn’t really count against what I was trying to say, since I was assuming that the alternative — minimum wage hike without a decrease in payroll tax — would address the problems you mention there. I don’t mean to beg that question or set it beyond dispute; rather, my thought was supposed to be, assuming that the two proposals give minimum wage earners the same paycheck and that the amount of that paycheck is decent, if one option decreases support for social security, medicare, and disability, then that’s a strike against it. I suspect you and I think differently about obligations to others, but one thing I don’t think is that people are generally obligated to support others when doing so means that they can’t support themselves. I’m not entirely sure what I’d want to say in a case like the one you describe, but my initial claim wasn’t meant to apply to cases like that, but to cases where the wage earner gets the same thing in her bank account but contributes more to programs that help others.

            Ok, gotta go read about the eruption of Vesuvius in Latin with some undergraduates. Vastly less complicated than taxation.


            • On the second point: yes, we were talking about different cases. The case I hand in mind was one in which a minimum wage hike was insufficient to address the relevant problems unless there was a decrease in the payroll tax along with it, in which case I see no ethical problem in the worker’s wanting the decrease, even at the expense of revenue to Medicare, Social Security, and disability. But again, if minimum wage workers are 4-5% of workers as such, and this 4-5% of the aggregate are paying payroll taxes on a minimum wage, I can’t imagine that we’re talking about that much revenue. The tax seems to be there simply to make the point that its absence would indicate “free riding.”

              I’m not a big fan of BHL, but I actually thought that this post made a good point. The post also calls to mind that minimum wage laws are, in practice, full of exception clauses and loopholes that seldom get discussed. The federal one involves exceptions for workers with disabilities, full time students, workers under 20 in their first 90 days of work, tipped employees, and “student learners.” That weakens the claims of those who insist that minimum wage laws have problematic unemployment effects on “teenagers.” The federal minimum wage applies in a rather uneven way to teenagers, and state laws might avoid the “teenager unemployment problem” by following suit.

              I was having a conversation about the minimum wage (and other economic matters) with a waitress (sorry, “server”) in a diner somewhere in upstate New York a couple of weeks ago. She pointed out that the minimum wage for tipped workers was pretty damn low and that the deduction scheme was rather complicated (the figure she gave me was much lower than the one on this chart, so I think she was referring to the wage before January 1, 2016). When I asked what she thought about increasing the minimum wage for tipped workers, she said, “Oh, Jesus, that’ll be a fuckin’ disaster. The wage’ll go up, yeah, but the tips’ll dry up, because everyone’ll think we’ll rolling in the dough.” I don’t know if that was a policy analysis or a plea for a good tip or both, but I gave her a good tip. Anyway, she was adamantly in favor of keeping the minimum wage where it was, on the premise that a lower minimum wage made diner patrons more tip-sensitive and more generous with their tips. I have no idea whether or not she was right.

              Liked by 1 person

  4. On the issue of social insurance versus public charity: there is a difference between (a) forcing everyone to join up to actuarially sound unemployment or old age insurance scheme (mainly for their own sake, but partly for the sake of achieving more-desirable states in society) and (b) forcing everyone who is able to pay taxes to help the needy (mainly so that they be helped, but also partly so that all of us together may exercise the virtue of charity collectively and so that we get more-desirable states in society). The rationale for the first sort of thing is not “ethically redistributive” – it is not forcing people to do things for the sake other people (though it is, at least secondarily, forcing people to do things in order to promote desirable social conditions that otherwise would be difficult or impossible to obtain). I don’t find such necessary-paternalism and necessary-social-good-provision justifications particularly troublesome. By contrast, the rationale for the second sort of thing is “ethically redistributive” – it is forcing people to do things for the sake of benefiting others. This is sometimes justified (as when vulnerable, innocent people would suffer otherwise), but it is prima facie troublesome. I realize that, as they are set up, Medicare and Social Security have mandatory employer contributions and so may have something of an ethically redistributive element (though they would not if the better-social-conditions-otherwise-unachievable rationale is what is to justify the coerced employer participation/contribution).

    However, your argument, Irfan, does not seem to be that, as we have set them up, our social insurance schemes are also charity schemes (either in function or in justification or both). Rather, it seems be something like this: the proper level of analysis is that of all of society and everything that the government does (or governments at all levels do) and what this whole system, all of it taken together, does is inherently coercively and redistributive. However, even if we grant this, the functional point does not directly address justificatory issues, at least at a certain ideal level that I’m concerned with. (If you are stuck in a system that coerces everyone for the benefit of others not just for the benefit of society, then perhaps you should seek proportional benefit or help the least well-off – or whatever, the question is one of how to coercively redistribute, not whether to.) Why isn’t this true: we should have distinct systems, with distinct justifications, for straight-up social insurance (coerced prudence that also benefits society) and public charity (coerced/collective altruism that also benefits society)? Also I’m not sure that the functional point is right. I’m skeptical about the holistic scope of analysis. We have systems that were set up on a social insurance justification (though with coerced participation by employers) and that have largely functioned in that way. This functionality (not to mention the corresponding justification) is compromised if we start running running a social insurance scheme as a charity.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s