Economics, Taxes and Spending

Were taxes really higher in the 1950s?

1.4.12 Biggs Chart 1

Tax rates were high in the 1950s, we often hear, yet economic growth boomed. So why should we fear higher taxes today?

One answer is that taxes in the 50s weren’t really high. Yes, the top marginal tax rate was 90%, but it applied to almost no one. What matters more is the average marginal tax rate – that is, the average rate paid on the next dollar of earned income. That figure tells you more about the incentives facing individuals working in the economy.

And based on data from a 2009 study by Robert Barro and Charles Redlick, the good old days in terms of economic growth were also pretty good in terms of taxes. Barro and Redlick calculated average marginal tax rates inclusive of federal income taxes, Social Security taxes, and state income taxes. In the 1950s, the average marginal rates equaled just 25%, versus 37% in the 2000s.

Moreover, even these data have several limitations.

•    First, they consider only the gross social security payroll tax. But workers in the 1950s received significantly more in Social Security benefits than they paid in taxes; thus, the true net Social Security payroll tax was negative. Today, most workers will receive less in Social Security benefits than they pay in taxes.

•    Second, the deadweight loss of taxes rises with the square of the tax rate. So the economic costs of taxation today are roughly 20% larger than in the 50s. Raising taxes hurts the economy more when taxes are already high.

•    Third, the overall tax burden has shifted toward high earners, who are in general more responsive to tax rates.

Are taxes today lower than in the prosperous 1950s? No, not at all.

28 thoughts on “Were taxes really higher in the 1950s?

  1. ” First, they consider only the gross social security payroll tax. But workers in the 1950s received significantly more in Social Security benefits than they paid in taxes; thus, the true net Social Security payroll tax was negative. Today, most workers will receive less in Social Security benefits than they pay in taxes.”

    I don’t think that is true. How could they receive more in benefits when SS is a pay-as-you-go system from the FICA tax?

    the only way that could have happened is:

    1. – a huge surplus has accumulated
    2. – benefits were increased – and then later decreased

    this does not pass the smell test… as I suspect the rest of this post fails also.

    it’s not the tax rate alone that matters. The question is how much money was generated – relative to spending.

    Rates were higher under Clinton but the budget was balanced. Your rates and revenue need to pay for your spending – and your debt.

    • Not only that, SS benefits weren’t even taxable until 1983.

      It’s the old AEI shell game.

      And this statement: “One answer is that taxes in the 50s weren’t really high. Yes, the top marginal tax rate was 90%, but it applied to almost no one.”

      That was the point. It was SUPPOSED to apply to “almost no one” as Bush’s Hubbard designed tax system didn’t capture ANYTHING from that stratum. Which is who pays for his supper.

    • Larry, perhaps I wasn’t clear enough. The incentive presented by Social Security includes the taxes you pay today net of the future benefits those taxes entitle you to (Google “net tax rate” and you’ll see this concept is pretty common among policy analysts). So if I pay a Social Security tax of, say, 10% but that buys me future benefits equal to 10% of my current earnings (in present value, to be technical) then my net tax rate is zero. So I’m not referring to the budgetary balance of the Social Security program at the time — that is, how much it collects and pays out in a given year — but the balance between current taxes and future benefits on an individual level.

      • Andrew – are you accounting for the fact that SS benefits are:

        1. inflation-adjusted
        2. include disability, death, and survivor benefits?

        try to buy a private annuity that gives you these features in today’s dollars.

        • Way to ignore his point…

          Your two points are judgments on whether SS is a good safety net; his point is that the net taxes on SS are higher now than before because the benefits you receive as a % of your income are less than what you paid into the program, as a % of your income.

          • re: ” Way to ignore his point”

            my apologies.

            re: ” his point is that the net taxes on SS are higher now than before because the benefits you receive as a % of your income are less than what you paid into the program, as a % of your income.”

            can we parse this a bit:

            what are “net taxes on SS”?

            are you talking about FICA as a percent of income or the fact that SS is now means-tested for those who make over 25K?

            sorry to be asking for the clarifications but I did “google” the “net tax rate” and did not get much.

            go slow here and let’s make sure we are clear about what we are saying.

  2. I’m not talking about the fact that part of SS is means-tested now. Basically think of it like this: If I’m making $50,000 now, and I pay (for ease of calculations) 10% FICA — that’s $5,000. If in the future, my benefits are $4,000, then my net SS tax becomes 2% of that $50,000 or $1,000 — I’ve paid $5,000 for $4,000 of benefits.

    Andrew’s point, and I’m not necessarily addressing whether his point is correct or not — I’m not an expert at SS financing — is that the tax rate in the 50s should be considered even lower, because people were getting more in benefits than they actually paid in. Whereas the current taxpayers will receive fewer benefits than they’ve paid in.

    • somewhat better…

      ” is that the tax rate in the 50s should be considered even lower, because people were getting more in benefits than they actually paid in. Whereas the current taxpayers will receive fewer benefits than they’ve paid in.”

      are we talking about the FICA tax ?

      is this about that tax as a fixed percent no matter the value of the dollar when FICA is paid verses the value of the dollar when benefits are received?

      I still do not see how the relationship would be different in the 50′s with respect to now or the 60s, 70s, 80s, etc.

      It seems to be a point, an assertion without much in the way of convincing evidence…

      why would we say, for instance, 5000, vs 4000 rather than 4000, vs 5000? what is the specific basis for the assertion?

      the reason I got off on the inflation index was that I thought (perhaps wrongly) that, that would have some effect on this relationship.

      this is a bit complex… eh?

      • Think about it this way: Let’s say that you pay $5,000 in taxes this year, and that additional year of earnings/taxes entitles you to $5,000 in future benefits (in PV). In that case, your net tax rate would be zero. If you earned more than $5,000 in future benefits your next tax rate would be negative; if less than $5k, your net tax rate is positive. In the early years of the system the next tax rate was negative because benefits were relatively more generous; today, with a higher retirement age and insolvency (which means that today’s taxes ‘buy’ less future benefits), benefits are relatively less generous. This link (http://www.american.com/archive/2011/march/is-social-security-middle-class-welfare) shows net tax rates by birth cohort, but you can translate that into averages by year.

        • there are lots of moving parts to this.

          are we talking about constant dollars ?

          why/how would $5000 in (FICA?) taxes yield you a different amount in the future? What would cause that?

          can you talk about any of this without incorporating the effects of inflation into it?

          I know you understand but there seems to be some shorthand here I’m not cognizant of.

          I’m a little reactive to some things published by AEI because they appear to synthesize data without clearly showing methodology and data…

          but I’m proceeding on this as if your point is valid but I simply do not understand it or more precisely, I’m not understanding why you think the value changes.

          • What we’re talking about is ‘present value’ dollars, which account for both inflation and interest (since a dollar today can ‘buy’ you more dollars in the future). Your Social Security benefits are based on your earnings, so the more you work (and pay into the system) the more you receive back. From that, you can calculate the relationship between taxes and benefits. The SSA paper I link to in another reply would explain in more detail how this stuff is calculated. My only point is that, a) for Social Security, taxes lead to benefits; and b) in the past, taxes lead to MORE benefits than they do today, so the ‘net tax rate’ in the past would be lower than today’s. This argument isn’t very controversial among the geeks who work on this stuff, but it’s not commonly talked about.

          • re: ” This argument isn’t very controversial among the geeks who work on this stuff, but it’s not commonly talked about.”

            OMG!

            I’m looking at your last reference and the Table B columns that say

            Effective
            nominal
            interest rate

            and
            Effective
            real interest
            rate

            can you show me the 1950 years that you say resulted in higher benefits verses years that resulted in lower benefits or am I on to the wrong thing?

          • I think the reference you’re looking to is interest rates on the trust fund, which is a totally different thing than the implicit interest rate paid by Social Security at any given time (just as a person in a traditional DB pension receives a benefit based on the pension formula, not how much the fund earns in a given year). To show things you’d have to work out someone’s entire benefit, which is do-able but can’t be shown here. What you want to do is look at the tables showing money’s worth ratios by birth cohort. Those show roughly the ‘deal’ received by people born in different years. Obviously, the early generations are more likely to have been working during the 1950s, and so the average net tax rate at that time would have been lower.

          • oops.. wrong table. Okay… which table (down lower?)… and which other years than the 1950′s would you compare to the 1950s?

            but I’m still not understanding what is the mechanism that causes these fluctuations…either

          • The SSA figures show the average lifetime money’s worth ratio (total benefits vs total taxes) for people born in a given year. What I’m thinking about here is the average net tax rate for everyone in a given year, which would be some sort of weighted average of the lifetime net tax rates for people who would have been working during the 50s. I know this is confusing in terms of generating the actual numbers; you’d need a spreadsheet and a lot of data to do that. My point was simply qualitative: that the net tax rate is a useful measure, and it was lower in the 50s than today.

          • re: ” My point was simply qualitative: that the net tax rate is a useful measure, and it was lower in the 50s than today.”

            and we are convinced of that from your post – how?

            is it any more than just an assertion?

            Not easy to say that you’ve made a compelling argument here.

            assuming your assertion has merit – would it also mean that other pensions, investments, annuities would be affected in a similar way or is this unique to SS?

          • I think I’ve explained it about as well as I can; if you want to know more, Google “social security” and “net tax rate” and you’ll find a number of papers that explain how it works and why it matters. So it’s not an assertion so much as an argument backed by logic and outside references. This argument doesn’t really apply to other taxes; with Medicare, for instance, you receive the same benefits no matter how much you pay in. Likewise with most general taxation. The logic does apply to other pensions and illustrates why the idea of the net tax rate makes sense. If you pay into a 401(k), for instance, you don’t regard that as a tax because whatever you pay in is eventually returned to you. Similarly with DB pensions, etc. They don’t really factor into the tax issue, though, because they’re funded by contributions to your employer rather than to the government and, in the case of 401(k)s, they’re voluntary.

  3. Andrew,

    Great article. I think your conversation with LarryG demonstrates just how big the task is when trying to explain this to the public.

    • ” Great article. I think your conversation with LarryG demonstrates just how big the task is when trying to explain this to the public.”

      I appreciate the time Andrew took with me in explaining it but I have comments:

      1. – if one is going to make this claim, they must:

      a. – make a layman’s Prima facie case – the topic
      and the actuarials are complex.

      b. – make clear why this would not also apply to
      other non-SS/FICA saving for future benefits.

      For instance, give the same circumstances, if
      I chose IN ADDITION to SS/FICA to set aside
      an equal amount in a fund or annuity – why
      would the outcome be different?

      I’m not yet convinced that this is unique to SS/FICA other than FICA is mandatory and other saving is usually not (except for employee contributions to Defined benefit pensions/annuities).

      c. – how would you fix this ?

      d. AEI has an unfortunate pattern of synthesizing data and not making clear their specific sources and their actual calculations and manipulations of the data and that seriously undercuts all similar posts; actually puts
      additional burdens of disclosure and completeness even
      on legitimate efforts.

      If you’re going to make the assertion to lay people, you
      have to provide a compelling case on the same level.

      • “I chose IN ADDITION to SS/FICA to set aside
        an equal amount in a fund or annuity – why
        would the outcome be different?”

        Because that wouldn’t be a tax, and the article is about taxes.

        “make a layman’s Prima facie case – the topic
        and the actuarials are complex.”

        What happens to be “layman’s” about the concept of prima facie evidence? But anyway, beside the fact that this post is about an academic paper, not many articles on economics or government finance could be written if they had to completely avoid complex tables of data.

        “how would you fix this”

        Fix what, and why? The post isn’t about SS benefits, it’s just an observation. Or do you mean the paper’s lack of addressing it?

        “If you’re going to make the assertion to lay people, you
        have to provide a compelling case on the same level.”

        He did (over and over) but you kept challenging it because it lacked the (wonkish) details. Then when you get a detailed explanation… you complain that it’s too complex.

  4. How is Medicare accounted for? In the same way you noted the disparity in SS taxation versus benefits, in the case of Medicare there is additional taxation assessed but also a benefit that didn’t exist at that point in time. To be honest I’m guessing that people who retired post 1950 received Medicare benefits without actually having paid into the system.

    • re: Medicare.

      there are actually 4 Medicares, Part A, B,C and D

      and only Part A is funded from FICA.

      The other 3 are not pre-paid through payroll taxes but instead sold by the govt as fee-for-service insurance to qualified folks who are inclined to voluntarily purchase it.

    • Joe, That’s a good point: I believe the data from Barro and Redlick doesn’t include Medicare, so you can effectively add that to the marginal tax rate. The Medicare tax used to be capped up until the early 1990s, but now applies to all earnings, so you’d get a slight rise over time. And you’re right: Medicare was a big wealth transfer to seniors beginning in the mid-1960s, who received it but didn’t pay for it. (Plus, if you believe the research from Amy Finkelstein at Harvard, the introduction of Medicare didn’t do much to improve health…)

      • Andrew, She, Amy Finklestein, didn’t deal well enough with spillovers e.g., hospital build outs and EMS which impacted rural areas much more than urban. Some of the decline rate of mortality in the non-medicare population could be explained by spillovers. Difficult to measure for certain but an interesting issue.

        One example that I often cite is the use of trusses to deal with hernias prior to Medicare. With Medicare perceived as an “earned” benefit seniors with hernias opted for an operation even though charity was available. While a hernia does not impact mortality quality of life certainly is impacted. And quality of life measures are difficult.

        So, there are two issues, spillovers and quality of life, that Medicare potentially impacted that have not been accurately measured.

  5. Te top nominal rate in the 50s was 70% using 91 or whatever clouds the issue, use real life numbers.Reagan complained he paid in the 70% bracket and thus made no more than 4 movies per year. So that shows how a progressive system works,RR did nothing and someone else got the chance to make some money. If so few paid at 70%, then why would they lower the rate in 1964 to 50%? If people could deduct nearly everything , why would they gladly pay 50% when they could pay nothing at 70%? And that rate started at less than $100,000, so how much could a millionaire deduct? It just doesn’t make sense.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>