The Growth Experiment Revisited (2013), by Lawrence Lindsey

In Full Archive by Nathan Lewis

The Growth Experiment Revisited (2013), by Larry Lindsey, was the last book on the “supply side bookshelf” that I hadn’t read. After many years focusing mainly on monetary issues, after I finished Gold: the Final Standard last year I figured I should catch up on all that has been done over the years.

July 6, 2018: The “Supply Side” Bookshelf

The book is, I am happy to say, very good! All of the books on the “supply side bookshelf” are good. Usually, books on economics are a mixed bag. Sometimes, they are such a bundle of confusion (especially on monetary topics) that it is hard to even define what the author is trying to say. We have had many posts through the years where I have lamented that I have to “interpret” the author’s words even to have something sensible that I can talk about. The collective talent among the still small group of “supply siders” — people like Art Laffer, Steven Moore, Steve Forbes, Charles Adams, Larry Kudlow and Brian Domitrovic is most impressive.

The book consists of most, perhaps all, of the text of The Growth Experiment (1990), which was Lindsey’s take on the policies of the 1980s, plus another hundred pages or so taking things up to 2013. Lindsey created a reputation for rather precise analysis of economic effects. It is, overall, a somewhat academic approach, and one that misses a lot of big-picture things in my opinion. In general, Lindsey positions himself as a sort of sober centrist, between the academic Keynesians and the “supply-siders” making seemingly-impossible (but in real life often achieved) claims, although he is consistent in his support of supply-side principles throughout.

It really seems to bother a lot of people that a reduction in tax rates can lead to more revenue. You would think people would be excited about this idea, but instead, it seems to irritate them at the deepest level. When this is phrased as “tax cuts will pay for themselves,” it drives them wild with disdain. The fact of the matter is, this is quite common. Between 1950 and 1970, Japan’s government cut taxes every year. Every tax cut was expected, through “static accounting” to lead to a decline in revenue. But, tax revenues increased — and they not only increased, they increased by sixteen times, due to high economic growth. Even the revenue/GDP ratio was higher in 1970 than it was in 1955. For me, to try to deny that these things happen is pretty much just delusion. But, Lindsey has a strong sense of diplomacy, which has been reflected in his career in high station.

Thus, we find that Lindsey concludes that the Reagan tax cuts — especially the ERTA of 1981-1984 — led to a decline in revenue from the income tax, compared to if nothing had been done. This includes an adjustment for higher economic growth, although of course this is a fairly vague assumption because it is hard to say what the economy would have done if taxes hadn’t been changed. Also, the “growth” of the 1981-84 period was badly depressed by the recession of 1982, which was largely due to monetary factors, exacerbated by a tax increase in 1982. So, the actual growth of that time does not really reflect the potential unleashed by the tax reform, or the higher revenues that would have resulted if not for these extraneous monetary/other factors. The calculated decline in revenue was much less than the “static” accounting estimates, but nevertheless, there was a decline, according to Lindsey. He calculated that the reductions in tax rates on high incomes did indeed “produce more revenue.” But, this was matched with reductions in tax rates on low incomes. Reducing (for example) the tax rate on low incomes from 15% to 10% doesn’t change the tax base much (Lindsey argues). Behavior is not affected. It just results in a 33% drop in revenue, more-or-less along “static” accounting lines.

But, this analysis does not take into account the effect of higher growth on the revenue of all other taxes. Obviously, if GDP is 10% higher, then the income from (for example) the payroll tax should be about 10% higher too. Also, a bigger GDP improves the revenue from State and Local taxes. Even if the Federal government is not a direct beneficiary of this, it is still real money — and State and Local governments that have more money are less likely to pressure the Federal government for some kind of spending handout. This applies also to spending as a whole. A healthier economy produces fewer demands for spending, there being fewer people and institutions in distress. So, you get an advantage from higher income from all taxes, plus a tendency for spending to be lower (at least as a percentage of GDP, although it might be higher in nominal terms because GDP is 10% larger) than if the economy were less healthy.

In other words, there is a strong sense of political expediency here — that we are trying to get the politically palatable result, rather than the real result, despite all the analytical detail. Nevertheless, the analysis is worthwhile, and required reading for those interested in these topics.

Lindsey has a detailed and insightful discussion about “excess burden” where he concludes that the “revenue maximizing rate” of tax, as per the Laffer Curve, is not an “optimal” rate, but much higher than that. The “excess burden” is basically the cost of the tax in terms of lost opportunities, or what amounts to slower growth. It basically rises as the square of the tax rate. In a hypothetical example, revenue is maximized at a 50% rate. This produces revenue of $25, compared to revenue of $24 at a 40% rate and $21 with a 30% rate. However, the marginal cost of the last $1 of revenue is an “excess burden” of $4.50. So, the total cost to the economy is $1+$4.50, or $5.50. We give up $5.50 of goodies, overall, to move $1 to the government. The government’s revenue is indeed $1 higher, and thus “revenue maximizing,” but the cost of this last dollar is so great that this option should be avoided. In the example, a 10% rate of tax produces a $0.50 “excess burden” on $9 of revenue. We give up only $0.50 of lost opportunities for $9 in revenue, but the overall tax revenue is lower. We get $9 of revenue at a 10% rate, compared to $25 of revenue at a 50% rate.

In practice, we would have more growth at a 10% rate than a 50% rate — enough growth that, after a fairly short period of time, the 10% rate on a much larger GDP would produce more absolute revenue than if the rate was 50%, although the revenue/GDP ratio is lower. Nevertheless, government services are paid from absolute revenue, not relative, so this economy would have the same government services at a 10% rate. Plus, with a much healthier economy, you would have a lot less unhappiness, social pathology, and demands for socialistic government spending. So, the discussion about “excess burden” only captures some of the problems with excessively high “revenue maximizing” tax rates. But, even this is important.

Lindsey has a nice discussion of “fairness,” concluding that is essentially impossible to achieve, and gives a brief thumbs-up to the notion of “uniformity” in taxes.

As James McCullough wrote in the last century (19th), the introduction of progressivity seems to leave the tax system “at sea without rudder or compass.” Proportional taxes at least unite the interests of the nation behind a specific standard, while progressive taxation, McCullough argued, invites political mischief.

Such has proved the case. … Perhaps it really would be better to draw a single line in the dirt and unite all citizens on one side of it. (p. 154.)

September 3, 2018: Recent Thoughts on Taxes #3: The Principle of Uniformity

The basic problem is the size of government, or spending/GDP which of course translates directly into desired revenue/GDP. It would not be very hard, I think, to impose something like a uniform 10% VAT or “flat tax” on every dollar earned, with no exceptions, from the first dollar to the last. This is already lower than the 15.3% payroll taxes paid by the lowest earners today (although this is mitigated by the earned-income tax credit). But, raise the rate to 20%, and everyone starts to balk. Should the very lowest earners, who struggle for every dollar, be paying 20% to the government? They pay more than this in many European countries, where payroll taxes higher than 20% are added to VAT rates in the high teens. Nevertheless, it is an uncomfortable proposition. Thus we come to some form of “progressivity,” which quickly becomes complicated and, ultimately, insoluble.

August 12, 2018: Recent Thoughts on Taxes: In Praise of “Regressive” Taxes

August 19, 2018: Recent Thoughts on Taxes #2: Is the Payroll Tax the Best Income Tax?

After many chapters of diplomatic centrist incrementalism, Lindsey loosens his tie with a total-revamp tax reform proposal in the last chapter. He calls this the “Keep It Simple Stupid” or “KISS” tax. The idea is to have one tax base, and one tax rate. In this case, the tax base is Value Added, so it amounts to a simple Value Added tax with no exemptions. This would replace the Federal personal and corporate income tax, and also payroll taxes, which together produce 90% of Federal revenue. Basically, it is a proposal very much in line with the FairTax plan, although it uses the VAT form which the FairTax guys think is distasteful.

September 30, 2018: The FairTax

October 10, 2018: More FairTax Stuff

Lindsey calculated that the tax rate that would be essentially “revenue-neutral” under this plan is … 17%. “Revenue-neutral” here basically means the same revenue/GDP. Of course, it would not really be “revenue neutral,” because growth would be much higher, and would thus lead to much more revenue even if the revenue/GDP ratio was the same. You could also opt for a somewhat lower revenue/GDP, and thus a lower tax rate, and figure that you are going to make up the difference pretty quickly anyway from growth, but that is another discussion.

Aiming for a lower revenue/GDP, at the outset, has some advantages. “Revenue-neutral” approaches tend to run into political difficulties, because there is inevitably a mix of some people who would end up with lower taxes, and some that would end up with higher taxes (with “static” assumptions, which are inevitable in these situations). Of course those that would have higher taxes balk, and then discussions become difficult. A politically easier solution is to make taxes lower for everyone. People seem comfortable if some people get big reductions in their taxes, if they get at least a little one. This tends to make a proposal that seems to produce less revenue, at least from a “static accounting” standpoint. In practice, revenue tends to be quite ample anyway, and the higher growth leads to more revenue before too long. But, this requires an appreciation of “dynamic” effects in taxes. It requires statesmen to look at all the red ink in “static” estimates of revenue, to know that this will all be wrong, to know that you don’t really know what will actually happen but that it will probably be pretty good, and then do it anyway. Many statesmen have done exactly this over the decades, but it is still uncommon.

Even “dynamic” estimates of revenue tend to hew pretty closely to “static” estimates, because people find that they lose credibility when they make guesses that are closer to what really happens in real life. As we have seen in many real-life examples, the results can be “incredible,” which means: “not credible.” The revenue from Russia’s income tax rose 41% in one year with the implementation of the 13% flat tax in 2001, which replaced a system with a 30% top rate. This is what really happens in real life, but if you were to say that “revenue will increase by 20% in the first year” in a cabinet meeting, you would be laughed out of the room.

Adding in some allowance for tax evasion, and perhaps generating a little more revenue to replace other Federal revenue sources, Lindsay proposes a 20% tax rate (inclusive, which is common for VAT), which is exactly what the FairTax people propose, before their “prebate” program.

One possible way of implementing a completely uniform tax system is to break it into a “sales tax” and an “income tax.” This is basically the “9-9-9 Plan” advocated by Herman Cain in the 2012 Presidential primaries (Cain was, for a time, the frontrunner in the primaries.)

Wikipedia on the “9-9-9 Plan”

The “9-9-9 Plan” would introduce a 9% Federal sales tax, and 9% “Flat” income and corporate tax, replacing all existing Federal taxes. The combination would be roughly equivalent to an 18% VAT or an 18% “flat tax,” with no deductions, on the first and last dollar earned. This is a nice solution. For whatever reason, I find it far more palatable than the 18% VAT or 18% Flat Tax propositions, assuming the principle of “uniformity.” The problem with this plan is basically a quirk of our Federal system. I think we have decided that it is simply too dangerous to introduce both Federal consumption-based (sales or VAT) and income-based taxes. These could easily rise to very high rates in later years, just as the original income tax, with a top rate of 7%, did so.

One solution to this conundrum is to designate a maximum rate. But, there might be times when we really do want a big rise in taxes, notably during a major war. Another solution is to mandate (probably requiring a Constitutional Amendment) the principle of uniformity–which, as we have seen, was part of the original Constitutional framework. The requirement that all taxes apply to all people at the same rate, and that there are no deductions or progressive rate structure, will tend to keep a political cap on rates.

Instead, Lindsey does two things. First, he keeps the EITC and then adds a $1000 credit per worker. This reduces net taxes on lower incomes. Then, he adds what amounts to a 20% employment income tax on income over $120,000. This is basically like today’s payroll tax, but instead of an upper limit on income it has a lower one.

Thus, we end up with the “20-0-20” plan: a 20% VAT, no additional corporate taxation, some rebates for lower incomes, and a 20% “flat” income tax on employment income above $120,000. The effective rate on upper incomes is 40%, although this is split into a 20% VAT and something like a 20% income tax. (It would be paid directly by the employer however, like a payroll tax, and is thus more like an “indirect” tax. There would be no separate income tax return.)

First, let me note that this is a lot like the FairTax “prebate,” especially including the $1000-per-worker rebate. But, what is up with the new 40% effective top rate? That doesn’t seem so hot to me. The basic rationale behind the high tax rate is … more revenue, as Lindsey is not trying to make a “revenue-neutral” plan here but ultimately one that creates higher revenue/GDP, in response to higher spending/GDP expectations. Lindsey aims for a 20% increase in revenue/GDP. I think there is also some urge to maintain the present “progressive” rate structure, with the 40% effective top rate comparable to the present 39.6% rate, and he is just making up an excuse to do so. The extra 20% surtax seems to apply only to employment income. The income of most wealthy people is in the form of the profits of corporations owned by the wealthy. Warren Buffett is wealthy because of his large shareholdings in Berkshire Hathaway. His salary is negligible (I think it was $100,000 a year for a long time). Buffett pays taxes on this at the corporate level. Berkshire Hathaway paid $2,076 million in income taxes in 2017. With 16.5% ownership of the company, Buffett indirectly paid $347 million in taxes — money which, since he controls the company, he could have paid to himself in the form of salary or dividends, and then spent it to have a good time. I don’t think I like the idea of taxing employment income at a 20% marginal rate, while corporate income is untaxed — although this is a mirror of the present situation, where employment income has roughly a 40% top rate while corporate income is taxed at 21%. A Flat Tax like the 9-9-9 Plan would tax both corporate and employment income at the same rate. (The VAT, though it is typically conceived as equivalent to a sales tax, is in many ways equivalent to a “flat” tax on corporate income.)

The FairTax plan raises the sales tax rate from 20% to 23% to fund the “prebate.” I think this is a better solution overall than maintaining the EITC and then adding an additional refund, although it does introduce a large administrative issue, with the payment of effectively monthly sums to every person in the United States. This could be done in much the same way as welfare programs today use something like a debit card to which a monthly sum is added. It is a very cheap and effective solution, although the prospect of expanding it universally remains somewhat daunting.

In general, Lindsey stumbles on an issue that will become a big deal going forward. We can make up any exquisite tax plan we like, but it doesn’t make sense to have the world’s greatest tax plan to finance a terrible spending structure that has gradually rising spending/GDP more-or-less baked in the cake — and to no clear advantage from this expanding spending, either. We would basically be reduced to being tax collectors for the expanding and unsupportable Welfare State. In the book, Lindsey is constantly figuring out ways to raise taxes. He is especially fond of “bracket creep,” which, he notes, applies not only to inflationary nominal increases in income, but also real increases in income. A larger and larger number of people find themselves in higher and higher tax brackets, paying higher and higher rates, which Lindsey tends to favor. Why don’t we reform spending at the same time? This would require a total reconstruction of all welfare and entitlement programs, since they are not really fixable simply by adding tweaks.

The easiest way to “reconstruct” these is simply to devolve them to the State level, where they should have always been to begin with. Social Security is somewhat problematic, so we will keep that at the Federal level for now although it should be reformed also. All existing health, welfare, education, housing etc. programs would be devolved to the State level. This would cut Federal spending commitments by roughly 40%. The Federal Government could simply say: “As of January 1, 2020, all Federal health and welfare programs will cease entirely.”

Then, State governments, which are already involved in all of these things already, would be free to implement whatever health and welfare policies they favor, and impose taxes appropriately to pay for them. They would have a year or so to prepare. California and Massachusetts would no doubt introduce “single-payer” healthcare in some form. Other states, like Texas and Utah, might take a more market-based approach. The same would apply to all welfare and other such “social” programs. Federal responsibilities would consist of Social Security (which would be eventually reformed and wound down, probably replaced by a “provident fund” system), the military, debt service and Federal pensions — basically, the Enumerated Powers defined in the U.S. Constitution.

This would reduce the Federal government’s spending/GDP to about 12%, even including Social Security. That could be funded with a Federal VAT (or uniform Flat Tax with no deductions) of about 12%, which could be imposed as a “uniform” tax with no surtaxes or rebates. Taxes at the State level would of course be higher. The most straightforward way to do this is to raise existing State sales taxes — basically, to implement the FairTax, but with both taxing and spending centered on the State level, rather than the Federal. You could even include a “prebate,” at the State level.

States would experiment with different solutions. Successful examples could be imitated, and mistakes would be learned from. Competition between States would place a natural limit on bureaucratic expansion and inefficiency, much as corporations experience. People who would rather live under a different regime could easily migrate to a State where like-minded people have converged. This was the original vision of the united States of America, and it remains a superior solution today.