Vero on Tyler on the capital gains tax


George Mason University economist Tyler Cowen disagrees with my assessment of the impact of the capital gains tax cut in 1997 on economic growth. He writes:

Why think the Capital gains tax cut gets the credit? With loss offsets in the tax code, isn’t the real rate on capital gains pretty low in any case?  And the last decade had a relatively low (nominal, published) capital gains rate but not fantastic growth results. Also, keep in mind that the biggest effects of a cut (or hike) in capital gains rates are on old capital, not new capital. New capital (and other factors) drives growth.  But the incidence of capital gains taxes on new capital is largely, through incidence, bounced onto consumers and labor.

I think he may be right. Darn. And then there is this:

Arguments that the maximum CGT tax rate affects economic growth are even more tenuous: Capital gains rates display no contemporaneous correlation with real GDP growth during the last 50 years. Although the effect of capital gains on economic growth may occur with a lag, Burman (1999) tests lags of up to five years and finds no statistically significant effect. Moreover, any effect is likely small as capital gains realizations have averaged about 3 percent of GDP since 1960 and have never been more than 7.5 percent.

What do you guys think? I’d be interested in any argument you may have against or in favor of this paragraph.

Vero’s post is here and do read the comments.  I could have put my first point more precisely.  I’d like to see an estimate of the real rather than nominal published rate of tax on capital gains.  This would take into account loss offsets, carry-over, borrowing against the gains and deducting the interest payments, the option of holding until death, and a number of other factors.  I mentioned only loss offsets, which do have a cap and thus one must distinguish between the average and the marginal capital gains rate.  A lot of the measured and paid “capital gains tax” is actually slid in from what would otherwise be counted as ordinary income.

In any case, I regard this as one of the great myths of some schools of economics, namely the power of the capital gains tax rate.  One can be agnostic on the matter (for one thing the timing of the tax cuts is endogenous), but the data do not show the rate to have a big influence on subsequent economic growth.


Capitalism is about building productive capital that generates (ideally) constant returns long into the future.

Capital gains is about pump and dump. In 1997, the capital gain tax rate was cut, and much was make about how this would promote growth by, for example, rewarding the granting of stock option to the workers in startups which would innovate, go public, pop, and reward the hard work of those who came up with great ideas.

Thus we got great ideas like was almost a sure fire way to get a high, almost tax free return if you are stuck in a staid technology company with a 6 figure income and a marginal tax rate of 40%. And popped ensuring riches taxed at only 15% when you cashed out in a year and a day and retired. Much better than labor income. too dicey? No worry, ENRON is solid because tech requires electricity.

Every stock in the NASDAQ was measured by stock price appreciation, not investment in R&D, in factories, in productive capital. So, every firm was driven to do whatever it took to drive up the stock price.

On the other hand, Warren Buffett gets very little benefit from the reduced capital gains tax because he is a capitalist. He buys, holds, reinvests for decades, and ideally, forever.

The mindset behind the capital gains tax cut is to promote the creation of NASDAQ bubbles, more, more ENRONs.

The goal if you believe in capitalism is to create more investors who act like Warren Buffett: buy and hold and reinvest the dividends the productive capital generates.

Another way to look at it: think of a energy producer, whether an oil well or a wind generator or solar array. These require putting in real cash to create the capital. Once that is done, the value is the cash put in. The invisible hand has matched the amount of cash to the return on investment. Energy will always have a market, just like food - only mass death will cut the demand. So, what good is a reduced capital gains tax rate for real productive capital asset creation. There is no gain.

Thus the reduced capital gains tax creates the incentive to put the cash into some pump and dump scheme to reap the lower tax rate instead of creating real productive assets you will hold for the rest of your life because they generate a constant stream of cash.

Capital gains cuts are just another version of trickle down, so of course the results are diffuse, hard to see and maybe not even real. But imagine if the stimulus had been in the form of a credit for all income taxes paid last year, with a minimum of a few thousand dollars, even for those who did not owe taxes. Debts would be paid down, consumption would accelerate and it would have had a very big influence on subsequent economic growth.

Bernanke’s biggest flaw is that he thinks this is only a credit contraction recession, so he can fix it with credit expansion tools. He doesn’t realize that he is up against (1) a generational shift to the Gen Xers who got buried in student loans, and view credit very differently from the Baby Boomers, (2) the retirement of said Boomers, the pig in the python with the largest cohort turning 48 ( the year of peak consumption per Harry Dent) in 2005 – did something change after that? And (3) technology, which slashes prices every year (OMG – deflation!). the book launched yesterday – Bernanke isn’t going to change – given his history, he can’t – and Obama doesn’t have any idea how awful his appointments were. It's inflation coming, not deflation. Don't ever bet against the Fed.

"Bernanke’s biggest flaw is that he thinks this is only a credit contraction recession, so he can fix it with credit expansion tools."

I'm sorry, I must have missed the day when Congress ceded stimulus-via-spending authority to Bernanke, and he failed to use it.

"Well, that settles. Speculation bad, ergo 100% capital gains tax rate."

How about,

100% capital gains tax if you sell within 1 year,
70% if you sell within 2 years,
40% within 3 years,
20% 4 years,
10% 5 years,
0% thereafter.

You still get liquidity -- you can sell any time -- but you can't get a quick profit.

You'd buy stock because of how well you think the company will be doing a few years down the road, not because you think other investors will get enthusiastic in a few days.

I predict Wall Street would be 100% opposed. They get their commissions off pump-and-dump. They have expensive equipment designed to handle billions of trades a day. If investors thought long-term, Wall Street would lose a whole lot of money.

Wouldn't that be good for the economy, though? And for people's pensions etc?

Real nonresidential fixed investment growth was stronger in the three years prior to the 1997 capital gains tax than in the three years after the capital gains tax. This implies the the tax cut hurt capital spending.

Interestingly, the BEA publishes data on nonresidential fixed investment by legal organization. That means you can break out capital spending by corporations, nonprofits and those organizations such as s corps that are subject to the capital gains taxes. Since the 1950s when the US has been cutting marginal tax rates and taxes on dividends and capital gains the share of nonresidential fixed investment accounted for by those subject to personal tax rates has fallen from 25% of business fixed investment to about 11% of such investment.
This means that the growth in capital spending has been by organizations not subject to the individual income tax. This data calls into serious question all the arguments that lower individual income taxes generates greater investments.

You can find the data in table 4 at the BEA.

I don't think this sort of macroeconomic analysis can be taken very seriously. Suppose we take for granted that nearly all innovation created over the past 15 years wouldn't have happened were it not for low capital gains rates. Few would question that these technologies have had a profound effect on productivity and on the quality of our lives, yet if we attempted to analyse these benefits with a 5 year lag we'd be hard-pressed to find statistically significant effects since they take a long time to work there way into our highly developed economy.

Suppose it takes 2-3 years on average for capital to be converted into reasonably mature products. Can we reasonably expect to see these products fully saturate the market in the remaining ~2 years? I don't think so. The effects of innovation take a long time to really work their way into the economy both directly and indirectly (causing others to be inspired or copy the innovations).

In short, we need a 10-20 year lag to find the real fruits of risk capital, but such analysis would be meaningless given the other number of factors that influence our economy. Further, simply because we are experiencing diminishing returns on (lower) risk capital today doesn't mean that these effects aren't important. We had a lot of low hanging fruit for economic growth in earlier eras -- both in terms of technological innovation and in terms of other economic developments.

It's better to start from first principles when answering this question. Do low capital gains rates have a positive effect on risk capital? The answer to this is unquestionably yes (if you talk to entrepreneurs, venture capitalists, institutional investors, etc). Does risk capital tend to create the innovations which boost productivity and make our lives safer and more comfortable? I think most people would agree with this. Again, it may be true that relative returns have diminished and may remain this way, but this doesn't mean that they're not critical to long run economic growth. Investors may also make mistakes and invest irrationally in speculative investments, but what is the alternative? Are we going to get more rational investment from government? I hardly think so...


Suppose it takes 2-3 years on average for capital to be converted into reasonably mature products. Can we reasonably expect to see these products fully saturate the market in the remaining ~2 years? I don't think so. The effects of innovation take a long time to really work their way into the economy both directly and indirectly (causing others to be inspired or copy the innovations).

Then, suppose we had a capital gains tax of 100% if you sell within 5 years, and 0% if you sell after 5 years. That should have an entirely positive effect on the innovation we want to encourage, right?

et's take this in small easy steps:
- Entrepreneurs assume risk when they start a business.
- The biggest (pecuniary) return to starting and growing a successful business is when you sell it -- i.e. the capital gain.
- Taxing that gain will lower the return to taking the risk required to try to realize that gain (just as raising any price would do).
- The lower the return, as risk stays constant, the less entrepreneurs take the risk to begin with.

Let's see. A guy takes a risk and invests in a business he starts, and is rewarded by the increased capital value of his business in excess of his cash and labor input because he has innovated.

This increased capital value is computed from the annual profit he reaps from the business. Eg., the market rate of return on invested capital (his cash plus labor) is 10%, but he delivers 20% ROIC. Thus his $1M in cash plus labor invested has a market cap of $2M. He operates his business for 50 years, taking a dividend/salary of $200K a year (not reinvesting and growing), with the durable capital being depreciated and retires, selling the business - its name, reputation, fixed capital for a $1M. After all, it was his innovation that created the $1M in excess capital, so the buyer is going to see lower returns if he doesn't add the same innovation.

How much capital gains tax is the owner of this small business going to pay over the lifetime of the business?

Isn't the capital gains tax zero in this simplistic example?

But the double market ROIC indicates a capital gain from risk taking and successful innovation?

Your model seems to be that a guy starts a business by investing a $1M in capital and labor, claims a 20% ROIC so he can sell the business for $2M, so he can abandon the business to its new owner who might or might not have the needed innovation to keep the RIOC at the 10% of $2M.

Your model is that the 20% ROIC instead of the market 10% is not an incentive to start and operate for a long time an innovative business? Instead the only reason to start a business is a quick buck from dumping the business ASAP?

The full social and economic value of the capital investment might take 20 years to fully accrue to society, but this doesn't mean that it's efficient to require entrepreneurs and other early stage investors to wait this long. They have bills to pay and their time is better spent with at earlier stages of development than in a very mature company (never mind that they don't necessarily fully capture these benefits--innovation often spurs other innovation, encourages competitors to work around their patents, has economic value in excess of what they're able to charge for it, etc...); others meanwhile are better suited to this more intermediate stage and/or have the capital... Efficient capital markets are critical to entrepreneurial activity whether it's selling to a larger company or getting liquidity via IPO, etc....

Let's see, an innovator needs to start up and immediately sell out at a price higher than his investment because his innovation,, will take decades to turn a profit, and it might not turn a profit because innovators will pop up to under cut the value of the innovators, like expanding its book market place to include pet stuff.

The thing is, Bezos didn't rush in to buy up the billion dollars in stock - he just had some employees who thought "hey, we can sell pet food and dog collars on". And that employee didn't get a big pop in income that was taxed at the low capital gain tax rate. (Any pop in stock price was realized by Bezos, millions of stockholders, and maybe the employee, but the employee most like got a higher wage on which not only high income taxes, but also payroll tax was paid.

And a more real world example - circa the late 90s, the top search engine was Altavista, known for fast results and simple display of results - google was a university research project (just escaping the lab). The view was the employees who developed Altavista and made it so attractive weren't sufficiently motivated by being paid a wage, so Altavista was in effect sold off to reap the huge capital gains - the history is complex, but Mike Capellas argued Compaq's purchase of DEC was a profit just from DEC's cash and the market cap of the firm built on Altavista.

By your reasoning, Altavista represents the bigger success because it was cashed out quickly for big capital gains, while google was the loser because its founders just held on to their innovation for years, giving up the opportunity for quick riches.

Pete L,

This is the internet where nobody knows you're a dog, so I will tentatively accept your claim to be a successful entrepreneur. And so I will suppose that you know how you do business far better than me.

However, you do not know how you would do business if the rules of the game were changed. I am confident that you would find ways to succeed.

All of your complaints about taxation which would discourage early sale for great profit, represent problems you could and would solve.

So I tend to think that the inconvenience to real entrepreneurs would be far overshadowed by the changes this rule would make for non-entrepreneurs who hope to Make Money Fast with low-tax capital gains.

Peter L, thank you for the rational response! This is refreshing and rarer than I'd like it to be.

I have no direct experience as an entrepreneur. I've dreamed of doing that, and it looks like it would require too much arcane knowledge of government regulation, venture capitalists, tax details, patent law, etc. As it is I get paid a little for my work. I have the impression that in business for myself, unless I understood all the details, I would wind up with a contract or a tax obligation etc which would leave me working very hard with nothing to show for it.

I do of course have some direct and indirect experience with the stock market. Some of my indirect experience comes from co-workers. I've heard the following story about a dozen times: "I had a plan, and it worked, I could make 30% a year playing the stock market. It worked. But then I saw how to do it better. I could make 50% on my money. And that worked too. Then I tried to make 90%, and I lost it all. And here I am, back in the work force." I personally found I could make 30%, but something about it reminded me of the "gambler's ruin" from probability class.... The economy grows 3%, and my money grows 30%. How long could I keep that up?

People say that over a period of years, the stock market accurately reflects the value of public companies. But over a period of days, it is a crap shoot. Intermediate times give intermediate results. The largest part of the volume is people and machines gambling. And my pension fund depends on a highly paid fund manager who gambles my retirement. I wouldn't want his job -- when he has one unusually bad quarter he will be fired. I don't want anybody to have his job, but I don't get a choice.

What does any of this have to do with entrepreneurs? Well, if an entrepreneur can grow his company to the point it successfully goes public, he can make a whole lot of money. Sell his stock for a high price and then retire or start a new one. I've heard it's no fun running your company after you've sold it -- you face stockholder relations and the possibility of hostile takeovers and filing required public data etc. But it can be a whole lot of money.

Also the company can sell more stock later. But stockholders tend to disapprove, they'd rather own a company with more debt than a company with diluted ownership.

And I've known people who sold analytical tools for investors. They were entrepreneurs. Investors are a market.

I've heard one convincing justification for the stock market. An entrepreneur creates a successful private company and manages it until he dies. His incompetent children then run it into the ground. He can avoid this outcome by going public, hiring a professional manager with an MBA, and putting his stock into a trust fund for his children.

So, the money you make working gets taxed. But money you make gambling on the stock market gets a much lower tax? What are we trying to encourage here? Gambling on the stock market does not improve the nation's GDP.

Oh! In 2007, 20.7% of total value added to GDP came from the finance/insurance/real estate industry. Every time anybody bought or sold stock through a US broker, the brokerage fee added to US GDP. That was part of the 20%. When a foreigner bought or sold our stock it improved our GDP and our foreign exchange. So the stock market contributes to the economy that way. Though not as much as in 2007.

Insider information is supremely important for the stock market, because people who believe they have insider information can make stocks go up or down in the short run, and somebody wins big from that.

What good is all this? We can't afford it. We could kind of afford it back when the economy was growing fast and it didn't much matter who got the money.

Eliminate the chance to win big fast. Invest based on the expected health of the company in 5 years. If you think you'll need the money in less than 3 years, don't tie it up in investments. But you can get out of it what you put in unless the price is low when you need to sell.

Would it work to tax capital gains for speculators but not for entrepreneurs and venture capitalists and people who buy initial offerings? Maybe. But that provides more opportunities for scams.

If we got rid of the chance to buy low and sell high tomorrow to a speculator who got the word a day late, where would the smart money go? Probably some of it would go into casinos. But the rest? It's hard to get 500% on your money in 5 years betting on an established company. They are likely to have steady growth with unsteady reverses. But you have a chance at a big win in 5 years with a startup. Or before that, with a private company that's just getting started. I think there's a strong possibility that actual entrepreneurs would be better off.

About your personal problems, being personally undercapitalised and needing to sell too soon, maybe when there are eager buyers but you haven't put in the time, and maybe even when the company looks real good now but might not fetch such a good price in a couple of years, I kind of feel for you. It ain't all possum and redeye gravy being an employee, either.

If you bust your ass for 6 years and make due with little salary and you suddenly have 95% of your (now-liquid) net worth in one place, it's very risky not to diversify a little.

That makes sense. My priority is to make "pump and dump" type things less attractive. Yours is to provide flexibility for legitimate entrepreneurs. Is there a way to have both? To discourage one and not the other?

I can imagine piecemeal changes to my scheme that could help some. Like, people who buy stock could face high capital gains taxes for a set time, but people who were assigned stock at the IPO could sell theirs any time? That might handle one of your specific concerns, but only one. Is there some reasonably simple way to get more of what we both want?

Most fundamentally though, you're glossing over the fact that IPOs (or acquisition/merger) provide liquidity for most of the investors that relied upon that opportunity when they made their investment. For instance, pension funds, insurance companies, and other institutional investors, who overwhelming fund the VCs, have operating budgets that require cash at certain intervals. Likewise for the other early stage investors.

What I primarily object to is the stock market gambling game interfering in investment. I'm not certain how best to moderate that.

My natural thought is to limit the profit from stocks that are sold too soon. Then a lot of the gambling goes away. And the opportunity to manipulate the stock price to affect the options market is reduced. The options market itself is reduced. Etc.

But I haven't run complete simulations. For all I know the result would be that gambling entities might buy large blocks of stock, wait until long enough, and then manipulate prices until they have cycled through all their old stock and everything is new. Then wait again. Maybe it would be more likely that we would have one entity manipulating prices at a time.

Would you really insist on 100% capital gains taxes on anyone that needs to relocate or otherwise sell in <= 5 yrs?

I'm not sure. It looks real murky. What if you actually improve a house so it's worth more, and then you need to sell? Would it be better to overinsure it and burn it down? Bad kind of potlach.

Say we had a graduated system -- you lose 100% of profits the first year, 80% the second year, etc. Then if your house is worth more you still get something extra out of it, if you've owned it for at least one year. If you think there's a good chance you'll need to leave within a few years, don't buy a house. It may be less risky to rent.

What I'd particularly want to slow down is passive investment as speculation -- the hope that somebody else will do something to make you rich because you had the foresight to buy things cheap that other people did not know would become valuable. This is a zero-sum game, and we are better off when smart people do productive things instead. Reducing capital gains taxes in general, is heading in the wrong direction.

...if radical changes were to be imposed on capital markets like you suggest the ripple effects on businesses like mine would be very real and rather damaging.

I don't want that. I want a way to discourage zero-sum passive speculation, while encouraging innovative entrepreneurs. I think we might need something radical along those lines, and ideally we would build the alternative pathways for businesses like yours while (or before) we dismantle the dysfunctional stuff. But something not so radical could be useful and maybe adequate.

Do you have any suggestions?

Comments for this post are closed