Originally Posted by Kurt Horner

For bonds, I looked at the Moody's Aaa series versus the Baa series. The lower-rated bonds seem to have returns 1.2% higher over the entire 90-year period. As a result, Baa bonds are very nearly at ROI.

I averaged the Baa rates for the 1946-2003 period and got 7.7%. Subtracting the 4.1% inflation rate yield 3.6%, that same rate as the prime rate, and lower than the ROI of 4.4% over the same period.

There are, of course, bonds of lower grade than Baa. I tried to find the quantities of new issues above and below the investment grade level, but was unable to do so. Since Baa is at ROI and there are definitely bonds below that grade, then an average interest near ROI is quite reasonable.

No it isn't. First of all Baa is not at ROI, but lower. Investment funded by bonds will be a mixture of security having yield that fall between the Azz and Baa rate. That is why these two series are given.

What you re thinking of are junk bonds, which have yield considered higher than Baa debt. Junk bonds are not used for capital investment, they are used for financial in much the same way as margin is used (except there is no collateral).

For loans, I went here, and see that all commercial and industrial loans for early May were at ~3%. Given the inflation rate of -1.3%, this would be roughly equal to ROI.

Reporting data for one month during a period of financial turmoil (e.g. an extremely rare negative inflation rate not seen in my lifetime) does not say much about interest rates relative to ROI over the long run.

This does seem like good evidence that -- in the actual economy -- equity returns exceed financing costs. As noted above, I do concede that this is possible. What's important to note here is that the conclusions above show quite plainly that the state plays a role in this situation.

Um no. The interest rate return for the 1871-1920 was unaffected by the state, recall I got 4% just like they. But return to equity then was 7%, higher that the interest rate return. The authors did not give an equity return for that period so that can really say nothing about what happened then.

During the 1945-2002 period I show equity returns that agree with what they have, but I show interest returns of 2.7% -- still lower than equity returns.

Now I will grant that the state was involved in creating what appear to be artificially low interest rate returns. After all the adjusted rate the paper's authors show is just about the same as the market rate in the 1871=1920 period suggesting that debt investors seek to conserve their returns.

This state intervention results in a lower cost of capital for businesses after 1920 than before. But my analysis shows that return to equity fell at the same time, removing any advantage gained by the artificially low interest rates. During the entire 130+ year period the return to equity remained stubbornly higher than the return to debt.

While the market for investment evens out the returns, the actual cost borne by businesses seems to diverge from this.

No it doesn't--see above.

I should note as well that the particular data sets used tend to skew us toward an analysis of the upper echelons of the economy. Partly, this is the data we have, but it seems likely that this difference is only an artifact of looking at the top of the economy and not the whole. Of course, for the situation to even out, that would mean that ROI is frequently lower than the cost of financing for smaller, less capitalized businesses. That is also a notable observation about the economy we have.

Why can't what the data shows without lots of handwaving be what the data shows? Why all the wembling? Methinks thou doth protest too much.

There are observations and there are "observations." Certainly this discussion demonstrates the difficulty in making a value-free analysis using economic statistics.

Um interest rates are interest rates. There are what they are.

Economics is less of a hard science than, say, physics due to a trickier observation process. The temperature of the fluid in a beaker is much easier to define and determine in a value-free manner than the GDP of a country. The latter analysis requires determining what is and is not "production," what is and is not "domestic" and strains to find a measurement device that puts all the various items being measured into similar units. Economic analysis isn't just comparing a bunch of beaker measurements, it's more like comparing a bunch of beaker measuring experiments, each with different methodology, and trying to determine the overall beaker temperature trend. Good luck trying to do that without injecting some personal bias!

Hah! Obviously you don't do science or you wouldn't be so naive. Beaker experiments, as you call them, are hardly so easy to define. Right now at work we have an issue about the effect on filtration rates and yield in a chemical reaction step of removing a solvent. One faction (the scientists) sees higher yields and good filtration performance. The other faction (the engineers) sees lower yields and not very good filtration performance. I personally saw one of their filtrations and I swear to God it went flying through the filter. Didn't happen for us, and I'll be goddammed if I can tell you why. As an engineer I get to make the final decision on how to proceed, and I am letting the solvent change go. The other engineer (who is on vacation now) will be mad at me, but what can I d?, production wants to try the no solvent approach and since, as I see it, the risk of long filtration times and other hassles will be their problem and they want it, why not try it? After all, I did see this process work even if it doesn't in my lab. Maybe it will work in production.

All that is not to say that I take the extreme Austrian school position that empiricism is pointless.

Well that's a good thing since empiricism is hardly pointless. Despite all the politics, scientists do eventually figure it out.

Rather, when you say "I observe X" using economic data, we have to be extra careful to ensure we're factoring in possible errors in the process by which you arrived at that observation. This is because, often, economic "observations" are actually conclusions themselves.

Um observing bond prices at close is more precise than most scientific measurements, Most of the data I have been dealing with is a helluva lot more precise and accurate than filtration rates.

Originally Posted by Mikebert

I averaged the Baa rates for the 1946-2003 period and got 7.7%. Subtracting the 4.1% inflation rate yield 3.6%, that same rate as the prime rate, and lower than the ROI of 4.4% over the same period.

Did you change your date range in between posts? Previously, you said there was a 1.5% difference between ROI and Aaa bond rates. I pointed out there was a 1.2% difference between Aaa and Baa. That would make for a 0.3% difference between ROI and Baa which is small enough to be no difference at all. Somehow, in this post, the gap is back up to 0.8% (which still isn't all that big).

In order to get the 1.2% difference I just subtracted Aaa rates from Baa for the entire 1919-2009 series. When I do that for 1946-2003, I get 0.9%, so the gap is 0.6%, still less than your quoted numbers above. I'm not sure what's going on there.

Neither of our date ranges are optimal. I propose August 1929 to November 2007, which would be peak-to-peak between the two most recent crises of the capitalist system. For that, you get 3.84% real return on Baa bonds. What's ROI for that period? (I'm guessing low 4% range.)

Originally Posted by Mikebert

What you re thinking of are junk bonds, which have yield considered higher than Baa debt. Junk bonds are not used for capital investment, they are used for financial in much the same way as margin is used (except there is no collateral).

That's a heck of a statement to make without substantiation. While I'm sure the percentage of high-yield issues used for mergers and acquisitions is much higher than for investment grade bonds, it's not all of them. It's probably not even a quarter of them. Obviously, the average cost of bond financing will be lower if you systematically exclude the higher-yield bonds. So, I have to cry foul here.

Originally Posted by Mikebert

Reporting data for one month during a period of financial turmoil (e.g. an extremely rare negative inflation rate not seen in my lifetime) does not say much about interest rates relative to ROI over the long run.

And I admitted that in my post. I suppose I could go through the other 20 or so reports near that link, but most of them still fall during the crisis.

Originally Posted by Mikebert

This state intervention results in a lower cost of capital for businesses after 1920 than before. But my analysis shows that return to equity fell at the same time, removing any advantage gained by the artificially low interest rates.

Aren't your pre-1920 data coming from a different source than after?

Originally Posted by Kurt Horner

While the market for investment evens out the returns, the actual cost borne by businesses seems to diverge from this.

Originally Posted by Mikebert

No it doesn't--see above.

I was conceding your point here. The cost to investors of equity versus debt is about the same (i.e. the market evens this out). To firms, there appears to be a gap.

Originally Posted by Mikebert

Why can't what the data shows without lots of handwaving be what the data shows? Why all the wembling? Methinks thou doth protest too much.

Because the "data" for many economic statistics aren't data -- they're results. Now the Baa yield data -- that's data. But how about, say, the CPI? There's no way that counts as raw data. Significant choices have to be made in relative weighting of prices in order to determine the CPI.

Originally Posted by Mikebert

Hah! Obviously you don't do science or you wouldn't be so naive. Beaker experiments, as you call them, are hardly so easy to define.

I happen to be a structural engineer, but I'm not in the part of the industry that tests new techniques. (There's some pretty serious bureaucracy on that side of the industry that would probably infuriate me.) Nonetheless, I do understand the difficulty in setting up controlled experiments. However, the odd experimental results that you describe can't even be attempted in economics. There's no way to truly test a policy beforehand, since we don't have spare economies lying around to experiment on.

The fact that even the physical sciences have experimental issues is one of the reasons that hyper-Austrian anti-empiricism doesn't fly. There's a reason why so many of the Mises.org crowd are also skeptics of evolution and climate science.

Also, I should note that once you actually do implement your new filtration system, you'll know pretty soon whether the engineers or the scientists got it right. Similar error checking is hard in economics too, since economists can and often do claim other variable changes as the cause of the "failure" of a policy. There's a lot going on at once in an economy, which tends to prevent adequate separation of variables. As a result, value judgments creep in. Which is why there is such a thing as "left wing" and "right wing" economics, but no one ever talks about "left wing" chemistry.

***

I just found an article on turn-of-the-century dividend payouts on the London market. The crucial part is on pages 10-11, where the authors note that dividend payments at that time were significantly higher proportions of profits than in the present US market. So there's some evidence of an increase in retained earnings as capital markets have become more regulated.

Originally Posted by Kurt Horner

Did you change your date range in between posts? Previously, you said there was a 1.5% difference between ROI and Aaa bond rates. I pointed out there was a 1.2% difference between Aaa and Baa. That would make for a 0.3% difference between ROI and Baa which is small enough to be no difference at all. Somehow, in this post, the gap is back up to 0.8% (which still isn't all that big).

In order to get the 1.2% difference I just subtracted Aaa rates from Baa for the entire 1919-2009 series. When I do that for 1946-2003, I get 0.9%, so the gap is 0.6%, still less than your quoted numbers above. I'm not sure what's going on there.

You hit on part of the difference. The rest was the period for return on equity. In my prior post I was using the value I got for 1929-2002 that I produced in order to compare to the value in the paper. When I did the Baa calculation I did a ROI calc for the 1946-2002 period and got 4.4%.

Neither of our date ranges are optimal. I propose August 1929 to November 2007, which would be peak-to-peak between the two most recent crises of the capitalist system. For that, you get 3.84% real return on Baa bonds. What's ROI for that period? (I'm guessing low 4% range.)

I got 4.25%.

That's a heck of a statement to make without substantiation. While I'm sure the percentage of high-yield issues used for mergers and acquisitions is much higher than for investment grade bonds, it's not all of them. It's probably not even a quarter of them. Obviously, the average cost of bond financing will be lower if you systematically exclude the higher-yield bonds. So, I have to cry foul here.

Why? The issue here is the cost of all finance, retained profits, corporate bonds and loans. Junk bonds only became a significant part of finance in the 1980's and the purpose to which they were put was mergers and finance. I'm sure junk is issued for other things nowadays, today there is debt for every purpose under the sun (and then some), but the idea that junk is used to finance a large amount of business investment is pretty far-fetched.. And since that and above prime bank loans are the only classes of debt that fall above the ROI, they have to make up a BIG fraction to outweigh the other sources of funds which fall below ROI.

Even the paper you cited uses Aaa rates as an estimate for overall cost of capital for business. So I do not think I am out of line for using Aaa rates as my proxy for cost of money to business. Your assumption of high-interest bank loans and junk bonds as the principal source of funds for investment is a lot more far fetched. I've never seen anyone use junk bonds as a cost of money proxy.

Aren't your pre-1920 data coming from a different source than after?

No both are determined from my E/R proxy. There is no "official" source for return on equity data before 1929. Nobody was collecting data then. However, the stock market existed and stock prices were reported in newspapers. Companies issued earnings reports and dividends were reported. We have good information on the stock index back to 1871, which was assembled by the Cowles Commission in the 1930's. I already talked about R and how I calculate it. E/R can be used as a very good proxy for real return on equity for the stocks in the index, and it is a not unreasonable assumption that the index be representive of the economy as a whole.

To test the validity of the assumption, I calculated E/R for the 1929-2002 period to get 4%, which the paper you cited says is the value obtained by NIPA. So my E/R proxy gave the same value as the actual figure in the post-1929 period (when actual data are available). If it worked after 1929 why wouldn't it work before 1929 period? One has to work with the data that are there. But surely using a tested proxy is better than simply assuming that the ROI before 1929 is the same as after 1929 as the authors of that paper did (see below).

Because the "data" for many economic statistics aren't data -- they're results. Now the Baa yield data -- that's data. But how about, say, the CPI? There's no way that counts as raw data.

How is it any different from the interest rate data? Both are indexes (averages) of market prices. You don't have to use the CPI, you can use a commodity price average. You get essentially the same result over the long run since prices for everything moves up more or less together on a longer time scale. If the value of the dollar is falling (and this explains almost all price inflation in the long run) then just about everything will rise by about the same amount. Over short (politically relevant) periods how one chooses to construct an index can have some impact, but over the longer run it doesn't.

However, the odd experimental results that you describe can't even be attempted in economics.

And they cannot be done in astronomy, meteorology, or paleontology either, but these are still sciences. And there are experiments in economics, particularly in behavioral economics.

What makes economics different from science isn't the ease of doing experiments. It is a preference for theory over empiricism, exactly what you have been doing here. You apparently believe that interest rate should be the same as return, and in that paper you cited, they mentioned that theory says this should be so (so it makes perfect sense for you to believe this). But when someone looked at the data and found a problem, what did the economists do? They redid the study and manipulated the data so it now fits the theory.

As a result, value judgments creep in. Which is why there is such a thing as "left wing" and "right wing" economics, but no one ever talks about "left wing" chemistry.

But of course there are value judgments in science/technology too. For example, you have the scientists and the engineers. We have different worldviews and so you can get factions on projects. One group favors one explanation while another favors a different one. What makes science different is conflicts are resolved by data (its that focus on empiricism). Scientists are usually not all that strong on reasoning (unless they are theoretical scientists). It's usually just easier to let the data speak for themselves, and if it is not clear, then go get more.

Thus, the opinion of even a well-respected researcher doesn't count for much at meetings unless he has his own data. If he has have no lab resource employed on the project then he has little influence.

With time, conflicts are resolved because enough data is generated. But most of the time decisions on what data to generate have to be made based on inconclusive data because data requires resources and there are always other projects that can use resources. So the filtration issue hasn't be resolved in the lab (as it should be) because I am not working on it. I have other projects and as far as i was concerned, eliminating the solvent did not yield benefits large enough to justify my working on that instead of another project. The chemist who is the champion feels otherwise and so has continued to work on it. Since I was the engineer what I say goes, I play a role in this. But since I hadn't generated any more of my own data and the politics of the matter favored adopting it, I went along with it, after the chemist made a strong claim that is will work. If it doesn't, it won't be me you was wrong.

I just found an article on turn-of-the-century dividend payouts on the London market. The crucial part is on pages 10-11, where the authors note that dividend payments at that time were significantly higher proportions of profits than in the present US market. So there's some evidence of an increase in retained earnings as capital markets have become more regulated.

I already noted this in the articles of mine I gave you urls for. Here in the US, before 1920 ROI was 7%. 2% was retained and the other 5% paid out as dividends. Since 1920 ROI has been 4.2%, the same 2% has been paid out leaving only 2.2% for dividends.

So the majority of profits were paid out over the 1871-1920 period, but only about half of the profits since 1920. This observations, which you have found evidence for in Britain suggests a greater return on equity existed in the past, allowing more dividends to be paid.

This greater return on equity was associated with higher real interest rates, as I mentioned before.

I am not read in economics so I don't know economic theory. Based on our discussion I have learned that the real interest rate and the return on equity is supposed to be the same. I believe that some idea like this was used in the central argument in the book Dow 36,000 . They argued that according to economic theory there should not be a risk premium for stocks (i.e. stocks should earn about the same as bonds, meaning that ROI should be about the same as interest) and so P/E should rise to about 100, meaning that the Dow at 10,000 was still undervalued and would have to rise to about 36,000 before it became fully valued. Thus one should buy in 1998 or 1999 to take advantage of this once-in-lifetime gain. If you accept their premise, their prediction was perfectly sound.

I thought their premise was absurd, and in my 2000 book, I explained why the Dow 36,000 guys had it wrong and that we would not see Dow 36,000 for several decades. I argued that a secular bear market was beginning and stocks would under-perform money markets for 20 years so stocks should be sold in 2000. Nine years later I think it is safe to say that I was right and they were wrong.

It now seems that their premise was valid according to standard economic theory, and so they were not crazy to write than book. Apparently the theory is simply wrong.

I wrote an article in 2001 in which I used my E/R proxy for return on equity. I noted as an aside that return on equity is higher than real interest rate. I figured it was a risk premium for business akin to the risk premium for equity than was being so clearly demonstrated by the behavior of the stock market in 2001. I had no idea that this claim was contrary to economic theory.

But my main focus in that article was about stock market valuation, not the issue we are discussing. I am quite sure I was using a government rate of interest in that article because the interest rates look lower than what I have been getting here. I had not yet obtained my data base of corporate interest rates and these kind of data were not so conveniently available in 2001 as they are today.

Going back to the underlying question - is the "opposed wealth accumulation" inherent to the capitalist system or caused by extraordinary measures by government and/or the 'extreme capitalists' themselves.

Could it be a result of both? That to some degree, it is inherent in the system but facilitated to some degree by extraordinary, and unjust, measures?

But, to really put the William James hat on, does it make a pragmatic difference?

Mike has already offered the solution of an activist government that would... now how could we say this, hmm...oh yea, .... tax the shit out of 'em (TSOE).

Now, obviously, Justin would see the danger in an activist government. So, let's keep it real simple - at some clearly specified level(s) of wealth accumulation, there must be an effective rate of taxation, say 80%. No "ifs," "ands" or "buts." With all their twisting and turning and accountants, these 'extreme capitalists' must show, in the final analysis, an 80% effective rate of taxation. The key would be to keep the notion of "income" as simple as possible and maybe put out a "finder's fee" for anyone who can find unreported income from our 'extreme capitalists.'

I think this is a much better solution for Kurt as well. Even if correct that the unwanted wealth accumulation is due to "extraordinary measures" and injustice, who cares? Let's TSOE! Much simpler than attempting to mitigate the myriad of probably uncountable "extraordinary measures."

The only question is what to do with all those new billions in the govt trough. Health care? Infrastructure? Paying down the debt?

Originally Posted by Mikebert

I got 4.25% {for ROI 1929-2007}.

OK, so that's a difference of 0.41% -- a resounding "maybe" for either theory.

Originally Posted by Mikebert

Why? The issue here is the cost of all finance, retained profits, corporate bonds and loans. Junk bonds only became a significant part of finance in the 1980's and the purpose to which they were put was mergers and finance.

No, that's when they got a lot of media attention because the use of such bonds in a high profile fashion was new.

Originally Posted by Mikebert

And since that and above prime bank loans are the only classes of debt that fall above the ROI, they have to make up a BIG fraction to outweigh the other sources of funds which fall below ROI.

. . . or have a substantially higher average interest rate. Let's say 40% of the market is A grade at ~5%, 40% is B grade at ~7% and 20% high-yield at ~12%. Then you'd get an average of 7.2%, higher than the B grade. It would help if we actually knew the relative distribution of these bond types, the percentage actually used for capital investment and the average interest for junk bonds -- but I don't think that data is readily available.

Originally Posted by Mikebert

Your assumption of high-interest bank loans and junk bonds as the principal source of funds for investment is a lot more far fetched. I've never seen anyone use junk bonds as a cost of money proxy.

I wouldn't do it either, I'm just saying that if you're going to use a proxy you need to be close to the overall average.

Originally Posted by Mikebert

No both are determined from my E/R proxy . . . We have good information on the stock index back to 1871, which was assembled by the Cowles Commission in the 1930's . . . To test the validity of the assumption, I calculated E/R for the 1929-2002 period to get 4%, which the paper you cited says is the value obtained by NIPA. So my E/R proxy gave the same value as the actual figure in the post-1929 period (when actual data are available). If it worked after 1929 why wouldn't it work before 1929 period?

Sounds good to me. Your ROI data is probably as good as one can get. Our disagreement seems to hinge more on the cost of raising funds.

Originally Posted by Mikebert

How is it any different from the interest rate data? Both are indexes (averages) of market prices.

True, but everything in the bond rate data is very similar and easily collected. Such is not true for the "basket of goods" used to determine the CPI.

Originally Posted by Mikebert

You don't have to use the CPI, you can use a commodity price average.

Which is better than the CPI, but still worse than the bond rate numbers.

Originally Posted by Mikebert

Over short (politically relevant) periods how one chooses to construct an index can have some impact, but over the longer run it doesn't.

There's also likely to be value judgment in just what constitutes the longer run.

Originally Posted by Mikebert

What makes economics different from science isn't the ease of doing experiments. It is a preference for theory over empiricism, exactly what you have been doing here. You apparently believe that interest rate should be the same as return, and in that paper you cited, they mentioned that theory says this should be so (so it makes perfect sense for you to believe this). But when someone looked at the data and found a problem, what did the economists do? They redid the study and manipulated the data so it now fits the theory.

Yep, but the same process occurs in other sciences, even physics. The typical reaction to an anomaly is to look for errors in data collection or interpretation, not to trash the whole theory. In most cases, the "anomaly" is just imprecision. Even when theories collapse, core elements of the old theory are often incorporated into the new one.

Originally Posted by Mikebert

But of course there are value judgments in science/technology too. For example, you have the scientists and the engineers. We have different worldviews and so you can get factions on projects.

But does the science vs. engineering split manifest in national politics? No. There are left-wing engineers, but is there left-wing engineering?

Originally Posted by Mikebert

I am not read in economics so I don't know economic theory. Based on our discussion I have learned that the real interest rate and the return on equity is supposed to be the same.

I don't think that conclusion would be universal, but it's a reasonable extrapolation from the Modigliani-Miller theorem -- that it shouldn't matter how a company funds its capital investment. (See here and the related article on M-M.) If this is true, then the market should be attempting to even out the cost of investing in capital, which, per above, it appears to do. Note that this theorem has a number of assumptions behind it that to varying degrees don't fully apply to the observed economy. This means that an equity premium could exist and M-M still be true. It also means that anyone making investment decisions based on M-M is a moron (see below).

Now, the specific notion of uniform interest rates is actually an Austrian school concept. Even there, however, we're still describing a long-run tendency of the market, not something which really occurs. Most economic theorems are like this -- they describe the trajectory of the economy, not a specific state.

Originally Posted by Mikebert

I believe that some idea like this was used in the central argument in the book Dow 36,000 . They argued that according to economic theory there should not be a risk premium for stocks (i.e. stocks should earn about the same as bonds, meaning that ROI should be about the same as interest) and so P/E should rise to about 100

A 1% earnings yield? That doesn't even make sense on theoretical grounds. Based on late 90s bond yields, P/E of ~30 would be have been more than enough to make bonds perform better. I'm not seeing how they got from a theoretical indifference between sources of capital funding to being outrageous stock market bulls.

It's important to note here that since theory construction is a deductive process, the validity of any predictions are dependent on correct premises. If your theory has certain initial assumptions and those conditions don't actually occur then one should expect deviations when you do empirical measurement. For example, I conceded above that an equity premium may exist for certain types of firms (seen from their perspective). In fact, it may even occur for the entire market (definitely temporarily, and perhaps long term*). What you could do, in the case of economics, is predict the way in which your measurements should deviate from the theoretical case.

* Although, I doubt that -- and there is our specific disagreement.

Originally Posted by playwrite

Could it be a result of both? That to some degree, it is inherent in the system but facilitated to some degree by extraordinary, and unjust, measures?

Well, I would say, no, not inherent in capital markets, but . . .

Originally Posted by playwrite

But, to really put the William James hat on, does it make a pragmatic difference?

Short term, no, it doesn't. I'm more than willing to accept policies of downward income redistribution as long as there is concerted effort to end policies of economic centralization. If my view of the economy is correct, then once enough of the centralizing policies are removed, the desire for downward redistribution will wither. If I'm wrong, or I'm right and the campaign for decentralization is less than successful -- then at least no one is completely screwed in the meantime. All I ask from liberals is to stop treating the centralizing policies as if they were brought down from the mountain by Moses.

Originally Posted by playwrite

Mike has already offered the solution of an activist government that would... now how could we say this, hmm...oh yea, .... tax the shit out of 'em (TSOE).

That, in and of itself, is not sufficient. You can tax the rich all you want, but if the primary government expenditures still promote and bolster centralization, then all you've done is complicate their budgets.

Originally Posted by playwrite

I think this is a much better solution for Kurt as well. Even if correct that the unwanted wealth accumulation is due to "extraordinary measures" and injustice, who cares? Let's TSOE! Much simpler than attempting to mitigate the myriad of probably uncountable "extraordinary measures."

Again, near term, this might be correct. Long term, though, you've just given the ruling class a whole bunch more toys to play with -- and given them cause to fight amongst each other for control of the political system.

There's also the matter of how precisely people are taxed, or more specifically, if different types of income are treated differently by the tax system. Right now, Wall Street brokers often pay a lower percentage of their income in taxes than their secretaries and neither of them pays the percentage of taxes that most celebrities do.

Originally Posted by playwrite

The only question is what to do with all those new billions in the govt trough. Health care? Infrastructure? Paying down the debt?

Definitely not infrastructure, and health care could easily be done badly too.

I would propose a Citizen's Dividend. Scrap all current income assistance programs and just give an equal lump sum to all adult citizens. In addition, roll most current taxes into a simple progressive income tax with capital gains taxed as income. Have the corporate income tax only be paid on retained earnings (dividends would taxed in the income taxes of the individual shareholders). The result of all that is a pretty level playing field between methods of earning income, increased income at the bottom and no disincentives to earning more income of a particular type.

Beyond that, make a concerted attack on corporate personhood, intellectual "property" and corporate welfare as well as a host of rules that muck up the health care system.

Originally Posted by Kurt Horner

or have a substantially higher average interest rate. Let's say 40% of the market is A grade at ~5%, 40% is B grade at ~7% and 20% high-yield at ~12%. Then you'd get an average of 7.2%, higher than the B grade. It would help if we actually knew the relative distribution of these bond types, the percentage actually used for capital investment and the average interest for junk bonds -- but I don't think that data is readily available... I'm just saying that if you're going to use a proxy you need to be close to the overall average.

And I maintained that Aaa rates are closer than Baa rates when you consider that most funds come from the compamy's cash returns, that otherwise carry a low return. Other researchers used Aaa rates or even government rates. I have never seen anyone use Baa rates for this sort of analysis.

Here's a different line of reasoning. When you talk about low grade bonds, their rate of interest is higher because of the increased risk of default. But when we talk about return on equity we are talking about the stock index: the S&P500 and its precursors, because that is the data that is avaliable going way back. Companies in the S&P500 are investment grade (blue chip) stocks, they don't go bankrupt at anywhere near the rate small caps do. Investment grade stocks should be compared to investment grade (Aaa) bonds. If you want to use Baa rates then you should compare that to returns from small cap stocks, which have given higher returns than the S&P500 over the long run.

Sounds good to me. Your ROI data is probably as good as one can get. Our disagreement seems to hinge more on the cost of raising funds.

If so then why don't you accept the result for the 1871-1920 comparison, when there is much less interference by complicating effects? Here the interest rate I got was 4% (same as they got in that paper) compared to ROI of 7%, a whopping 3% difference.

The same process occurs in other sciences, even physics. The typical reaction to an anomaly is to look for errors in data collection or interpretation, not to trash the whole theory.

Yes, but in the case of science the old theory was based on empircal data, that's what makes it science. In economics and other social studies the theory isn't based on data.

But does the science vs. engineering split manifest in national politics? No. There are left-wing engineers, but is there left-wing engineering

Economics used to be called political economy. Politics is central to the subject matter. Left wing and right wing views are part of the subject matter of the field and cannot be divorced from the study of it. Look at the discussion we are having here.

The way you see the world is colored by your political beliefs. When you see data that contradicts your beliefs your first response is to question the data, to argue that it cannot be valid because it disagrees with logical derivations from premises you believe. Later you attempted to get some data of your own. For example, now you are trying to use low grade corporate bonds as your proxy for the cost of money applicable to capital investments. You choose this because they are biased towards what you want to show. The paper's authors started with Aaa bonds, which is a more defensible assumption, but then they had to manipulate the data to get it to show what they wanted.

Now the same thing happens in science, except its not about national poltics because this is not relevant to the subject matter of science. For example, I believe that we should multiply phase separation times by the cube root of the scale up factor. So when my 80 gram lab run took 45 minutes for the phases to separate to give a good-looking interface (in my opinion) I argue it will take like 15 hours at the 600 kg scale. On the other hand since we had already run the old process successfully at the 66 kg scale, and we got good breaks in less than an hour, I figure we will get breaks within 2 hours at the 600 kg scale.

The chemist disagrees. He has gotten good breaks with the new process in 10 minutes and figures we will get good breaks in 1-2 hours in the plant based on his previous plant experience.

Now obviously we choose the premises based on our positions. As an engineer I tend to support the other engineer, who is dead set against the new process. The chemist supports the results his technician has gotten and is a big booster for it. The other engineer does not like that this technician goes to meetings as if she were a associate scientist and expresses opinions about technical issues. The other engineer and chemist have clashed before. I tend to play a mediator role. As I have said I do not know what is the best thing to do. You usually don't and so politics and beliefs matter a great deal as to how projects go and what is believed about the way things work technically.

Science is a human process, just like an other kind of sausage-making. What makes it different is old premises are revisted with new data. Nobody questions the data first, as happens in non-scientific fields. If you get a wierd result you don't ignore it, you repeat the observation. Most of time you get something else and by the time you have made enough observations to be sure of what your data say, the weirdness is gone. However if you get a persistent weird result, then you publish. Others don't ignore what you report, they will attempt to repeat what you did. If they can't then your weird result probably reflects some oddity in your experimental set up. It is just like the initial weird result that doesn't get repeated. This is what happeend to cold fusion.

But if the weirdness persists, then people start looking at the theory and the premises. This is how quantum mechanics developed.

I don't think that conclusion would be universal, but it's a reasonable extrapolation from the Modigliani-Miller theorem -- that it shouldn't matter how a company funds its capital investment. (See here and the related article on M-M.) If this is true, then the market should be attempting to even out the cost of investing in capital, which, per above, it appears to do. Note that this theorem has a number of assumptions behind it that to varying degrees don't fully apply to the observed economy. This means that an equity premium could exist and M-M still be true. It also means that anyone making investment decisions based on M-M is a moron (see below).

Why is one a moron if they base investment decisions on M-M? Because it does not work? If it doesn't work, then how valid can it be?

Now, the specific notion of uniform interest rates is actually an Austrian school concept. Even there, however, we're still describing a long-run tendency of the market, not something which really occurs. Most economic theorems are like this -- they describe the trajectory of the economy, not a specific state.

My understanding of Austrians is they eschew empricism entirely. What they do is akin to what natural philosophers did before there was science. These philsophers derived their theories on reasonable premises and used them to explain their observations. They were all wrong, as later scientific investigation has revealed.

I don't think the philosphical approach is a useful way to understand how the world works because it is next to impossible to guess what the right premises should be.

A 1% earnings yield? That doesn't even make sense on theoretical grounds. Based on late 90s bond yields, P/E of ~30 would be have been more than enough to make bonds perform better.

Their argument was based on the observation that earnings have grown at a ~2.5% real over the long run. As P/E is bounded, growth in E translates to growth in P which gives the capital gains fraction of total return. With a 100 P/E, dividend yield would be about 0.4%. Add this to 2.5% and you get 2.9%, which is about the same as the 2.7% real return from investment grade bonds.

What has happened in the past was a dividend yield of about 3% was added to this 2.5% to give a fundamental return of 5.5%, which they argued was unrealistically high; the market should correct for this excess return by shifting the prices upward. They argued that the ongoing bull market was simply this adjustment taking place and not an asset bubble as others claimed.

Similar arguments from a market efficiency standpoint were used to argue that the high home prices a few years back, or the high oil prices last year were not speculative bubbles. I don't think there are many, even on the right, who still claim that the 1998-2000 stock market or the 2003-2006 housing market were not bubbles.

It's important to note here that since theory construction is a deductive process, the validity of any predictions are dependent on correct premises. If your theory has certain initial assumptions and those conditions don't actually occur then one should expect deviations when you do empirical measurement.

Theory construction is not really deductive. You don't know what the intial premises are (that is what you are trying to find out). Look at the phase separation issue. I am claiming that it will take much longer to settle at large scale because the individual globules of each phase have farther to move in a big tank than in a small one. This distance should scale as the cube root of the tank volume.

But the process could be dominated by hindered settling, in which case the concentration of globules is what slows separation. Settling rate is limited not by movement of globules, but by the rate of coalescence of small globules into a bigger ones. which then quickly separate out. Coalescence rate should be a function of the system conditions, composition, temperature pH etc., but not system size.

Eather one of us could be right. I don't know.

We are not going to go anywhere on this interest rate thing. It is obviously important to you that interest rate and return on capital be the same.

Let's return to the orignal main issue. I said concentrations of wealth occur naturally under the capitalist process. You disgreed, but then we got carried away on this side issue.

Originally Posted by Mikebert

Here's a different line of reasoning. When you talk about low grade bonds, their rate of interest is higher because of the increased risk of default. But when we talk about return on equity we are talking about the stock index: the S&P500 and its precursors, because that is the data that is available going way back. Companies in the S&P500 are investment grade (blue chip) stocks, they don't go bankrupt at anywhere near the rate small caps do. Investment grade stocks should be compared to investment grade (Aaa) bonds. If you want to use Baa rates then you should compare that to returns from small cap stocks, which have given higher returns than the S&P500 over the long run.

OK, now that argument makes some sense -- you're trying to compare large cap stocks with the type of bonds those companies tend to issue. Unfortunately, this doesn't address my underlying argument, which I'm going to have to explain better below.

Originally Posted by Mikebert

If so then why don't you accept the result for the 1871-1920 comparison, when there is much less interference by complicating effects? Here the interest rate I got was 4% (same as they got in that paper) compared to ROI of 7%, a whopping 3% difference.

I'm not sure about their interest estimation in the earlier period. The reason I cited that paper was to demonstrate how apparent differences between debt and equity are actually balanced.

Originally Posted by Mikebert

Yes, but in the case of science the old theory was based on empirical data, that's what makes it science. In economics and other social studies the theory isn't based on data.

Sure they are. One can observe human behavior and extrapolate from those observations. The data is different because the subjects have will.

Originally Posted by Mikebert

For example, now you are trying to use low grade corporate bonds as your proxy for the cost of money applicable to capital investments.

No, I'm trying to include them in the average value while you are trying to exclude them. The use of proxies is dangerous since you have to be sure that such proxies are actually a representative sample of the whole system. Since I'm only arguing that M-M should hold long term for the entire market as a whole, looking at particular segments of it should produce errors one way or another. For example, using only the upper echelons of the economy seems to produce a slight equity premium. The reverse should be true if you used junk bonds and penny stocks. Baa and small cap would probably be close to even (and it is). Determining which of these is closer to the overall average is hard because we don't have any means with which to weight these against each other.

All, I'm saying is that the data you have doesn't conclusively demonstrate an anomaly. What would? If junk bonds and penny stocks produced an equity premium, then I'd have cause for revision. If the equity premium (in the mid-range) was large (say, over 1%), then there would be cause for revision.

Originally Posted by Mikebert

You choose this because they are biased towards what you want to show. The paper's authors started with Aaa bonds, which is a more defensible assumption, but then they had to manipulate the data to get it to show what they wanted.

You're misunderstanding the methodology here. They didn't "manipulate the data" -- the M-M theorem has premises that they know the data doesn't perfectly conform to. Thus, they're attempting to correct for deviations from the ideal case. Think of it sort of like the ideal gas law. It gives excellent predictions, if the gas you are using fits within particular parameters.

Originally Posted by Mikebert

Why is one a moron if they base investment decisions on M-M? Because it does not work? If it doesn't work, then how valid can it be?

It doesn't "work" as a tool for investment advice because its idealized conditions obviously don't exist in the real world -- except perhaps on a long time scale.

Originally Posted by Mikebert

My understanding of Austrians is they eschew empiricism entirely. What they do is akin to what natural philosophers did before there was science. These philosophers derived their theories on reasonable premises and used them to explain their observations. They were all wrong, as later scientific investigation has revealed.

Many of them do so -- because Mises did. Hayek didn't though. It depends on which Austrian school types you're talking about.

The natural philosophers had no other option. Economists have a hard time of it too.

Originally Posted by Mikebert

Similar arguments from a market efficiency standpoint were used to argue that the high home prices a few years back, or the high oil prices last year were not speculative bubbles. I don't think there are many, even on the right, who still claim that the 1998-2000 stock market or the 2003-2006 housing market were not bubbles.

Yes, but you'll also notice that bubble assessments were rampant among Austrians in both of those periods. The Austrians had the proper skepticism about market efficiency theories (as do behavioral economists). Weak-form EMH may exist, but that doesn't mean that you can't get blindsided near term, nor does it mean that markets can't be systematically tricked. EMH just means that all information regardless of quality is incorporated into the price of financial instruments. What if the information is distorted? The Austrian school's credit cycle theory explicitly predicts that credit expansions cause systemic information errors. So EMH, if it holds at all, can only hold on a very long time scale where multiple credit crises are evened out. Since the Modigliani-Miller theorem would only fully apply in a market without information distortions, we can conclude that, if at all, only the corrective process of a long time scale could possibly cause it to hold in the actual markets we have.

That's why I've been particular about how we test the matter at issue. There are many circumstances in which it shouldn't hold true, on its own terms.

Originally Posted by Mikebert

Theory construction is not really deductive. You don't know what the initial premises are (that is what you are trying to find out). Look at the phase separation issue. I am claiming that it will take much longer to settle at large scale because the individual globules of each phase have farther to move in a big tank than in a small one. This distance should scale as the cube root of the tank volume.

All of that is deductive logic. Now you're going to do an empirical test and see if the data conforms to it. Even if the data shows otherwise, that doesn't demonstrate that volume relationships can't have an effect on these sort of processes, just that those considerations don't govern for this particular system. It's similar with economics. The existence of disequilibria doesn't mean that markets don't move towards equilibrium.

Originally Posted by Kurt Horner

All, I'm saying is that the data you have doesn't conclusively demonstrate an anomaly.

This is the point. You are calling it an anomaly. But is it? Why should one believe the Austrian notion that interest and capital return be the same? It's an anomaly for you because you favor the Austrian views on this issue.

Originally Posted by Kurt Horner

The use of proxies is dangerous since you have to be sure that such proxies are actually a representative sample of the whole system.

A proxy isn't necessarily a sample. Consider a pH probe. You stick the probe into a sample and measure the voltage generated by the probe. The pH is linearly related to the voltage. Here the voltage is a proxy for the pH. The way you know you are not getting into trouble is to validate the probe measurement by using it to measure the pH of a similar sample of known pH and seeing how well it agrees with the correct value.

I validated my E/R measurement by comparing it to the NIPA figure and seeing that it gave the same results---not by showing that the index was representative of the economy as whole.

Now the NIPA figures are for the entire economy while my proxy is just for large cap index. Yet the proxy works, despite not having any small companies.

This could mean that the index used for E/P is a representative sample of the whole economy. It could also mean that the portions of the economy not covered by the index are different from the index, but they collectively do make up a sufficiently large fraction of the economy to make their difference affect the average.

If the first case were true, then small companies, who likely face more expensive debt than large companies would have a greater incentive to refrain from borrowing and use retained earnings for investment than would large companies. This would imply a greater propensity of large companies to employ debt (which because of their size would be tend to be high grade) and smaller companies to eschew debt (likely to be low grade). Tis would tend to make low grade a smaller fraction than high grade for that portion of funding not obtained from retained earnings.

If the second case were true, then the sort of companies that would use low grade debt are a pretty small fraction of the economy. In this case too the fraction of low grade debt would be small.

I'm not sure about their interest estimation in the earlier period. The reason I cited that paper was to demonstrate how apparent differences between debt and equity are actually balanced.

It would seem that this data, coming from before the time of taxation, when the Federal government was very small (perhaps a third the size of state and local government) should be the easiest to use for our purposes. Their data agrees well with my data, which I got from:

Smith, Walter B. and Arthur H. Cole, Fluctuations in American Business, New York: Russell & Russell, 1935, p 173-184.

Their stock index is not very good. Like professor Siegel I have an index that goes back to 1802 and which includes some of their data, but I don't use the pre-Civil War data much because of the thinness of the early indices. Nobody seems to have serious problems with the Cowles stock index after 1871. As the bond market was more developed in the post-1871 period than the stock market, I would expect bond indexes to be good in that period.

Why do you have a problem with the data? How come you buy the later interest rates, with all the manipulations, but not the raw data they present from the earlier period?

Is it because they don't agree with what you believe? The authors of that paper post the same rates as I get and use them as evidence in favor of your views. They do so because they have no data for return on equity for the early period and so assume it is the same as the NIPA data from after 1929. Since this post-1929 value is equal to the 1871-1920 raw interest rate, they see this as validation of their beliefs that interest = return on equity. You, having accepted the higher return on equity of that period, now have to question the interest data to get the facts to line up with your beliefs.

Originally Posted by Mikebert

We are not going to go anywhere on this interest rate thing. It is obviously important to you that interest rate and return on capital be the same.

All that I've ever been strident on is the principle that they should tend to move toward equality, and that an economy that did balance in this way would still grow.

We probably should drop this particular line of inquiry though, since we've hit the limits of the data and have still not reached conclusive results one way or the other.

Originally Posted by Mikebert

Let's return to the original main issue. I said concentrations of wealth occur naturally under the capitalist process. You disagreed, but then we got carried away on this side issue.

It's not that much of a side issue, since (I think) you we're using the equity premium as a means of showing why capital markets cause wealth concentrations. But, let's return to post #821 . . .

Originally Posted by Kurt Horner

This has the same assumption that I note above, that management, maintenance and utilization of capital can be done by one person regardless of the quantity of capital. If, instead, there are limits you get a situation like this:

I have 200 units of capital and you have 100. After a year I have earned 20 units and you have earned 10. Both of us spend 5 units on consumption but I have to hire an employee. He also consumes 5 units. The result is a system where there are three persons gaining income from 105 units of capital each and capital ownership does not lead to inequalities of wealth.

It's possible that the diminishing capacity to manage capital kicks in at a level higher than 100 units. That's fine. All that means is that eventually you hit a productivity limit beyond which anyone who lost out in the random walk is cut back in as an employee and differential gains from capital ownership immediately cease.

Another possible failure might be that I only offer 4 units of pay to my employee and use the extra unit to grow my capital faster. There are a number of problems with this theory. First, one has to assume that not only do I only offer 4 units but that owners of greater capital in general only offer 4 units. If workers were interchangeable and no interpersonal factors intervened, this might occur. However, the actual result of offering 4 units is that only desperate people would take that offer. If the labor market is competitive or small amounts of capital easy to obtain, then there are no desperate laborers and as a result I must offer 5 units of pay in order to get competent labor. Failure to get competent labor means either a) I have to work harder or b) my returns will be below market.

Historically, the advent of industrial wage labor occurred as a result of land rights being taken from peasants. In England, for example, the Enclosure Acts created large numbers of desperate laborers who would then take the 4 unit wage since the capital they previously owned had been stolen from them. (No random walk here -- "deliberate stomping" would be more accurate.)

So, here's how it works. Capital accumulation should occur up to the point where true economies of scale (from actual physical and technical considerations) exactly cancel the information advantages of direct control by the operators of the capital goods. Below that point, firms tend to grow and above they tend to shrink. Now if the state systematically insulates firms from the costs of large size then that point of optimum size will move up and the economy will centralize more than is strictly necessary. In addition, this will shift the types of capital goods produced into forms that are appropriate to the increased scale -- making it likely that future observers will see the economies of scale as caused by the technology, rather than the state intervention driving the type of capital being created.

Originally Posted by Mikebert

This is the point. You are calling it an anomaly. But is it? Why should one believe the Austrian notion that interest and capital return be the same? It's an anomaly for you because you favor the Austrian views on this issue.

Because, the tendency of markets to move toward equilibrium is the cornerstone of all economics regardless of school. Applying that principle to interest rates, implies a long-run tendency toward equalization. To refute it, you've got a few options. You could a) propose an entirely new basis for economics or b) explain why interest would be "different" from other prices or c) identify a persistent cause for disequilibria in a particular direction in this market.

You haven't attempted a) or b) and c) is, as noted above, potentially compatible with my position. In fact, even those schools of economics that would be uncomfortable with my argument are all taking option c).

Originally Posted by Kurt Horner

So, here's how it works. Capital accumulation should occur up to the point where true economies of scale (from actual physical and technical considerations) exactly cancel the information advantages of direct control by the operators of the capital goods. Below that point, firms tend to grow and above they tend to shrink.

This isn't capitalism. See below

Now if the state systematically insulates firms from the costs of large size then that point of optimum size will move up and the economy will centralize more than is strictly necessary.

The state doesn't insulate firms. The firms insulate themselves, using the state as the means.
************************************************** **************
The unit of action in capitalism is the capital, not the firm. Just because the great capitalists of the late nineteenth century were associated with huge firms doesn't mean you have to do it that way. A capitalist can amass a fortune by buying multiple firms in special situations, as in my example (and as Buffet has done) although that is slower than growing a firm.

Also very small firms can create large concentrations of wealth (e.g. E-bay, Amazon, Google). Although you have to be lucky to do this.

Once the concentration is present, then the rent-seeking can begin. After all, the easiest way to grow your wealth is to move the size limit on your firms upward so they can grow larger, saving you the effort of continuously finding new firms to acquire in special situations, or the need to be lucky. This is what the nineteenth century capitalists did, but they did so only after they had developed the necessary concentration of wealth to purchase the influence necessary to achieve their goals. Railroad interests were able to get one of their own elected president in 1860, well before corporations were invented. Their invention simply made it easier for the magnates.

Posted because a)it's cool, and b)it tends to support the libertarian/anarchist contention that Hobbes was full of crap, and that people, as social beings, are fundamentally predisposed to cooperate and help each other without being forced to. The contention the article offers that attempts to control the behavior of spontaneously-ordering groups run the range from useless to counterproductive is just a bonus.

Worth reading in its entirety, but I like to excerpt.

Yet studies by Drury and others suggest the bonds that form between strangers in crowds are very robust, and although people might help family members first in an emergency, they will also help others irrespective of their connection to them. "What is really of interest," says Drury, "is why so many people - strangers without any formal organisation, hierarchy or means of communication - join together and act as one."

So where does this inclination come from to empathise so strongly with others on the basis of shared fate alone? Nobody is really sure, though it appears to be uniquely human. As Mark van Vugt at the University of Kent, UK, and Justin Park at the University of Groningen in the Netherlands point out, no other species appears to have the capacity to form rapid emotional attachments to large, anonymous groups (The Psychology of Prosocial Behaviour, published by Wiley-Blackwell next month).

Originally Posted by Justin '77

Posted because a)it's cool, and b)it tends to support the libertarian/anarchist contention that Hobbes was full of crap, and that people, as social beings, are fundamentally predisposed to cooperate and help each other without being forced to. The contention the article offers that attempts to control the behavior of spontaneously-ordering groups run the range from useless to counterproductive is just a bonus.

Worth reading in its entirety, but I like to excerpt.

Thanks, Justin, very interesting article!

Originally Posted by Justin '77

it tends to support the libertarian/anarchist contention that Hobbes was full of crap, and that people, as social beings, are fundamentally predisposed to cooperate and help each other

In the context of crowds. They said nothing about other contexts.

Originally Posted by Mikebert

In the context of crowds. They said nothing about other contexts.

What 'other' social contexts are there? A 'crowd' is the next social base unit up from a 'nuclear family'.
They're talking about the nature of the first emergent building blocks of every social system. That's pretty damn widely applicable.

Originally Posted by Justin '77

What 'other' social contexts are there? A 'crowd' is the next social base unit up from a 'nuclear family'.
They're talking about the nature of the first emergent building blocks of every social system. That's pretty damn widely applicable.

You are overgeneralizing. just because groups of people behave in Virtuous Way X in a physical crowd or large size doesn't mean that will apply elsewhere.

Originally Posted by Odin

You are overgeneralizing. just because groups of people behave in Virtuous Way X in a physical crowd or large size doesn't mean that will apply elsewhere.

Nobody said that (except, to a certain extent, the authors of that study). At this point, all there are are observations and some first attempts to try to explain them.
What's interesting is that the dynamic they have observed appears to be fairly robust -- in the sense that the size and specifics of the focal point of the group does not appear to play a qualitative role, and appears to have only a minor quantitative role.

Of course, anyone observing how people actually behave could have offered a ton of anecdotal evidence along those same lines (and people have, all the way back to Lao Tze). But we live in a world with a strong empirical bias, so a 'study' with 'data' automatically is cooler than just pure bird-watching.

Originally Posted by Justin '77

What 'other' social contexts are there? A 'crowd' is the next social base unit up from a 'nuclear family'.

In the article they were talking about dense aggomerations of people, in which there may be a dozen or more strangers within a few meters of an individual. Most people spend a pretty small fraction of their time in such close proximity to large numbers of strangers.

Right now I am at work. There are 2500 people at our facility all involved in complex projects involving a lot of interpersonal interactions with a common purpose. My workplace certainly qualifies as a social base larger than the nuclear family. And yet right now, and for much of the day, there are perhaps 4-10 people within 10 meters of me, most of whom I know and and most who are separated by walls. This is hardly a crowd in the sense of what they were talking about in the article.

Originally Posted by Mikebert

In the article they were talking about dense aggomerations of people, in which there may be a dozen or more strangers within a few meters of an individual.

Its funny how you seem to have read something into the article that not only isn't anywhere in it (or in any of the research referenced in it), but which is actually directly contradicted multiple times in it.

There are a couple segments dealing with even down to one-on-one interactions in isolation.

Maybe you could point to where you pulled your claim from?

Originally Posted by Justin '77

Its funny how you seem to have read something into the article that not only isn't anywhere in it (or in any of the research referenced in it), but which is actually directly contradicted multiple times in it.

There are a couple segments dealing with even down to one-on-one interactions in isolation.

Maybe you could point to where you pulled your claim from?

The article I read talked about big crowds:

In a study to be published in the British Journal of Social Psychology (DOI: 10.1348/014466608X357893), a team led by John Drury at the University of Sussex, UK, talked to survivors of 11 crowd-based disasters or near-disasters, including the 1989 Hillsborough stadium crush that killed 96 soccer fans, and a free concert by Fatboy Slim on Brighton beach in 2002 that was swamped by 250,000 people, four times as many as expected, and led to around 100 injuries.

The team found a similar pattern of solidarity and cooperative behaviour in a study of the suicide attacks in London on 7 July 2005, which led to crowds of commuters being trapped underground

A good example is the poll tax riots in London in 1990, when protestors from a wide spectrum of backgrounds and interest groups joined forces in the face of what they saw as overly aggressive police tactics.

As Mark van Vugt at the University of Kent, UK, and Justin Park at the University of Groningen in the Netherlands point out, no other species appears to have the capacity to form rapid emotional attachments to large, anonymous groups

And in summary

What are the lessons from all this? One of the most important is that the current approach to managing crowds, which is all about control and containment, can be counterproductive. Police tend to assume that people in crowds are prone to random acts of violence and disorder, and treat them accordingly. But aggressive policing is likely to trigger an aggressive response as the crowd reacts collectively against the external threat. This is why many researchers consider kettling to be a bad idea. "You're treating the crowd indiscriminately, and that can change the psychology of the crowd, shifting it towards rather than away from violence," says Stott. He has found that low-profile policing can significantly reduce the aggressiveness of football crowds, and that if left alone they will usually police themselves.

The whole article is about behavior in large crowds, dense agglomerations of people. Where does it talk about behavior in isolation?

Well everyone knows that Hobbes' moral psychology was wrong, and for good reason too! It's quite simply refuted by general human decency. I suspect that many statists tend to fall back on these shady grounds due to straw-grasping resistance to radicalism, not the merit of Hobbesian arguments. We acknowledge the essential nature of humanity every day; somehow quite a few individuals seem to forget the way things are when attempting to counter the big bad libertarian army.

Originally Posted by Mikebert

The whole article is about behavior in large crowds, dense agglomerations of people. Where does it talk about behavior in isolation?

The article did say thing about dense agglomerations, but it also said (this is where that wholу 'contradicting your take' comes in):

The researchers divided a group of volunteers into two according to whether they overestimated or underestimated the number of dots in a pattern - a deliberately arbitrary distinction. They then told each person that they would be talking to someone either from their own group or the other, and that they should arrange some chairs in preparation. Those who had been told they would be talking to a member of their own group placed the chairs on average 20 per cent closer together than those who had been told they would be talking to a member of the other group

and

a 2005 experiment on English soccer fans by Mark Levine at the University of Lancaster, UK. He found that supporters of Manchester United who had been primed to think about how they felt about their team were significantly more likely to help an injured stranger if he was wearing a Manchester United shirt, rather than an unbranded shirt or one of rival team Liverpool. However, fans who were primed to think about their experience of being a football fan in general were equally likely to help strangers in Liverpool shirts and Manchester United shirts, but far less likely to help someone wearing an unbranded one (Personality and Social Psychology Bulletin, vol 31, p 443). This shows the potency of group membership, and also how fluid the boundaries can be.

That is, behavior that is observed in large groups was successfully experimentally replicated in one-on-one interactions where the 'group' was physically not present at all (or, in fact, was a totally virtual artifact)...

I'm surprised you missed that.

Originally Posted by Justin '77

The article did say thing about dense agglomerations, but it also said (this is where that wholу 'contradicting your take' comes in):andThat is, behavior that is observed in large groups was successfully experimentally replicated in one-on-one interactions where the 'group' was physically not present at all (or, in fact, was a totally virtual artifact)...

The article did not merely "say things" about behavior in dense agglomerations (crowds) it was about behavior in crowds. The point of the article was that group behavior in crowds is more often positive than negative. The traditional view of the "madness of crowds" and the "ugly mob" is largely wrong.

The first experiment you cite shows prejudicial behavior exhibited by an in-group versus an out-group. This sort of in-group/out-group mechanism has been studied extensively in an effort to understand such things as lynch mobs. I recall a study in which students were divided by eye color and found to exhibit prejudicial behavior based on eye color when none had been displayed earlier.

The first experiment shows that groups identification can be based on a competely arbitrary, ad hoc categorization and that they can form rapidly. That is, people are capable of developing in-group sensibilities based on nothing more than their mere random presence in an aglomeration of people.

The second experiment shows that prejudicial behavior is not necessarily negative (e.g. lynch mobs), as was the focus of earlier work. It can be positive (giving aid preferentially to in-group members).

When you combine the two experiments together a hypothesis can be formed that says the mere presence in a crowd during some sort of diaster can produce in-group prejudice that makes people in the group more likley to co-operate with each other. That is the "saneness of crowds" should be a more frequent occurance than madness, and that ugly mobs should be the exception. The rest of the article was about examples of actual behavior in large crowds that showed these properties.

That was what the article talked about, behavior in crowds.

Apparently you wish to take this sort of behavior (and things like pedestrians crossing the street in an orderly fashion) to argue that because humans spontaneously create order, their societies do not need government.

That is quite a leap.

************************************************** ****************
There is a much more basic issue that you need to resolve before any sort of argument along these lines is more than wishful thinking on your part:

For the vast majority of the time humans have lived on Earth, they have lived in small egalitarian bands that lacked government. That is, ungoverned human societies are the norm for our species.

However, almost no humans live in ungoverned societies today. If the ungoverned society is superior to governance, how did governance ever get a foothold, much less completely replace, the tens of thousands of ungoverned societies that existed all over the world 10-15 thousand years ago?