Was Hitler a vegetarian?

I always heard loose talk about Hitler being a vegetarian. Personally, I have no idea if he ate meat or not, and I don't think it's particularly relevant to anything, but I did come across a reference to what might be an authoritative source speaking on the matter. On page 41 of the book Entropy and the Magic Flute by Harold J Morowitz, he refers to an article:

An article translated from the German publication called Die Wiesse Fahne volume 14, 1933, attributes the following quote to Dr. Goebbels: "Adolph Hitler does not drink alcohol or smoke and, moreover, is a vegetarian." The unsigned article goes on to say:

"Brother National Socialist, do you know .. that your Fuhrer - out of persuasion and out of love for the world of animals - is a vegetarian ... ?

"... that your Fuhrere is the strongest opponent of any form of animal torment; especially of vivisecton ... "

This is obviously Nazi propaganda. Everything Goebbels said was Nazi propaganda. After all, he was the Nazi minister of propaganda.

So, there it is, on the internet, a suitable reference. Now, all you need is the appropriate volume of Die Weisse Fahne. That's The White Flag, though Google's translation system has some trouble with it thanks to the double ess.

Slave Labor Can Cost More Than Free Labor

As I noted recently, I've been reading Lebergott on Manpower and Economic Growth, and he pointed out something that at first glance appears unlikely, slave labor can cost more than free labor. This seems odd. After all, slaves do what they are told, or else. Why should they cost more for a job than someone who freely contracts with you?

The reason is rather interesting. Slaves had collective bargaining. We tend to think of slaves as being tied to a particular plantation and not as part of a market economy. Sure, slaves could be bought and sold, but the real money was in hiring them out. Slaves cost money, and money meant an opportunity cost. If you had fields of cotton, it might make sense to have your slaves cultivate it, but it might make more sense to hire them out to a brickyard. This was intensified when cotton cultivation started moving west to Louisiana and Texas where productivity per acre and per slave was much higher than in Georgia or South Carolina. It even made sense to hire out your slaves to someone in another state, even if it meant having your own fields lie fallow.

Consider how this works. Someone runs a plantation or a sawmill or whatever and needs labor. They can put a sign to hire free workers and negotiate with each worker separately. If one worker is too greedy, they can just hire another. Free men, in some ways, are even cheaper than slaves. Getting fed is that worker's problem, not his ex-employer's. Alternatively, and especially if they need a lot of workers, they might negotiate with a slave owner and hire an entire team, complete with an HR department with its whips and manacles. Of course, there are fewer teams of slave to hire than individual free men, so their seller has more market power. They have to take the whole team or leave it. While using slaves might be simpler, they might wind up paying more per hour of labor than if they had built a team of free men.

This sounds rather fantastic, but Lebergott has the numbers to back it up.

In a way it makes sense. An individual worker is at a disadvantage when hired by a large enterprise. One way of fighting the capitalist collective is to form a workers' collective, and negotiating a contract on an enterprise by enterprise basis, or even better, on an industry wide basis. These workers' collectives are sometimes called unions and sometimes called contractors or subcontractors.

The difference between a union and a contractor who provides a class of laborer for an enterprise is who gets to eat the surplus cost produced by the collective bargaining. Union dues tend to be rather modest when compared with salaries. Contractors tend to take a big chunk off the top, and while some of it goes towards benefits, a lot of it goes for executive salaries, return on investment and so on. While companies tend to resist unions, they have no problem with contractors, at least not on a philosophical basis.

There was nothing in this system for the slaves. They weren't paid. They just worked.

Your Share Of The GDP

There is a general feeling nowadays that economic growth has nothing to offer the average American. Yes, the gross domestic product (GDP) has been growing, but no one feels much better off. There are a lot of ways to characterize this disconnect. I've chosen to look at one simple ratio, your share of the GDP. The simplest form of this ratio is the ratio of the GDP to the population, but it is more enlightening to consider the ratio of that GDP share to the average worker's wages. Does working hard and playing by the rules get you more of the GDP or less of the GDP than it used to.

---

The Data Sources

First, we needed a source for the gross domestic product. A bit of searching brought us to a rather useful data source, a web site called , and particularly to their GDP database access page. This provided not only the GDP from 1790 to the year 2000, but the population as well. We augmented this data with a few numbers from the Census Bureau and the Bureau of Economic Analysis to take our numbers through 2007.

It was harder to find a good source for earnings data. The Census Bureau does provide such data, but is surprisingly coy about it. Consider the page below from the 1910 statistical abstract. It provides wages and hours, but only as percentages of relative to the baseline average of wages and hours from 1890-1899. Curiously, they do not provide the actual baseline numbers.

Later census statistical abstracts aren't quite so co. May include a useful table of average annual earnings of full time employees. These usually have an overall average and the state by state averages. Unfortunately, this table is not always present. Luckily, in honor of the bicentennial in 1976, the Census produced a wonderful set of historical statistics. This is actually available as a download, though we originally owned a print copy. This was two oversized volumes which provided us with many happy hours of browsing. Did you know that eggs cost nearly a dollar a dozen in the late 1800s? That put the 1970s inflation into perspective.

One chart we found was series D 722, 723 and 724, Average Annual Earnings of Employees: 1900-1970. Given the various means used in obfuscating earnings data, it was quite surprising to find such a useful chart. As you can see below, this chart does have its defects. Series D 722 runs from 1929 to 1970 while D 723 runs from 1900 to 1960. No one series covers the entire time period.

This called for some splicing. We mainly used the D 722 series, but computed a correction factor based on the overlap from 1929 to 1960. This let us adjust the D 723 series to cover the period from 1900-1928. This is methodologically suspect, if not actually bogus, but it gets us a smooth series that generally reflects the actual ups and downs as experienced by people in the workforce. This is the golden age of mashups.

Getting from 1970 into the current century was fairly easy once we noticed a URL in a more recent version of this table in one of the Census Statistical Abstracts. This led us to the Bureau of Economic Analysis which has replaced the Office of Business Economics noted on the table above. They had all the data we needed in four separate series 6.6A through 6.6.D. According to the notes in the 1976 census blockbuster, the historical D 722 series was developed by Lebergott and reflects the earnings of a full time employee as reported by a statistically weighted cross section of employers. The related series D 723 and D 724 reflect lower wages due to unemployment, so our analysis is based on the optimistic assumption that full time work is available.

---

Annual Earnings as a Share of the GDP

The chart below plots [earnings / (gdp / population)]. It shows how much of the economic pie you are likely to get by working full time.

The chart has its ups and downs, but they are not all in the places that one might expect.

From 1900 to 1970, a worker's share of the GDP varied around 1.5. If everyone got exactly a fair share, as in some theoretically egalitarian state, this ratio would be about 1. This reflects the fact that not everyone worked for a living, at least not in the cash economy. Young children, pensioners, the disabled, and most married women were not in the workforce.

The ratio was higher during the Roaring 20s. That was a period of rapid industrialization, rising living standards, urbanization and increasing productivity. Historians still argue whether this was a result of the the new income tax or the outlawing of alcohol.

The ratio skyrocketed during the Great Depression. If the Roaring 20s increased the value of work, the Great Depression validated it. If you managed to have a job, you were getting a lot more of the GDP pie than during better times. Of course, the GDP was smaller. "Nice work if you can get it", as they used to say.

The Great War (World War I) and World War II lowered the ratio. War time is usually considered boom time with massive government spending and jobs for everyone. Unfortunately, war time spending is not all that efficient. While the GDP increases, a lot of that product is in the form of military victory which is hard to carry on a balance sheet.

After World War II, the ratio rose into the 1960s. This wasn't exactly the golden era, but the GI Bill had created a new generation of educated workers, American industry ruled the world, and unions were powerful in and out of the workplace. Many saw the 1960s as an echo of the 1920s. There were cultural and economic parallels between the Roaring 20s and the Go Go years, but the resemblance was superficial.

In the mid-1960s the ratio began to fall, and it has fallen since. The Vietnam War may have been one cause, but there was no post-war recovery in the mid-1970s. It is interesting to speculate about the fine details of the curve in its descent, but they are all fine details. The oscillations of earlier years are gone. The recession of the early 1980s barely registers on this chart. Neither does the recession of the early 1990s. The Roaring 90s were marked by a mere slowing of the decline, and the Bush years as a slight increase in the slope.

The ratio has dropped from around 1.5 to a bit over 1.0, a drop of about a third. It doesn't quite take two people working full time to live as well as one did during the 1950s and 1960s, it takes about one and a half. Our living standard may have improved, but the value of a day's work as a share of all goods and services has fallen.

The share of GDP has been less volatile since World War II. Perhaps the most intriguing result is the relative stability of the ratio since the 1950s. There are ripples and wiggles, but even the troughs of recession and spikes of recovery have been muted against the overall trend. Since we are looking at the earnings of full time work, this may because more workers are part time. It also may reflect increased government intervention in the economy serving to dampen the swings.

A CLOSER LOOK

It is worthwhile taking a closer look at the data. Were there dramatic changes in population or GDP? Consider the chart below.

The orange line is population (in units of 10,000 people for the purposes of charting). We can see the population rise early in the century, a slow down during the Great Depression, a sharp rise during the Baby Boom, and a return to a more ordinary regime in the 70s. Population is cumulative. Most of the baby boomers are still alive, so the dramatic rise in population after World War II might play a role.

The green line is the GDP (in billions of nominal dollars), and the GDP rises dramatically throughout the era. Naively, it appears that the GDP rises exponentially, though this may be an artifact of inflation, rather than increased production. Since we consider wages in nominal dollars, we do not need to adjust earnings for inflation. The issue is not standard of living, but rather of relative value.

The blue line is GDP per person (in nominal dollars). While Malthus predicted exponential population growth and arithmetic economic growth, we are seeing arithmetic population growth and exponential economic growth, so the exponential GDP swamps the arithmetic population curve and each person's share of the GDP rises exponentially, assuming it is divided evenly.

The purple line shows annual earnings (in nominal dollars), and it appears to grow exponentially, at least partly an artifact of inflation. This number may appear suspect. In 2006, annual full time pay averaged nearly $47,000, much more than one would earn working 2,000 hours at a $14.50 median hourly wage. As noted earlier, this is an optimistic analysis, assuming that full time work is available.

The red line is annual earnings as a fraction of the individual share of the GDP (multiplied by 10,000 for the purposes of charting). It is the ratio of the purple line and the blue line, and we have traced its rise and fall. There is no obvious explanation for this ratio in this chart. The blue line (GDP per person) and the purple line (annual earnings) seem to rise in tandem. They are farther apart in good times and closer together in hard times, and it is this subtle difference that shows up in the gyrations and descent of the red line.

ONE EXPLANATION: WORKFORCE SEX

One explanation for the declining ratio of earnings to GDP share might be the increasing number of women in the workforce. Children, old people, pensioners, the wealthy and, until late in the century, women were not expected to be in the workforce. If a greater portion of the population is working, then each worker is likely to take home a smaller share of the GDP.

The following chart, based on Bureau of Labor statistics, shows the percentage of working age men and women in the workforce since 1948.

Men left the workforce while women joined. The percentage of men working fell by more than 10%, while the participation rate for women nearly doubled from the low 30s to nearly 60%. The chart below combines the two series to provide a closer look at overall workforce participation.

The late 1950s and early 1960s were the low point. The baby boom ended in 1964, and the chart shows women going back to work, but the big changes didn't start until the late 1970s and it continued through the 1980s. The participation rate rose from around 56% to around 63%. The ratio of the participation rates is 8 to 9, so we could account for a inverse fractional decline in GDP share per worker of about 1/9 or 11%, not the roughly 1/3 we have observed.

The participation rate has been rising since the early 1960s, when it was below 56%, but it only rose to 58% by 1970. Still, we can see an increase starting in the mid-1960s, the end of the baby boom, and continuing into the 1990s. There were recessions and set backs, but the percentage of the population in the workforce rose and continued to rise throughout this period of decline in share of GDP. Even more interesting, the rise in workforce participation seems to have leveled off in the 1990s, varying in the low to mid-60s.

ANOTHER EXPLANATION: WORKFORCE WORK

Another explanation to consider is the change in who does what. The chart below shows the percentage of workers in each major employment sector. It is based on the 1900-1970 Series D 127-141 of employment by sector, combined with the BLS 1939-2007 data from http://www.bls.gov/ces/cesbtabs.html, their Series B workforce by sector dataset, which uses a similar, but not quite identical categorization. (In other words, this was yet another series I had to build by judicious splicing).

The chart is cumulative, so that the each share is represented as the distance between two lines, and the sum of all the percentages is representated by the teal line across the top at 100%.

The blue line at the bottom is mining. The distance from the blue line to the purple line is share of workers in construction. This number has fallen, though the building boom of the 2000s is visible. The distance to the yellow line represents the share of workers in manufacturing and the distance from that line to the dark purple line is the share in services. The percentage of workers in manufacturing has fallen dramatically since mid-century, but the percentage of service workers has increased to take up most, but not all, of the decline. The distance to the red line represents the share of workers in trade, retail and wholesale, and the distance to the green line the share of workers in the financial sector. Neither of these sectors have changed much since mid-century. The distance above the green line represents the percentage of workers employed by the government, federal and local. This percentage peaked in the 1970s and has declined since.

We can see both structural changes, particularly between the start of the century and mid-century. The nation was nearly half rural in 1920. Still, the overall trend has involved a decline in the manufacturing sector, with a long continuous decline since the end of World War II. In terms of GDP, American productivity has not declined over this period, despite rising population, but earnings as a share of the GDP have.

Where has the rest of the GDP gone? Why doesn't a full time job give one a bigger share?

Some may have gone to Social Security and other retirement payments. People born after the Great War were not only eligible for Social Security, but they were also likely to have government insured private sector pension plans. Many have also saved money for retirement. The economic condition of retirees has been much better since the 1970s when compared with earlier eras, and people have been living longer so the nation has aged.

Some may have gone to the corporate sector, with higher corporate profits, higher dividends, increased capital gains and greatly increased pay in the executive suite. The distribution of goods and services is largely political, and the corporate sector is more politically powerful now than it was mid-century. It is now acceptible to begrudge a worker his health plan or three weeks of paid vacation, but not his boss's ten million dollar salary or hundred million dollar bonus.

Whereever the rest of the GDP growth has gone, most Americans realize that they are not going to get all that much of it by working at a full time job. Worse, full time jobs, without layoffs or reduced hours are increasingly hard to find. Part time workers get even less of their share of the GDP. Well into the 1960s, it was possible for a worker to conflate his or her own financial well being with such an abstract concept as gross domestic product. In the new century, any such worker would be considered delusional.

Some Lebergott

I was reading some Lebergott the other day, in particular his Manpower In Economic Growth. It all started back at the bicentennial in 1976. The Census Bureau had put out a special commemorative two volume edition of statistics for the first 200 years of their operation. It was a wonderful, if massive, tome. It had population, incomes, production, and so on. How much did a farmhand make in 1880? What did a dozen eggs cost in 1925? It was all there. Then I moved, and I just couldn't move this monster with me. Then, the Census posted the two complete volumes on the internet. It's all image scans in PDF, but there it was. I was reunited with this magnificent time waster.

So, I started looking at incomes because Americans have been working as hard as ever, but they haven't been getting paid more for it. This, despite rising productivity. There was a wonderful set of series, D722-727, giving the incomes of full time employees from 1900 to 1970. I was fascinated. It was better than Ripley's believe it or not. Did you know that real income had risen four-fold from 1900-1960? Naturally, I had to learn more about these wonderful numbers, and here again, the Census came through. At the bottom of the table was a footnote and the name Lebergott.

I've been wading through Manpower In Economic Growth for a bit now. Lebergott actually tried to get some idea of what Americans were earning back in the late 18th and through the 19th century, but it's all pretty sketchy. Slaves, for example, weren't paid, and a lot of people were paid in kind. If you were a farmhand, you might wind up with some bushels of wheat, and you had to sell or barter to get them ground, buy clothing, pay your rent, and buy other stuff. In other words, you were part time farmhand, part time wholesaler and job lotter.

Anyone who can write "The results are unequivocally confusing." is someone who takes the data seriously. Too many people look at the raw data and don't get confused properly. One thing he pointed out was the Adam Smith didn't understand America all that well. Smith went on at length about the importance of specialization. If you've read any economics, you know the parable of the pinmakers, where breaking this task into a series of specialized subtasks could produce marvelous productivity gains.

Well, America, especially in the early 19th century was full of generalists. Even your farm hand might have to do a stint wholesaling, and the typical American career might include terms as a shop keeper, a farm, a rail splitter, a teamster, and so on. He gave one example of a mason putting down his trowel, spending some time working in a drug store, and then setting himself up as a doctor. America didn't have Europe's castes and guilds. You could be anything you wanted, as long as you kept the customer satisfied.

On page 121 Lebergott goes so far as:

"We actually speculate that a dominant, significant force making for great productivity advance was the very absence of division of labor."

Wow! So much for Adam Smith.

"What, after all is characteristic of the effective division of labor? It is that there is one preferable way and, finally, the approved and only right way."

Specialization works in highly mature industries, but when circumstances change, and they were quite different in the new world as opposed to the old, someone has to figure out a new "best way" of doing things. In Europe, the work would have been done by specialists, but in America it was done by amateurs. Some would botch the job. Most would just muddle along.

"But a significant group took an entirely different approach to the work - because of a difference in perspective, laziness or sheer genius. It was these who made major technological discoveries, who made substantial departures in the organization of work and enterprise."

Not only did America listen to the "aggressive deviant", but America was full of people from all over Europe. In England, there was the English way. In Germany, there was the German way. In America, there was no one American way, so many ways would be tried, and the best ways chosen.

I always like an economist who looks at the data, and looks beyond the dogma. Adam Smith was not wrong as a rule. Much of his work was insightful, as when he notes that on occasion an invisible hand might lead individual profit seeking to benefit the common good. (Smith never represented this as a rule, only as a possibility). Adam Smith was writing early in the American experiment. The real disruption was only beginning. I'm inclined to believe that he would have found much to appreciate in Lebergott's analysis.

The Magic of the Gold Standard

A lot of people think that the gold standard is magic. They are probably thinking about some ancient era when all we had was gold, and no one had credit. In the 1930s, the rule in the United States was that for every $25,000 gold bar the federal government had in reserve, they could issue $71,000 in reserve credit to registered banks. That's nearly a multiplier of three. Those banks, could then issue a combined credit of about $525,000, based on their $71,000 worth of reserves which were, in turn, back by the $25,000 chunk of metal sitt2ing in a vault somewhere.

That's an effective multiplier of about 21. But where did that come from? That's easy, 21 is roughly the product of 2.84 and 7.39. Where did those numbers come from? That's easy, they made them up. The magical, wonderful, natural, organic gold standard was based on bars of gold and made up numbers. I suppose the financial community raised a ruckus whenever one of those numbers got adjusted one way or the other, but they raise a ruckus now.

I'm not exactly sure of how the the gold standard was all that different from what we have nowadays, except that we've dropped the gold bar part.

Balancing the Budget

We already know what a balanced budget looks like. We had a surplus under Clinton, and now we have a deficit under Bush. Part of this is the result of Bush's big spending programs: the War in Iraq and his failed tax cut stimulus program. That's right. We know the zero point. The only reasonable way to think of the tax cuts is as spending.

Credit Worries - 1936 Revisited

I was recently reading a article about the Chase Bank in a 1936 Fortune magazine. 1936 was in the middle of the Great Depression, and the management of Chase was worried about there being too much credit. If nothing else, the Depression was not caused by a lack of money. Banks like Chase weren't making all that many loans, and they were parking their money in treasury notes paying a fraction of a percent. There was money, but it wasn't moving much.

Their big worry was the $30,000,000,000 in loan making capacity that the various banks had on their books. The entire Roaring Twenties boom had been fueled with reserves of only $14,000,000,000, so having double this amount, just sitting around and looking for bubbles was worrying. Unlike modern bankers, Depression era bankers had had enough bubbles for a lifetime and were hoping for a more stable economy.

So, how much is $30 billion? The GDP in 1935 was about $73 billion, so it was about 40% of the GDP. That's about $5 trillion today, but we have a lot more people in the United States, so we probably have an equivalent number at about $2.5 trillion. That's still a lot of money, but what is striking is the ratio between the two numbers. The ratio is about 80 to 1. A nickel hot dog would run about $4 and a $500 new car would run about $40,000 today. A factory worker making 50 cents an hour would be earning $40 an hour today. You might find yourself paying $4 for a hotdog or $40,000 for a new car, but there aren't a lot of factory jobs, or jobs of any kind, paying $40 an hour these days. Obviously, our economy has restructured a bit.

Of course, there was no big bubble in the 1930s. The economy actually lost ground in 1937. What happened to that $30 billion? It stayed in the banks invested in Treasury notes paying less than 1% per annum. If nothing else, FDR could borrow cheaply. Basically, no one wanted to take a risk and lend money to anyone but the most reliable borrower.

There was investment in the 1930s, but it was fueled by cash flow and, now and then, by the sale of stock. Unlike boom periods, when stock prices are linked to earnings, during the 1930s, stock prices were linked to liquidation value. This made it hard to raise money, but if you bought stock at FDR's inauguration, assuming you had some money to do so, you would have done well.

Are we likely to get into such a situation today? Probably not. In the early 1930s, there was no Federal Reserve to cushion the economic collapse. If the economy took you down, you went down. On the other hand, don't expect high yields on your CDs for a while.