Consider the theory: 2x > 0. I would say x, may not be positive, thus the theory is wrong. Any non-positive number is a counter example. All I needed was one to disprove the theoy, and I’ve found an uncountably infinite amount.

“QED also requires that you don’t commit the informal fallacy of wrong conclusion: the existence of an exception has little bearing on the likelihood of an outcome.”

There is no likelihood of an outcome as there is no probability distribution. You’d have to make some assumption about what x’s can be chosen with what frequency for me to tell you something about likelihood.

As for your example, I choose E) nothing can be said about likelihood. This is akin to asking, given 1000 randomly selected math problems what is the probability the answer will be positive. It’s a meaningless question, because there is no specified distribution from which we are drawing math problems. If we went to 1000 people on the street, almost all of the math problems would give positive answers. But that tells us nothing about the distribution of problems more generally. In the same way, what is the distribution two people can exist over all possible states and relative frequencies. It is a meaningless question.

The better way to proceed is just to put your assumptions forward first and make sure they prove your argument. For instance, Ricardo assumed two countries, within countries people were identical, across the two countries there were differences. Specifically, one country has an absolute advantage in wool and wine, but the other has a relative advantage in wine. Both countries have full employment and there are no transaction costs or transportation costs across countries. From here, he derives comparative advantage.

It’s an important theory that has earned Ricardo very deserved status in the profession. Does it actually describe the majority of international trade? No, it doesn’t. Under Ricardo’s theory trade comes from differences between countries, and countries specialize almost completely. In the real world, trade occurs most commonly between similar countries trading similar goods (US and Europe selling cars to each other) or countries geographically close (US and Mexico/Canada). So new trade theories evolved that better explained the real world. Does that mean comparative advantage has no effect? No, of course not, it just means its effect is in the context of other factors driving trade.

Okay. Your last reply persuades me that you are not being disingenuous.

“The status quo may not have productivity growth, thus preserving it will not lead to increased productivity. QED.”

Unfortunately QED requires that you don’t make a formal deductive error: from “may” not have growth one can only deduce “may” not lead to increased productivity. QED also requires that you don’t commit the informal fallacy of wrong conclusion: the existence of an exception has little bearing on the likelihood of an outcome.

Let us not forget that my assertion is that the expected result, all else being equal, of adding people to free trade, is an overall increase in productivity, compared to the nontrading state. As a simple illustration, I gave you the case of just two people.

We are, of course, talking about populations of people. So let me try this approach.

You have 1000 experiments each of which is simply as follows: Two people who are currently self-sufficient discover one another. They each are given the choice about whether or not they will trade with each other.

Let’s be clear:

1. The choice is left up to the individuals. We are not assuming any particular choice per se. We are assuming only that they are humans, acting as we expect humans to act.

2. The environments for the individuals may vary in a random or even haphazard way across experiments, but the environment immediately pre- and post-choice within each experiment is unchanged. Only the effects of the choice cause any change.

3. You don’t know the environments in which the experiments are taking place except to the extent the experiments are possible (the individuals have been self sufficient).

Now which of the following do you think is true?
A) It is most likely overall productivity would increase.
B) It is most likely overall productivity would decrease.
C) It is most likely overall productivity would not change.
D) Two or more of the above are equally likely.

“Okay, so if you think people usually choose in their best interests, how can you think the choice to maintain the status quo unless a mutually beneficial alternative exists, does not on the whole lead to increased productivity?”

But here is the simplest way to show that your argument does not follow.

The status quo may not have productivity growth, thus preserving it will not lead to increased productivity. QED.

If you’ll notice, the texture of the discussion thus far has been, you posit a theory, I produce a counter example, you not understand and rinse and repeat.

If you lay out the argument, you’ll see that it requires many assumptions that may or may not hold. This is a perfect example of why modern econ is done rigorously with math and formal proofs. What “sounds true” does not necessarily logically follow.

Okay, so if you think people usually choose in their best interests, how can you think the choice to maintain the status quo unless a mutually beneficial alternative exists, does not on the whole lead to increased productivity?

BTW, the answer is as bargaining or transportation costs rise trade will decrease and could stop between the two individuals (often called autarky in economic models). In other words, costs affect the equilibrium, BECAUSE people persue their self-interest.

I think you didn’t read my whole post. My point, once again, is that they are not worse off, if they do not do it. It seems your confusion is a little deeper than I thought. Do you know what it means to have a choice? Do you think that when people have a choice, they generally choose for, or against, their personal best interests?

You seem to not understand that many different forces can be acting on something at one time. Consider an object traveling at a certain force in one direction. If the object travels into the wind, it will travel slower. If it travels with the wind, it will travel faster. The wind could be a large or a small effect depending on how fast the object is traveling and how aerodynamic it is. Does it make any sense to say, “oh, you say the wind can move objects, but I threw a baseball into the wind and it went just fine. So where is this wind thing?”

“”So we could ask, under these assumptions what happens as bargaining costs rise holding all other things equal?”

The answer: Almost certainly nothing, since people usually don’t choose to act in ways that make them worse off.”

Again, you aren’t thinking about the assumptions you are making. If trading gains each X units of welfare, but the bargaining costs or transportation costs are much higher than X, then they aren’t better off trading. They’re worse off.

There could be a huge population of Martians with a vast supply of oil or diamonds with a huge demand for salt water, and we still wouldn’t have any gains from trade with them, because the transportation costs would be so high.

The answer: Almost certainly nothing, since people usually don’t choose to act in ways that make them worse off.

You seem to forget that each party has a veto–each can choose to continue as they were, or to proceed with cooperation. That is, after all, what is meant by “free trade”. You have to wonder why they would’ve chosen cooperation if they thought they would be worse off.

It is not a matter of whether or not a counter example can be imagined, but given these circumstances, how likely do you think it would be that people would make a choice to be worse off than they already are?

I find it very interesting that you take this stance. What are your thoughts on comparative advantage? Do you think it is a false theory? Do you think that specialization contributes to wealth? What do you think is the ultimate source of wealth creation? What is the source of the unprecedented wealth creation in countries with a greater capitalist presence?

It seems we have found a point of fundamental disagreement between us. I’m looking forward to your answer, but I think that will be end of the discussion as it seems the walls are closing in on this discussion rapidly.

“No, but humans are probably dumber than they would be if birth rates were equal across incomes.”

So your dumbing down rule works sometimes but not others? During times of bad nutrition it doesn’t work, otherwise it does? Human evolution 100K years ago caused it to not work then, but to work now?

From 1854 – 1914, RGDP/capita annualized growth rate is 1.36%.

From 1915 – 2008, RGDP/capita annualized growth rate is 2.21%.

Are you sure you weren’t using RGDP and not RGDP/capita? If you have different data from NBER, please let me know, I use this site frequently and would like to know if it’s wrong or disputed.

The text makes it sould like you just used RGDP, “the US economy has actually grown less rapidly since 1914 than it did before.”

You are making several assumptions to get to your conclusion. I’m attacking them. That is not something you can say ceteris parabis to and make go away. For instance, in your example, what if bargaining costs are very high–it takes time to haggle out a deal. Maybe one person is even deaf! What if they have to work on different parts of the island (maybe that’s where the coconuts are) and there are substantial transportation costs (coconuts are heavy!)

Ceteris Paribis is holding all other things equal. So we could ask, under these assumptions what happens as bargaining costs rise holding all other things equal?

On the other hand, I’m attacking the theory. I’m attacking the assumptions. The point is that the theory is not self-evident like you seem to think it is. It rests on several assumptions that may not be true.

“So you have that increasing productivity countering the effect of a growing proportion of dumb people. I don’t see how you can tell which is the more influential factor.”

I can’t, that was my point. The effect on my example could be offsetting the effect of your example. The point was your model makes simplifying assumptions that may not be true.

“But a more obvious point is–do you really think that humans have been getting progressively dumber over the last 100,000 years?”

No, but humans are probably dumber than they would be if birth rates were equal across incomes. There are obviously several confounding variables that push it the other way. One is evolution. Another (on smaller time horizons) is nutrition.

“But it should be cumulative. We didn’t lose the shovel when we found the steamship.”

But realize we are measuring growth of real GDP/capita, which is on a positive trend. Suppose an economy has no growth of any kind. The growth rate is 0. GDP is 100. Population is 1. Enter new technology. Shovel that increases GDP to 110. The growth rate is 10%. But the next year, the GDP is still 110, and the growth rate has fallen to 0. The insight is that when technological change is causing the growth, one needs change to continue to keep getting growth. If you have lots of great inventions one year and fewer in the next, the growth rate goes up and then down.

YOU: Again, no it doesn’t. I gave you several potential counterarguments.

You gave me several confounders. I hate to take this back to such an elementary level, but that basic level was my point: Two people on a desert island who don’t know each other exist are trying to live. After a few years they discover one another and decide to work together for mutual benefit. You don’t think their production would stand to increase?

“Now if technology is constant, the RGDP/capita will change from the average of the incomes of smart and dumb to just the income of the dumb people.”

I think I see why we are talking past each other. You don’t seem to believe that cooperation and specialization increases productivity. My point was merely that if you go from 100 dumb people trading to 10000 dumb people trading, you would still expect RGDP/cap to increase. Maybe not as much as if those people were smart, but it would still increase. So you have that increasing productivity countering the effect of a growing proportion of dumb people. I don’t see how you can tell which is the more influential factor.

But a more obvious point is–do you really think that humans have been getting progressively dumber over the last 100,000 years?

“Don’t confuse technology with things that use microchips. The first shovel increased productivity much…”

But it should be cumulative. We didn’t lose the shovel when we found the steamship. And population growth shouldn’t detract from it–a shovel is no less useful to one person just because 100 other people have it. Whether technology growth after 1960 was more or less than before (I still suspect it was more), it most definitely should not have resulted in a fall, or even stagnation, of RGDP/cap.

You gave examples of setbacks in the post-1960 era, some of which I think ring true–particularly bad monetary policy. But those examples bolster Selgin’s point about the negative effect of the Fed on economic growth.

Again, no it doesn’t. I gave you several potential counterarguments. You concede the governance point, so I’ll move on to the other two.

The IQ point: The reason I put “ceteris parabis” in quotes was to emphasize that the argument holds everything else constant.

So this point: “The IQ effect you mention should always be at play. So productivity growth should not happen over human history, but of course the opposite is quite the case.” Isn’t valid, because technology growth was being held constant.

Try a thought experiment. Imagine there are two types of people smart and dumb. Only the dumb people have babies. The babies are usually dumb but sometimes smart. Obviously, over time there will be more and more dumb people than smart people. In fact the percent of smart people in the economy will approach zero over time. Now if technology is constant, the RGDP/capita will change from the average of the incomes of smart and dumb to just the income of the dumb people.

Does that mean you can’t have RGDP/capita growth if all things aren’t held equal? Of course not, you could still have technology based growth, and the technology growth could trump the IQ factor.

“Technology has been exploding in recent decades, especially since 1960. It should have the opposite effect that you describe.”

Don’t confuse technology with things that use microchips. The first shovel increased productivity much more than the first iphone, the first steamship and the first locamotive much more so than the first spaceship.

Technology is something that converts hours worked into GDP. It’s a stylized fact of the business cycle that hours worked per capita doesn’t change much over time, thus RGDP/capita should come close to mirroring productivity increases over time. But while that holds true in the post-WWII era, I don’t know if it holds true over the very long horizons we are considering.

Just remember all of these factors are happening all at once in the data.

1992 – 2000 was a technological boom time, with a lot of productivity growth. (RGDP/capita 2.49%)

1980 – 1992 had some financial turmoil, a major real estate bubble burst, not so much productivity growth (RGDP/capita 1.94%)

1972 – 1980 had large negative oil shocks, bad monetary policy, less productivity growth (RGDP/capita 1.86%)

And 2000 – 2008, it drops considerably to 1.21%, this is driven by the tails, 2003 – 2005 was 2.36%.

I just can’t tell you how significant that is to me. ;->

Naturally there are many real and potential confounding factors. The c.p. excludes everything but the simple notion of a growing population of free traders. Certainly that would predict a growing RGDP/cap. One then has to ascertain the effects upon that growing base of those factors that aren’t really being held equal.

I have to say, though, that I don’t quite understand the point of declining IQ and technology.

Technology has been exploding in recent decades, especially since 1960. It should have the opposite effect that you describe.

The IQ effect you mention should always be at play. So productivity growth should not happen over human history, but of course the opposite is quite the case.

Additionally, you might expect that if IQ were dropping, then, c.p., productivity would fall. But the productive effects of trade exist to some degree regardless of IQ. So, with a growing trading population, it is not clear which effect would dominate.

Your point about political boundaries is a good one. Perhaps it is such restrictions on free trade that dampen RGDP/cap over time.

But it would appear that RGDP/cap really is dropping, so you are right that SOMETHING must be doing it. I just don’t see how the IQ and technology arguments can explain it.

I feel the need to point out at this point that this is your theory and not something the economics profession sees as generally true. In general, large countries don’t grow faster than small countries. In fact, population growth and growth rate generally have a negative correlation. It should be obvious that there are many counterarguments to your simple theory. For instance, governance may be difficult as countries get large (institutions more difficult to form). Also, there is the complicating factor of trade. Does Florida gain increased specialization or innovation if many people move in from Georgia? Probably not, as there is already a lot of trade between the states. And here is one adapted from a Bryan Caplan argument. Income is positively correlated with IQ, people with low IQs have more babies than high IQs, IQ is hereditary. So over time, the percentage of low IQ people increases, “ceteris paribis” RGDP will go down over time.

“It is a little disturbing, however that RGDP/cap fell after 1960. It should not fall.”

Don’t forget the old stand by, the Solow growth model. American growth is mostly driven by technological change–invention and innovation. This is quite prone to chance: penicillin, Einstein, Edison… history is riddled with examples of random occurrances or individuals with large effects, and there are many more we can’t observe.

Also, things like wars and oil shocks affect growth. Also, potentially population make up affects RGDP. If a baby is born, RGDP goes up, RGDP/capita goes down. If an immigrant comes to the US RGDP goes up, RGDP/capita could go up or down.

“What do you think about the idea of the immigration waves pulling down the 1914 data point, thereby biasing down 1.38% and biasing up 2.23% (i.e. the jump may not really be as big). Do you know of any data that would permit a rough calculation of that effect?”

I think it potentially explains some of the difference. If 1% of people are added with half the average productivity it lowers RGDP/capita a little less than half a percent. Do it again and gdp/capita is down 1% So low skilled waves could have a noticable effect. Though I doubt my numbers match the actual numbers very well.

“I wanted to point out that the data actually support the opposite of Selgin’s point”

Your second post of numbers lends some support to your assertion, but only because 2.23% from 1.38% seems like a lot larger jump than 1.38% from 1.36%. But I hope you realize that your first post did NOT have the data to illustrate your point, seeing as RGDP/cap is expected to grow with or without a Fed.

It is a little disturbing, however that RGDP/cap fell after 1960. It should not fall. It should not even stay the same. It should increase. It is suggesting that some negative economic factor is more than negating the positive productivity effect of adding more people to the trade pool.

What do you think about the idea of the immigration waves pulling down the 1914 data point, thereby biasing down 1.38% and biasing up 2.23% (i.e. the jump may not really be as big). Do you know of any data that would permit a rough calculation of that effect?

Here is dividing pre-1914 and post 1914 into two equal periods

1790 -1852, 1852 – 1914, 1914 – 1961, 1961 – 2008

Real GDP per capita 1.36% 1.38% 2.23% 2.17%

So growth rate of RGDP/capita growth rate will prove to be very difficult to pin down. It is the change in a trend line and likely to be very small compared to the changes in the GDP/capita series.

I don’t think looking at GDP series will lead to a very convincing argument to for or against the fed. Mostly, I wanted to point out that the data actually support the opposite of Selgin’s point, and I think that makes him look uninformed or disingenuous.