A follow-up on language and unemployment
Vaidas Urba left me this comment after my previous post:
One more data point – South Tyrol (Alto Adige), a german-majority province in Italy is colored blue in the map.
That left me wondering about other evidence on this question. So I looked at the percentage of residents with German ancestry, by state. Here are the top ten:
Here are the unemployment rates for the top 10 states, in order:
2.8%, 3.4%, 2.8%, 3.1%, 3.8%, 3.1%, 3.8%, 4.5%, 3.8%, 5.0%.
The national rate is 4.5%, so those do tend to be low unemployment states. But there are also states with small African-American populations, so I’m not sure if this is significant. FWIW, my home state of Wisconsin (unemployment = 3.4%) does have significantly more African-Americans than states like Washington (unemployment = 4.7%) and West Virginia (unemployment = 4.9%). And yet Washington has more high tech industries than Wisconsin.
Someone could do a multivariate regression with percent German ancestry and other ethnic groups as well to see if there is actually anything there.
PS. South Dakota has a very large Native American population (nearly 10%). I recall that the tribes in western South Dakota suffer from a lot of poverty. That makes me surprised by their 2.8% unemployment rate. Are Native Americans often counted as “not in the labor force”?
Tags:
3. May 2017 at 07:37
comment
3. May 2017 at 08:02
I would guess that many Native Americans on reservations are not in the LF because they are not actively looking for work (why would when you know no jobs are available).
This paper is dated but says the same thing: https://www.bls.gov/opub/mlr/1982/07/rpt3full.pdf
3. May 2017 at 08:25
Jeff G,
You said, I would guess that many Native Americans on reservations are not in the LF because they are not actively looking for work (why would when you know no jobs are available).
A 2.8% unemployment rate is essentially a representation of what full employment looks like. In other words, the jobs are there for those willing and able to do them. They may not pay enough to cover their reservation wage (pun intended, although probably confusing to those who are unfamiliar the economics term “reservation wage”), but they are indeed available when the unemployment rate is that low.
3. May 2017 at 08:42
Happyjuggler, I’ve read that reservations don’t have good property rights, which discourages investment. South Dakota’s a big state, and the jobs may be in other locations. I’ve also read that alcoholism is a big problem out there.
3. May 2017 at 09:04
Ok, I am going to issue a mea culpa here; I seem to have radically underestimated the issue of “isolation from urban centers” factor.
From wikipedia (not always a reliable source, but here it is anyway):
Five of the state’s counties are wholly within the boundaries of sovereign Indian reservations.[91] Because of the limitations of climate and land, and isolation from urban areas with more employment opportunities, living standards on many South Dakota reservations are often far below the national average; Ziebach County ranked as the poorest county in the nation in 2009.[92] The unemployment rate in Fort Thompson, on the Crow Creek Reservation, is 70%, and 21% of households lack plumbing or basic kitchen appliances.[93] A 1995 study by the U.S. Census Bureau found that 58% of homes on the Pine Ridge Indian Reservation did not have a telephone.[94] The reservations’ isolation also inhibits their ability to generate revenue from gaming casinos, an avenue that has proved profitable for many tribes closer to urban centers.
3. May 2017 at 10:38
Canada appears to present an interesting story along these lines as well.
Dont have time to do anything deep but here are some dated maps:
https://www.quora.com/Why-does-Canada-treat-French-as-an-official-language-when-the-number-of-people-who-identify-themselves-as-French-ethnicity-is-very-small#!n=12
http://www.canadianbusiness.com/economy/gallery-canada-unemployment-rates/
3. May 2017 at 11:40
Jawohl ja, it must be the German ancestry. What else could it be?
3. May 2017 at 13:44
Naive question, in the 1960s we often saw countries with unemployment rates of 1% to 2%. How was that possible? Was it really just a bunch of socialist make-work projects?
France had a sub 1.5% unemployment rate from 1960 to 1966.
https://fred.stlouisfed.org/series/LRUN74TTFRA156N
Germany’s unemployment rate was sub 1% until the oil crisis
https://fred.stlouisfed.org/series/LMUNRRTTDEQ156S
3. May 2017 at 14:57
Correlation does not equal causation.
When will this blog make it to Econ 101?
3. May 2017 at 17:32
MF,
Since when did Austrians concern themselves with things like data analysis or econometrics… or are you no longer an Austrian given that your a nationalist Trumpista? Nevermind that… we could probably just review some of Andrew Jacksons python scripts on this topic, that would set us straight.
3. May 2017 at 18:28
Cooper:
Man oh man, the 1950s and 1960s were a great time.
Real per capita income in the US grew by 75% through those two decades!
It is a curious thing.
Since the 1960s, in the US the top income tax rate has been reduced from 90% to 40%; unions have been destroyed; global trade has boomed, whole industries deregulated, such as transportation, finance and telecommunications. De facto open borders for migration.
The Fed has brought inflation under control, supposedly a huge bonus.
Yet in some regards the US economy is atrophying today and wages have been dead since 1972.
Orthodoxy in any endeavor is the right course only in stasis.
Orthodoxy may be defensible, say, in religion. After all, if we believe in God, then I suppose we can contend there are beliefs that are eternal.
Orthodoxy in macroeconomics?
It can become ossified, self-reverential, inaccurate—at worst, mere quackery.
4. May 2017 at 00:48
What’s the causal mechanism? Is German heritage somehow more conducive to accepting any form of work? Does the German language enable people to think more rationally and hence be more employable? Or is German culture more conducive to better monetary policy?
What about half German, half other? Do unemployment rates with mixed Germanic populations perform half as well? Is there a lazy race that destroys the good German bits?
Or perhaps this is nothing more than regression to the mean or the tendency of humans to see patterns in random data.
4. May 2017 at 01:20
That German/Central European low unemployment rate maps quite nicely across the Hapsburg possessions. The higher German rates are in the core of old Prussia.
Not quite sure how I’d try to explain causality but there’s certainly – some – correlation there.
Yes, East Germany and socialism, of course, but the Czech Republic and Slovakia have very low unemployment rates and they were both socialist and Hapsburg.
4. May 2017 at 04:02
‘ Does the German language enable people to think more rationally and hence be more employable? Or is German culture more conducive to better monetary policy?’
When I taught myself a little German many, many years ago, I remember thinking that ‘Now I understand why all the great composers were German; It’s in the language.’
I.e., with three genders and two objective cases–nouns and their adjectives have different endings depending on whether they are objects of verbs or prepositions–and verbs at the end of the sentence, the language ‘sings’. IIRC, there are 16 different ways to say ‘the’ in German.
Though Weimar monetary policy was nothing to write home about.
4. May 2017 at 08:39
@ Patrick R. Sullivan:
If you like cases, you should have studied Lithuanian (though I think Lituanian lacks a word for ‘the’).
4. May 2017 at 18:25
“What’s the causal mechanism?”
I suppose it would be Protestant vs. Catholic cultures. The Protestant work ethic speculated that some sort of Calvinist/Luthern heritage explains a higher work ethic. One could say the Puritan heritage of the northeast and the German heritage of the Midwest explains higher productivity and lower unemployment.
This is necessarily very imprecise and I’m not convinced. In Europe’s case, I’m more convinced by specific labor market reforms which Germany did and Romance language countries did not do.
For America, probably it’s that these Midwest states don’t have African-Americans or Appalachians with Scots-Irish ancestry. I think Washington has a local issue with declining lumber sales.
4. May 2017 at 20:25
OT
The RBA hold rates flat despite inflation being below target and economic growth slow. Why?
High house prices and worries about “financial instability.”
Aussie runs large trade deficits.
http://ngdp-advisers.com/2017/05/04/reserve-bank-australia-ponders-financial-instability-house-prices-soar-amid-trade-deficits-rba-straitjacket/
5. May 2017 at 03:01
“I suppose it would be Protestant vs. Catholic cultures.”
This could be tested in Switzerland, where you have French-speaking Protestant cantons (like Vaud) and German-speaking Catholic cantons, like St. Gallen
5. May 2017 at 05:37
@Miguel Maderia, looks pretty falsified to me. St Gallen has 2.42% unemployment (https://www.bfs.admin.ch/bfs/en/home/statistics/regional-statistics/regional-portraits-key-figures/cantons/st-gallen.html) and Vaud has 4.97% unemployment (https://www.bfs.admin.ch/bfs/en/home/statistics/regional-statistics/regional-portraits-key-figures/cantons/vaud.html).
5. May 2017 at 18:17
Students:
“Since when did Austrians concern themselves with things like data analysis or econometrics”
Since the start of Austrian economics, because that is when Austrians began refuting the assumptions behind the methodology of econometrics
“or are you no longer an Austrian given that your a nationalist Trumpista?”
Nope, still anti-Trump, still an anarchist
6. May 2017 at 04:36
So many sexists in France, not wanting LePen to become the first female President…
Where are all the “It’s her turn” feminist Hillary supporters, and why do they prefer Macron?
6. May 2017 at 05:56
Cooper, Europe had a much smaller welfare state back then. And of course the post war reconstruction boom helped. Virtually all parts of the world (except China and India and Africa) boomed in the 1960s.
Tim, I don’t think Slovakia has low unemployment. Look again.
Matthew, You said:
“I suppose it would be Protestant vs. Catholic cultures”
Nope, Belgium shows its language, not religion. And isn’t southern Germany also Catholic?
6. May 2017 at 08:12
MF, can you give me the lowdown on the Austrian refutation of econometrics?
6. May 2017 at 09:21
Steve F:
The lowdown:
Econometrics presumes the existence constant, time-invariant operating causal relations between homogeneous observable variables.
There are no such constants in the field of individual human knowledge and action, hence econometrics is the wrong method to use.
That’s the summary.
It shouldn’t be controversial to consider, because in over 80 years of econometrics, not one econometrician has discovered any empirical constant. Physicists and chemists have been discovering hundreds of constants in their fields, which any glossing of university/college science textbooks indexes can show.
The way to know there are no constants is to study the method of econometrics and understand what logically must be true about the econometricians themselves by virtue of using the methodology. That is to say, you can use self-referential analysis to test whether econometricians practise what they preach about how they learn about the world.
Clearly, when an econometrician conducts a test, they are presuming that their own minds are NOT constant, that their minds will be different from the time they start the test to the time they claim to have learned something new about the world. For why else do the test?
Given econometricians presume a distinct absence of constancy of their own minds, and given their minds determine their actions, it follows that their actions are not constant either. Note that the absence of constancy in their minds and actions is not merely a side effect or some cosmic trick. It is necessary in the method of econometrics itself.
When economists started borrowing from the method of physics and chemistry, they didn’t realize that the reason it worked in physics and chemistry is because the subject matter of physics and chemistry is that which does not know or learn or act. When they started applying that to subject matter which does know, learn and act, namely people, themselves, they committed a logical error.
Austrian economics is not anti-science. It is anti-scientism. Meaning, it rejects, rightfully so, the notion that empiricism is and should be the sole method to learn anything about anything in the universe, now and forever.
Most scientists today have no idea about the historical evolution of the method they now use came to be. Most believe that the scientific method as they know it has been around for thousands of years, oppressed some of the time, allowed to flourish at other times, when in reality it only arose during the 1920s, from a particular school of philosophers at the University of Vienna Austria of all places.
6. May 2017 at 09:27
Action presupposes a causally structured observational reality but the reality of action which we can understand as requiring such structure, is not itself causally structured. Instead, it is a reality that must be categorized teleologically as purpose-directed, meaningful behavior.
6. May 2017 at 09:27
Why are there no posts on this blog about what is going on in Venezuela?
http://globalnews.ca/news/3428049/venezuelan-tanks-plow-through-crowds-as-2-month-protest-continues/
6. May 2017 at 09:54
MF, thanks.
That sounds like the same premise an Austrian in my intermediate macro class has given me. I’m not sure what to make of it, because I think it is obviously true but that doesn’t mean econometrics don’t have descriptive or predictive power to degrees. Perhaps the lack of constants means an econometric model could be just minutely wrong yet still provide a large amount of valuable information.
The expectations-augmented Philips curve is a pretty simple econometric model that AFAIK holds up quite well in most circumstances. To help clarify for me, could you explain why using this is wrong?
6. May 2017 at 11:14
Steve F:
“but that doesn’t mean econometrics don’t have descriptive or predictive power to degrees”
The naive prediction of “If it happened yesterday then I predict it will again happen today” could also be considered having predictive power ‘to degrees’. But saying there is predictive power ‘to degrees’ is really just another way of saying that guesses can sometimes turn out correct.
“Perhaps the lack of constants means an econometric model could be just minutely wrong yet still provide a large amount of valuable information.”
If there is a lack of constants, then what are the ‘independent variables’ in the equations really supposed to be?
“The expectations-augmented Philips curve is a pretty simple econometric model that AFAIK holds up quite well in most circumstances.”
Why not the other circumstances?
“To help clarify for me, could you explain why using this is wrong?”
If you saw me drive to a certain office building Mondays to Fridays each week, and after say 3 months of observations you made a prediction on a Sunday that I will go to the office building the next day on Monday, then assuming I did go the next day, then the method of econometrics (physics) tells us that there is a law of nature at work here, akin to the gravitational constant (6.67408×10−11 m^3 / kg^1 / s^2).
Why is the gravitational constant observed as the same no matter where or when it is measured?
Then compare that to the expectations augmented Phillips curve. Why is not observed to have the same constants no matter where or when it is measured? Why is the model presented as time and location invariant when it really isn’t? See the problem?
6. May 2017 at 11:26
Steve F:
The “global warming -> pirate attacks” model has ‘predictive powers to degrees’
https://i.imgur.com/tqoD3om.gif
The expectations-augmented Phillips curve isn’t actually observed to ‘hold up quite well in most circumstances’. We are never in a controlled experimental condition. You are in fact understanding that which you are observing from a Phillips curve based belief. You first start with various beliefs about money and employment, and then you “see” it in the historical data, sometimes more so, sometimes less so. But the theory itself stems from assumptions, not observations.
Compare the Phillips curve to the global warming/pirate attacks curve. The reason you are open to the former but scoff at the latter is precisely because pure observations are not the determining factor. You have assumptions about money and employment, and about Earth temperature and pirates, such that when you look in history you “see” a causal relationship with money and employment, but not with the temperature and pirates.
6. May 2017 at 11:38
MF, what then is the solution? How do economists develop certainty?
7. May 2017 at 16:40
Yes, southern Germany (Bavaria at least) is Catholic.
8. May 2017 at 14:22
@bill
The federal state in the southeast of Germany is called “Bavaria” but it’s actually Franken + Oberpfalz + Schwaben + Lower Bavaria + Higher Bavaria. And then in the southwest there’s also Baden and Wurttemberg. I grew up in the south of Germany and it always felt like 50:50 between Catholics and Protestants (except maybe Oberpfalz and Lower Bavaria).
Huge parts of the map Scott presented in the last post look indeed like a map of Protestantism vs. Catholicism. The similarities are quite astonishing, the only huge parts that don’t really fit seem to be Poland and Austria. The rest fits surprisingly well.
It also fits quite well when you look where Protestant leaders like Luther, Hus, Zwingli, Calvin, Wycliffe and Simons are coming from, or when you look where Calvinists and others moved to (Böhmen, Hungary, Siebenbürgen) – or rather had to move.
8. May 2017 at 14:30
Well even Poland/Czechia fits because it’s basically the Prussian/Böhmian part that’s blue. And Austria might fight as well when you with the Habsburg theory some other guy mentioned.
9. May 2017 at 06:47
MF,
“Econometrics presumes the existence constant, time-invariant operating causal relations between homogeneous observable variables.”
Wrong. Have you never heard of Bayesian Inference? Bayesians have been treating parameters as random variables rather than fixed frequencies for about 200-300 years… A simple way to describe the view is that parameters are assigned probabilities which are interpreted as a reasonable expectation based on the state of knowledge (which is a mixture of prior and current experimental evidence). Common sense reduced to computation, as Pierre Simon Laplace would say.
As a result, there are many econometricians that do not hold your presumed view.
9. May 2017 at 12:31
@Christian List,
I only lived there for a year – near Rosenheim.
12. May 2017 at 20:16
Steve F:
“MF, what then is the solution? How do economists develop certainty?”
Well for one thing the positivist empiricist method, the most dominant method used by mostly government financed economists the world over, expressly forbids certainty in any belief. The concept of “falsification” is quite clear that ANY belief of some truth, even those based on repeated factual historical observations, could in principle be falsified by future observations as new data comes in that is inconsistent with the original theory.
If you really want to know something for certain, then you have to find what is necessarily the case both inside of you and outside of you, in your mind, whatever you think, whether you think anything correct or incorrect, there is something always there that logically cannot be any other way no matter what you try to do or think. THAT is what can be known with certainty.
So what is that you may ask? Well, because absolute truth must at least include what is true about us, about our minds, and because we are not located on the other side of the shoulder off Orion, we are in ourselves first and always, finding some absolute truth must start with our own minds. It cannot start in places we do not exist. By the way, deep down absolute truth is believed by everyone, including the most radical skeptics, including the author of this blog (he has belief in the absolute truth that the human mind is categorically unfit to know absolute truth, which is itself a claim about the absolute truth of the nature of the human mind; in other words he has an absolute belief in a self-contradictory ideology). Absolute truth is inescapable. We are all living in a reality that is absolute.
So what is an absolute truth about us? It must be a reality that cannot be denied even if we tried to deny it. That is, it must be true even when we say it is false. So what is true when we say something is true and when we say it is false? It is something there when we say or do ANYTHING. The answer is action. Regardless of what we say or do, we are acting. This is an absolute truth of the universe. Even if we try to say no, that is wrong, that is false, that has no basis, that has no evidence, that is just semantics, that is just a silly triviality, that is just a truism, that is just dumb, etc, etc, etc, etc, in every single case, in every single attempt, necessarily so, you are acting when you are doing that.
So why is action here? Why does it exist? We don’t know. And we don’t have to know. It may not even have an ultimate cause. Suppose you were absolutely omniscient. Suppose you knew everything that could possibly be known about the universe. Right down to the exact infinite number of points all across the universe. Even then, you would know the why of everything. But then what? After you know the why of everything, it would “only” be everything, and nothing more. You would not know the why to all those why’s because we already stipulated you knew all the why’s. So you would have to just think that what you know, doesn’t have any prior reason or cause.
That is you right now, just without the omniscient part. You don’t have to know everything back to the ultimate beginning of everything. You only need to know “a” absolutely certain point in the infinite space of the all, and that absolutely certain point, is your action.
Descartes started and stayed with “I think therefore I am”, but that is only a mental certainty, that has no bearing on any “external” certainty. Mises extended Kant, and discovered that Action is the absolute bridge that connects the certainty about ourselves with certainty about our surroundings. Action is both internal and external, ideological and materialist, etc. It cannot be denied, as denial is itself an action. Action is the use of scarce means to achieve ends. We must act precisely because we ourselves are scarce. Labor and knowledge are the most valuable of that which is scarce. Valuable for either self-sufficient purposes, or for exchange purposes.
Action it turns out is complex enough that we can extract from it further absolute truths, such as the law of marginal utility, the law of demand, even the quantity theory of money. A lot of confused econometricians and other mainstream economists don’t understand how to think about all of this, they have a superficial self-serving view of it, and they try to use their own flawed methods to understand it, which is why most of them have falsely concluded that these various laws have been “empirically falsified”. They don’t know how to think counter-factually and abstractly. They believe they should be able to see these laws taking place in a similar or same way that physicists see laws in controlled laboratory experiments. The law of demand is not falsified by observing for example some goods having an increased demand after the price increases. The law of demand is a logical constraint, not a prediction of people’s choices over time in response to price changes. The tool “all else equal” is too subtle and too complicated for empiricists to really grasp.
There is a logic to our action. Not in the sense that we make choices that are “logical” like “see moving truck, get out of the way” or “see raging fire approaching, run the other way”, but rather there is a logic in the structure of anything and everything we could ever do. Anything and everything we could do, or anything and everything that any actor in the universe could do, is “constrained” (terrible word by the way, but I see it used so often so meh) by certain absolute truths, of categories. The concepts of profit, loss, means, ends, these are logical categories, not empirical ones. Any empiricist who tried to confirm them or falsify them by experiment, would be subject to them and personally experiencing them no matter what the outcome of their experiment. Their experiment after all would be about them seeking to gain something, avoid (subjective) losses by not doing that thing, they would be using scarce means (their bodies, their equipment and tools, etc) and they would be seeking an end (to acquire knowledge about the world).
12. May 2017 at 20:31
Student:
“Econometrics presumes the existence constant, time-invariant operating causal relations between homogeneous observable variables.”
“Wrong. Have you never heard of Bayesian Inference? Bayesians have been treating parameters as random variables rather than fixed frequencies for about 200-300 years…”
Student, you have to try first then respond. The constancy is still presumed in the “parameters as random variables”, the constancy is in the relationship between the variable placeholders themselves.
You telling me that X is a function of Y and Z, where Y and Z can take on “random values” is the same thing as saying there is a constancy in the relationship between the concepts X, Y and Z, namely, that X is a function of Y and Z (by way of either addition of Y and Z, or some other operand).
Bayesian inference is just masking the constancy by pretending that relaxing the numerical values of the placeholders themselves regarded as constant in their relationship over time, is somehow the same thing as not making any constancy assumption and letting the observed data tell us the relationship. But the relationship was already presumed as constant.
“A simple way to describe the view is that parameters are assigned probabilities which are interpreted as a reasonable expectation based on the state of knowledge (which is a mixture of prior and current experimental evidence). Common sense reduced to computation, as Pierre Simon Laplace would say.”
Yeah yeah, I get all that, this is old hat, what I am talking about is the way you arrange those “parameters”, and that you first assume that the relationship between them over time doesn’t change, a constancy, and then you “plug in” the quantified observations you see to find what the specific expectation is going forward.
“As a result, there are many econometricians that do not hold your presumed view.”
Oh but they do, I am telling you what they themselves assume.
There is in fact no meaning to the statement “There is a one-sixth chance that when I roll this (unloaded) die it will turn up as a 3 in the next roll”. What it is, is a metaphor that refers to something that does have meaning, which is “If I roll this unloaded die a trillion times over time, then 3 will be rolled approximately 1/6 of the number of rolls”. (Assuming we don’t go all Wittgenstein and pick apart that statement until we’re left with a Gödel-like pile of goo)
https://mises.org/library/correct-theory-probability
17. May 2017 at 18:22
Um no, what more can I say…. you completely misunderstand everything. I give up on this… I cannot argue with a stone.
17. May 2017 at 18:26
You clearly don’t understand the difference between frequentist methods and Bayesian methods. Frequentist methods grew out of the study of frequencies like rolling a dice. Bayesian methods are based on propositional logic where the truth or falsisity of a hypothesis is unknown. You sir are a guy the pretends like he knows what he is taking about by writing a lot of words when in fact you are missing the crux of the matter completely.