Wealth Redistribution or Wealth Return?

[Loren Pechtel]

Did you present the information that any unemployment went up as a result or that hours worked dropped as a result? All you did in that thread was say that rising minimum wage would cause an indiscernible increase in the unemployment rate. Wouldn't it be easier for you to prove your case by showing how jobs decreased or workable hours decreased?

1) You're changing topics.

2) If you can't measure it you can't measure it. The problem is that 99% of workers aren't minimum wage and that provides a huge amount of noise for the signal to hide in. We have had one case where the signal didn't have a ton of noise to hide in and it was very clear.

You can't extrapolate a trend from a single datum.

Your conclusion is exactly as poorly supported by data as you claim your opponents' positions to be.

Null results don't actually show there is no effect, they actually only place an upper bound on the size of the effect (showing that it's not larger than the noise floor.) Since that constraint is way beyond any expected result the studies prove nothing. Thus it comes down to 1 data point shows a clear effect and a whole bunch of data points that are meaningless.

It's not good data but the best fit is that there's an effect that normally doesn't go above the noise floor. Saying there is no effect leaves your side to explain why we saw a huge effect in the one instance where a result above the noise floor would be expected. Occam's razor, ball's in your court.

 

[Loren Pechtel]

Did you present the information that any unemployment went up as a result or that hours worked dropped as a result? All you did in that thread was say that rising minimum wage would cause an indiscernible increase in the unemployment rate. Wouldn't it be easier for you to prove your case by showing how jobs decreased or workable hours decreased?

1) You're changing topics.

2) If you can't measure it you can't measure it. The problem is that 99% of workers aren't minimum wage and that provides a huge amount of noise for the signal to hide in. We have had one case where the signal didn't have a ton of noise to hide in and it was very clear.

That is why you'd need to show either min wage jobs or hours worked decreased, otherwise, you are making a claim that isn't supported by anything. Saying it is infalsifiable doesn't work in your favor.

I'm not saying it's infalsifiable, but that normally nobody would be stupid enough to set up a situation where it could be falsified. It took an unintended outcome to actually create a test case--and we got the expected result.

There are many things where scientific study of an outcome is unacceptable because of the anticipated bad outcomes. Thus we must rely on approximations and observing what happens when the situation happens anyway. For example, in chemistry the LD50 of a chemical can never be tested, it's always either an animal approximation (which occasionally will prove wildly wrong) or observations of accidents/suicides.

 

[bilby]

Null results don't actually show there is no effect, they actually only place an upper bound on the size of the effect (showing that it's not larger than the noise floor.) Since that constraint is way beyond any expected result the studies prove nothing.

... including the thing you are absolutely determined to claim.

 

[Loren Pechtel]

Null results don't actually show there is no effect, they actually only place an upper bound on the size of the effect (showing that it's not larger than the noise floor.) Since that constraint is way beyond any expected result the studies prove nothing.

... including the thing you are absolutely determined to claim.

You're still not addressing the point.

We have one situation where we would expect to see an effect--we saw an effect.

We have a whole bunch of situations where we wouldn't expect to see an effect--and we didn't see an effect. Continuing to do studies which have an expected result of null does absolutely nothing to discredit the one situation where we saw something. You're trying to bury it with bogus data.

 

[bilby]

We have one situation where we would expect to see an effect--we saw an effect.

"We" don't expect to see anything. We look at what is there, and report what we actually see, regardless of expectations. Good research is blind. You have a vast number of tests, and only a single datum - so no conclusion can be drawn.

You have a bias towards a particular conclusion, so you refuse to accept that a single datum is meaningless, because you're desperate for there to be some support for the conclusion you prefer, and the one data point happens to break in your preferred direction. But it's in what is obviously an atypical situation - it tells you nothing.

If you suspect that a coin is biased towards heads, and you toss it once, get heads, and then lose the coin on the second toss, the only valid conclusion is that you have no way to tell whether the coin was biased. Your "expectation" that you would get heads isn't a valid input into any conclusions. Maybe the coin was biased; Certainly the researcher was.

The "point" is that your really, really, wanting one datum to be admitted as evidence of a trend doesn't make that actually possible. And that goes doubly for a datum that's from a wildly atypical environment - Samoa isn't a similar economy to that of the USA as a whole. It's not even particularly similar to that of the only US state that's a Pacific Island.

The number of possible confounding variables is vast, even if we were to discard the massive bias that arrives under cover of "we would expect".

I wouldn't expect; I would seek out hard information. As would the many researchers whose results you casually dismiss. Goose sauce is gander sauce; If the data don't support a conclusion, then the data don't support your "expected" conclusion.

 

[laughing dog]

Null results don't actually show there is no effect, they actually only place an upper bound on the size of the effect (showing that it's not larger than the noise floor.) Since that constraint is way beyond any expected result the studies prove nothing.

... including the thing you are absolutely determined to claim.

You're still not addressing the point.

We have one situation where we would expect to see an effect--we saw an effect.

We have a whole bunch of situations where we wouldn't expect to see an effect--and we didn't see an effect. Continuing to do studies which have an expected result of null does absolutely nothing to discredit the one situation where we saw something. You're trying to bury it with bogus data.

No, you are using an outcome from literally isolated and tiny economy to make a generalization about any economy. You persist in ignoring context and basic economics because they contradict your ignoronomics.

Whether or not a specific increase in a minimum wage is an empirical question that depends on a multitude of factors. Nothing you post changes that reality.

 

[Loren Pechtel]

We have one situation where we would expect to see an effect--we saw an effect.

"We" don't expect to see anything. We look at what is there, and report what we actually see, regardless of expectations. Good research is blind. You have a vast number of tests, and only a single datum - so no conclusion can be drawn.

I'm referring to historical--how well reality fits the predictions. Thus it is reasonable to have an expected outcome.

And we do not have a single datum. Rather, we are testing the hypothesis: "increasing the minimum wage will result in unemployment of some minimum wage workers". We have one case where an effect above the noise floor would be expected--and we saw it. We have many cases (and you say research should be blind--yet engaging in research where the null outcome is expected is not exactly being blind) where any signal would clearly be swamped--and none was detected.

This is not enough data to quantify the effect but certainly enough to see that it exists. Whereas your side keeps focusing on 0 = 0 and pretending it means something while sticking your head in the sand about the case where we got the expected outcome.

You have a bias towards a particular conclusion, so you refuse to accept that a single datum is meaningless, because you're desperate for there to be some support for the conclusion you prefer, and the one data point happens to break in your preferred direction. But it's in what is obviously an atypical situation - it tells you nothing.

One data point certainly can tell you something, it just can't quantify it. You're acting like some of the NASA people that tried to ignore the cold elephant on the Challenger launch.

I wouldn't expect; I would seek out hard information. As would the many researchers whose results you casually dismiss. Goose sauce is gander sauce; If the data don't support a conclusion, then the data don't support your "expected" conclusion.

I dismiss the results of researchers who set out to measure something which is way below the noise floor and pretend there's some importance to finding that there's no detectable signal.

 

[Loren Pechtel]

No, you are using an outcome from literally isolated and tiny economy to make a generalization about any economy. You persist in ignoring context and basic economics because they contradict your ignoronomics.

Whether or not a specific increase in a minimum wage is an empirical question that depends on a multitude of factors. Nothing you post changes that reality.

I am using the outcome from the one situation where it wasn't obviously just going to be swamped by noise. The fact that it was isolated and small is why it could produce a stronger signal, nobody would have done something so bone-headed to an economy they were in charge of.

 

[laughing dog]

No, you are using an outcome from literally isolated and tiny economy to make a generalization about any economy. You persist in ignoring context and basic economics because they contradict your ignoronomics.

Whether or not a specific increase in a minimum wage is an empirical question that depends on a multitude of factors. Nothing you post changes that reality.

I am using the outcome from the one situation where it wasn't obviously just going to be swamped by noise. The fact that it was isolated and small is why it could produce a stronger signal, nobody would have done something so bone-headed to an economy they were in charge of.

That response missed the point - you are generalizing from a literal isolated incident in a very specific context - a context that is relevant anywhere else.

 

[bilby]

Whereas your side

I don't have a "side".

I have a position; I have arguments to support that position. But I am not playing a team sport; I am not responsible for, nor beholden to, the arguments or positions taken by others who oppose your nonsense; and I can assure you, I do not have a "side".

You're not the champion of a popular cause with a huge crowd of supporters at your back, facing off against a mob of indistinguishable opponents who oppose you only because they're suffering from groupthink; You're just another person on the Internet whose arguments stand, or fall, on their own merits (or lack thereof).

When you say "your side", you are implying a false dichotomy fallacy. You should stop doing that if you want to be taken seriously.

An Internet Discussion Board is not a two party system.

 

[Loren Pechtel]

No, you are using an outcome from literally isolated and tiny economy to make a generalization about any economy. You persist in ignoring context and basic economics because they contradict your ignoronomics.

Whether or not a specific increase in a minimum wage is an empirical question that depends on a multitude of factors. Nothing you post changes that reality.

I am using the outcome from the one situation where it wasn't obviously just going to be swamped by noise. The fact that it was isolated and small is why it could produce a stronger signal, nobody would have done something so bone-headed to an economy they were in charge of.

That response missed the point - you are generalizing from a literal isolated incident in a very specific context - a context that is relevant anywhere else.

The difference was one of size, not of kind.

Big effect, we see it. Smaller effect that's basically impossible to see above the noise floor, of course we don't see it. All that does is set an upper bound--and that upper bound is huge. The minimum signal that could in theory be detected is 10% of the minimum wage workers immediately losing their job--and in practice you would need an insane amount of data to see this as that last digit of the unemployment rate is quite noisy. Our one actual data point involved a far higher than normal number of minimum wage workers so it could be seen.

 

[laughing dog]

No, you are using an outcome from literally isolated and tiny economy to make a generalization about any economy. You persist in ignoring context and basic economics because they contradict your ignoronomics.

Whether or not a specific increase in a minimum wage is an empirical question that depends on a multitude of factors. Nothing you post changes that reality.

I am using the outcome from the one situation where it wasn't obviously just going to be swamped by noise. The fact that it was isolated and small is why it could produce a stronger signal, nobody would have done something so bone-headed to an economy they were in charge of.

That response missed the point - you are generalizing from a literal isolated incident in a very specific context - a context that is relevant anywhere else.

The difference was one of size, not of kind.

Big effect, we see it. Smaller effect that's basically impossible to see above the noise floor, of course we don't see it. All that does is set an upper bound--and that upper bound is huge. The minimum signal that could in theory be detected is 10% of the minimum wage workers immediately losing their job--and in practice you would need an insane amount of data to see this as that last digit of the unemployment rate is quite noisy. Our one actual data point involved a far higher than normal number of minimum wage workers so it could be seen.

You are not addressing my point with that blather.

 

[bigfield]

And we do not have a single datum. Rather, we are testing the hypothesis: "increasing the minimum wage will result in unemployment of some minimum wage workers". We have one case where an effect above the noise floor would be expected--and we saw it. We have many cases (and you say research should be blind--yet engaging in research where the null outcome is expected is not exactly being blind) where any signal would clearly be swamped--and none was detected.

Can someone link these research results, please? I see a hint of an interesting discussion about statistics but I have no idea what data you're all talking about.

And forfucksakes, if you're going to whine about how we've gone over this before then at least provide a link. Every time you write that Loren I just assume gone over this before means you wrote some one-liner about how you've "looked at the data".

 

[Loren Pechtel]

No, you are using an outcome from literally isolated and tiny economy to make a generalization about any economy. You persist in ignoring context and basic economics because they contradict your ignoronomics.

Whether or not a specific increase in a minimum wage is an empirical question that depends on a multitude of factors. Nothing you post changes that reality.

I am using the outcome from the one situation where it wasn't obviously just going to be swamped by noise. The fact that it was isolated and small is why it could produce a stronger signal, nobody would have done something so bone-headed to an economy they were in charge of.

That response missed the point - you are generalizing from a literal isolated incident in a very specific context - a context that is relevant anywhere else.

The difference was one of size, not of kind.

Big effect, we see it. Smaller effect that's basically impossible to see above the noise floor, of course we don't see it. All that does is set an upper bound--and that upper bound is huge. The minimum signal that could in theory be detected is 10% of the minimum wage workers immediately losing their job--and in practice you would need an insane amount of data to see this as that last digit of the unemployment rate is quite noisy. Our one actual data point involved a far higher than normal number of minimum wage workers so it could be seen.

You are not addressing my point with that blather.

You need an explanation that covers the two data points we have. I have yet to see this addressed, just endless derails to avoid addressing it.

 

[Loren Pechtel]

And we do not have a single datum. Rather, we are testing the hypothesis: "increasing the minimum wage will result in unemployment of some minimum wage workers". We have one case where an effect above the noise floor would be expected--and we saw it. We have many cases (and you say research should be blind--yet engaging in research where the null outcome is expected is not exactly being blind) where any signal would clearly be swamped--and none was detected.

Can someone link these research results, please? I see a hint of an interesting discussion about statistics but I have no idea what data you're all talking about.

American Samoa. US minimum wage laws suddenly ended up applying, much of the economy was working below that point. Major unemployment resulted. There are endless attempts to debunk this relationship by trying to measure changes to the unemployment rate from minimum wage hikes--oops, about 1% of workers are at minimum wage. Unemployment is only measured to one digit past the decimal place (and that digit is shaky.) Look at what happens if 10% of minimum wage workers were suddenly laid off--10% * 1% = .1%. Thus you can't hope to detect anything that doesn't result in laying off of 10% of minimum wage workers within a month.

(Beware of a standard deception--the advocates for a higher minimum wage usually refer to those working at or below minimum wage. Oops--the latter category is tipped workers, they typically end up well above minimum wage.)

And forfucksakes, if you're going to whine about how we've gone over this before then at least provide a link. Every time you write that Loren I just assume gone over this before means you wrote some one-liner about how you've "looked at the data".

Because the eternal requests for sources are really just a form of derail to avoid addressing the issue.

 

[Bomb#20]

Sorry for the delay, real life yada yada...

... ANd, like Politesse, I am also one who just hates a system that makes being a kleptocrat legal - that enables the obscenely wealthy to take from those who labor.

The second accusation to fly is that I want to “take the means of production” by which they mean the things that were purchased by amassing money off of someone else’s physical labor, and which are then artifically protected … essentially, “once I bully or coerce someone to give it to me, it’s now my honorable sweat that I claim as my own.” But I’ve never said that I wanted means of production collectivized - only fairly valued.

Then there’s the “but ideas are worth something! You stupid laborers don’t understand!” Except I’m a patent-holder myself.

I am someone who objects to a system that protects bullies and kleptocrats.

What makes you think the CEOs and owners are kleptocrats? What makes you think the obscenely wealthy "take" from those who labor? Is it "Like Politesse"? I.e., because the CEO doesn't go out and do the digging? Are you, like he, in the grip of the Labor Theory of Value? Do you think the source of profit is "amassing money off of someone else’s physical labor", as though physical labor were all it takes to produce? Do you put the words "it’s now my honorable sweat" in the mouth of your cartoon villain because you think it's a fact of objective morality that only sweat is honorable so your cartoon villain can reasonably be expected to agree with you about that? The Labor Theory of Value is metaphysical drivel with no more empirical evidence for it than the Qi theory of medicine.

Why do you accuse the guy who owns the means of production of bullying or coercion? Do you think when two people trade, one of them being better off automatically means the other one must be worse off and only traded because he was bullied or coerced? Do you think in a just world nobody would ever get better off from a trade? Or do you decide someone was bullied and coerced even when he's the one who sought out the trade in the first place because he figured the trade made him better off? What makes you think you know what's good for him better than he does?

As for ideas being worth something and stupid laborers not understanding, laborers by and large aren't stupid; socialists are. Laborers enter deals with capitalists because laborers intelligently know doing so is in their best interests. Socialists try to torpedo laborers' deals with capitalists because socialists stupidly talk themselves into believing those deals are not in laborers' best interests because "property is theft" or "the source of value is labor" or some similar metaphysical gobbledygook and ignore the fact that whenever socialists have cut capitalists out of the production process, production fell off massively and they usually caused a famine.

Patentable ideas are not the only kind of ideas that are worth something. If socialists think ideas of the form "If you and I collaborate to make an X, the X will be owned by Y" are not a part of the means of production and aren't worth anything, then stupid socialists don't understand.

 

[laughing dog]

No, you are using an outcome from literally isolated and tiny economy to make a generalization about any economy. You persist in ignoring context and basic economics because they contradict your ignoronomics.

Whether or not a specific increase in a minimum wage is an empirical question that depends on a multitude of factors. Nothing you post changes that reality.

I am using the outcome from the one situation where it wasn't obviously just going to be swamped by noise. The fact that it was isolated and small is why it could produce a stronger signal, nobody would have done something so bone-headed to an economy they were in charge of.

That response missed the point - you are generalizing from a literal isolated incident in a very specific context - a context that is relevant anywhere else.

The difference was one of size, not of kind.

Big effect, we see it. Smaller effect that's basically impossible to see above the noise floor, of course we don't see it. All that does is set an upper bound--and that upper bound is huge. The minimum signal that could in theory be detected is 10% of the minimum wage workers immediately losing their job--and in practice you would need an insane amount of data to see this as that last digit of the unemployment rate is quite noisy. Our one actual data point involved a far higher than normal number of minimum wage workers so it could be seen.

You are not addressing my point with that blather.

You need an explanation that covers the two data points we have. I have yet to see this addressed, just endless derails to avoid addressing it.

Been there, done that. In the other hand, you keep blathering on the Samoan minimum wage hike as if it is the philosopher’s stone instead of an isolated incidence in a specific context that is unlikely to similar to other situations. Nothing you have posted rebuts the observation that the effects of a minimum wage increase are an empirical matter that depends on the context.

 

[bilby]

And we do not have a single datum. Rather, we are testing the hypothesis: "increasing the minimum wage will result in unemployment of some minimum wage workers". We have one case where an effect above the noise floor would be expected--and we saw it. We have many cases (and you say research should be blind--yet engaging in research where the null outcome is expected is not exactly being blind) where any signal would clearly be swamped--and none was detected.

Can someone link these research results, please? I see a hint of an interesting discussion about statistics but I have no idea what data you're all talking about.

American Samoa. US minimum wage laws suddenly ended up applying, much of the economy was working below that point. Major unemployment resulted. There are endless attempts to debunk this relationship by trying to measure changes to the unemployment rate from minimum wage hikes--oops, about 1% of workers are at minimum wage. Unemployment is only measured to one digit past the decimal place (and that digit is shaky.) Look at what happens if 10% of minimum wage workers were suddenly laid off--10% * 1% = .1%. Thus you can't hope to detect anything that doesn't result in laying off of 10% of minimum wage workers within a month.

(Beware of a standard deception--the advocates for a higher minimum wage usually refer to those working at or below minimum wage. Oops--the latter category is tipped workers, they typically end up well above minimum wage.)

And forfucksakes, if you're going to whine about how we've gone over this before then at least provide a link. Every time you write that Loren I just assume gone over this before means you wrote some one-liner about how you've "looked at the data".

Because the eternal requests for sources are really just a form of derail to avoid addressing the issue.

You were asked for a link to the research results. Not for your personal reminiscence of the time when you saw this fabled research.




By the way, the "noise floor" you keep mentioning, that obscures the effect? That's the evidence that your position is oversimplified.

What do you imagine causes the signal to become noisy below a fairly high threshold?

Your simple model says that increasing minimum wage by a very large amount, in a single day, can cause a significant rise in unemployment. But it also says that this relationship is not detectable (due to "noise") for small increases in minimum wage, or for large increases spread over long periods of many smaller steps. Why would that "noise" exist?

Could it be that the system you're attempting to describe with a very simple model is, in fact, a very complex and chaotic system, with multiple feedbacks, some positive and some negative, which act to render simple models useless?

Could it be that the "expected" job losses from your simple model are, in reality, swamped by other factors, which quite possibly include job gains caused by the same changes that could lead to job losses under other circumstances?

There is a perfectly sound case that can be made that increases in minimum wage can, in some economic conditions, lead to an increase in the size of the economy, and as a result, to increases in the number of jobs available. Of course, there are a vast number of factors that make such a thing difficult to predict with great accuracy; It's usually easier to simply do it, and then look for any consequences.

The only situation where you have any data is for a massive increase in the minimum wage in a tiny and highly constrained economy with a minuscule number of industries, and a tiny number of employers in each of the major industries that does exist.

Yes, I would expect that large economic intervention in such an extreme situation would result in serious problems.

No, I do not think that this tells us anything particularly useful about small interventions in much larger and more open economies. And nor should anyone who has noticed that the unemployment data in such large economies is very noisy.

That noise is the sound of things being too complex for your simplistic models to handle.

 

[bigfield]

American Samoa. US minimum wage laws suddenly ended up applying, much of the economy was working below that point. Major unemployment resulted.

Link?
What year did this happen?
How much did the wage change?
How much did unemployment go up?
What were the trends in both of those metrics before and after?
How did wages and unemployment change in other jurisdictions around that time period?

What analysis did you do to establish that this change in min wage and unemployment is representative of relationship between min wage and unemployment in general?

On what grounds can you reject the null hypothesis, which is that min wage is not related to unemployment?

 

[Bomb#20]

Outside of some academic ivory tower, in the real world you don't get to call people excrement and still be presumed to be a person having good will toward them. It doesn't work that way.

You're the one who characterized my criticism as being of "the wealthy". Wealth is not the problem, I'd be very happy if everyone had the things that make a person "wealthy" in this world. But those who abuse the power they have to devalue and manipulate the labor of their employees, whoever they are, deserve no respect. They're criminals, and not out of necessity but out of greed.

You repeat the classic accusations Christians traditionally made against the Jews for centuries, only against the "wealthy <insult> class" instead, and then you clarify that you don't mean all wealthy people and some of them are okay, just like every anti-semite ever who clarified he doesn't mean all Jews and some of them are okay and he has Jewish friends, and you seriously think that's enough to make you a non-bigot?

Out of curiosity, just how many of the wealthy did you mean to exclude from the wealthy "meanie-head" class you disparaged? So far the only way you've offered for a stockholder or CEO not to be guilty of abusing his power and devaluing and manipulating the labor of his employees is "goes out and does the digging". You've made it abundantly clear that you don't need to actually see a person abuse his power to declare him a criminal -- you're transparently deducing it from your armchair, relying on a prescientific philosophical theory of economics that was superseded by scientific economics back in the 1870s. Since nearly all the wealthy have stock portfolios it seems unlikely your belief system would support your exempting very many from your insult, or from your accusation of criminality.