Republicans survive Recall effort in WI, what will happen in PA, PGH? (Carnegie: buying, calculating)

[MathmanMathman]

Originally Posted by BrianTH

Just for the record, a satirical joke is not a logical argument, and while one might argue that joke embodied an unfair caricature, it is a category error to compare such a caricature to a strawman argument.

True, but it was a set up so the poster could promptly knock it down sort of approach.

[caroline2]

Originally Posted by hempfield mania

Democrats fall short in Wisconsin recall elections - politics - More politics - msnbc.com

Think there will a renewed effort to "break" unions collective bargaining power in PA/PGH now that it seems the "backlash" as some members on here have discussed, does not really seem to be there?

I'm not sure it will even require much of an effort. Unions spent at least a $17 million per net seat gain -- seats in a state senate! -- and still failed to shift any majorities.

The unions were aware of their own irrelevance beforehand, which is why the tens of millions spent on recall campaign ads had nothing to do with a defense of collective bargaining.

Unions... yawn.

[hnsq]

Originally Posted by BrianTH

Labor is a resource.

By the way, I should have noted sooner that the CEO as such is actually just another employee. Presumably you mean "CEO" to stand in for the people who actually supplied capital to the business.

The CEO controls the capital for the business, and the CEO has the responsibility for the entire process.

I'm actually not sure what point you are trying to make. If the person who supplied the ingredients/oven and the baker don't agree to collaborate, then no cookies get baked. Presumably they agreed to collaborate because they each thought they would benefit. None of that tells you anything about how many cookies each should get.

But what if there is a person who would love to bake the cookies for less money, but due to a union artificially imposing conditions, that person cannot work?

So I'm just objecting to the notion that somehow the person who supplies the oven and the ingredients is inherently in a different position from the person who supplies the labor. Those are all necessary factors in the productive process.

But this is a fact...not every job is of equal value to a company. You can close your eyes and point and find someone who can do manual labor/blue collar type of work, however there are very, very few people who can perform the duties of a CEO. Not every job is of equal difficulty. Labor is easy to find. Quality top level management is not.

Already done.

sure. Let's look at your sources. Here is a quote from one, "And since Shedlock’s work indicated that our GDP is wildly overstated, so too is our productivity growth." Starting an article by making wild assumptions (such as productivity is the only contributing factor to GDP growth) is not the best way to go.... Another quote: The dramatic drop in unionization in the United States from 1979 to 2005 did not lead to faster productivity growth than in the seven largest European countries with union density greater than 60% The author is claiming that as less unionization has not led to growth exceeding that of European countries, that unions are productive. That is a backwards inference which disregards lurking variables surrounding nuances of cultural factors.

Looking at another study you posted, the authors state: The measure of interest is the total productivity differential between unionized and non-unionized firms. So the authors look at total productivity differences between whole firms, without regard to industry, region, or market segment? One cannot compare apples to oranges and draw accurate results. In addition to that, please look at the author's calculation of P-bar for the weighted average productivity. They weight purely based on sample size, of which they artificially generated 1000 iterations of statistical forecasting from 53 studies. A sample of 53 is not statistically significant, regardless of the compounding effects used. The very basis of this study is exploding out a small sample by a factor of 1000 to get a statistically significant result! How can I take that seriously?

It actually isn't much of mystery why this would be true. Lower-wage workers may well be less productive workers. Unions also negotiate for things like better training, better health care, safer and more pleasant workplace environments, and other things that could contribute to productivity.

So what happens when you simply take the lower wage worker and pay him/her more? They don't suddenly start working harder! If a person wants better health care, they are more than welcome to quit and apply for a job that already has that health care. Why is a union needed for that?

But maximizing productivity isn't necessarily in the interests of upper management--maximizing their own compensation is presumably their goal, and they may make more money for themselves even if their firm is less productive provided that they can get for themselves a larger share of the firm's income. In other words, even if fewer cookies are baked, if they can keep more of the cookies out of the hands of the other workers, they may personally end up with more cookies that way. Unions are thus likely serving the function of helping to cut off practices that would be bad for firm productivity, but good for upper management.

Maximizing personal compensation is NOT the goal of upper management. I haven't read a thing farther from the truth. What is the point of a corporation? To maximize investor (shareholder) value. Upper management works towards THAT goal. Unions work to give someone who (though their work alone) only deserve one cookie three instead. And in fact, in the process of giving those two extra cookies to the worker, the union completely destroys half a cookie in the process. There is a net of LESS cookies in the system because of the union. How is that a good thing?

I'd have to see the details of your deadweight loss argument to understand where you are going wrong. But at a guess, you are modeling this as a transaction between the prospective employees and the firm as an abstract concept, with the firm's interest being in maximizing productivity. In the real world, it is a transaction between the prospective employee and upper management, with the implications I just noted.

Edit: By the way, it is a little strange to talk about collective bargaining creating "deadweight loss" in this context, because it presumes certain alternatives. To flip the point around, most firms "collectively bargain" automatically--individual units typically don't have the discretion to negotiate compensation and hire as they see fit. So by restricting individual units from freely transacting for labor, are firms thereby creating a "deadweight loss"?

So I also suspect this argument depends on assuming the existence of the union is a variable, but the existence of the firm is not. But you'd have to explain why we should approach the problem in that way, as opposed to either assuming they are both variables, or in fact neither is.

deadweight loss is an economic term. Not a vague concept. read this article. It shows how to calculate deadweight loss due to the minimum wage, however the calculation for minimum wage loss and loss due to collective bargaining is mathematically identical. I am actually very impressed with the calculations in that article. They are very accurate.

Look at this graph:

the intersection of supply (how many cookies the worker gets) and demand (how many cookies all companies are willing to pay) is the equilibrium price. It is the level where everyone is paid and happy (or at least employed). Now when a union bargains for wages, they set the minimum number of cookies at the upper line. Each worker gets more cookies - they are now higher up on the supply curve. The problem is the triangle labeled 'DWL'. That is a very real number which is cookies (aka money) that simply disappears from the economic system. It doesn't go to management or wall street or the government, it is gone. It isn't available to anyone because of the price floor (collective bargaining agreement). Deadweight loss is when the top potential (if everyone was as efficient as possible) of the economy as a whole is decreased due to some external regulation. Brian - read the article I linked and it gives the math so that you could actually calculate this loss yourself.

[gg]

Originally Posted by hnsq

Maximizing personal compensation is NOT the goal of upper management.

The CEOs of the 50 firms that laid off the most workers since the onset of the economic crisis took home 42 percent more pay in 2009 than their peers did...

This only goes to '05, but you get the idea:

http://consumerist.com/images/resources/2007/04/changeinceopaygraph.jpg (broken link)

[hnsq]

Originally Posted by h_curtis

The CEOs of the 50 firms that laid off the most workers since the onset of the economic crisis took home 42 percent more pay in 2009 than their peers did...

This only goes to '05, but you get the idea:

You DO realize that a CEO does not set his own pay, don't you?

[ditchdigger]

Maximizing personal compensation is NOT the goal of upper management. I haven't read a thing farther from the truth. What is the point of a corporation? To maximize investor (shareholder) value. Upper management works towards THAT goal.

Then why do they take a paycheck?

[gg]

Originally Posted by hnsq

You DO realize that a CEO does not set his own pay, don't you?

Yeah sure and I have a bridge to sell you in the desert. All the execs hang out play a round of golf and discuss how much they can all make and how many lowly people they can lay off so they can make more. Then they think of ways to outsource to China or where ever, so... guess what? They can make more money for themselves. Then they cut some other deals in some back room to manipulate whatever they can to.... make more money for themselves.

The trickle down thing didn't work very well. If you look at how much the super rich make compared to the lowlifes, you will see the gap is getting bigger and bigger.

[BrianTH]

Originally Posted by hnsq

The CEO controls the capital for the business, and the CEO has the responsibility for the entire process.

I don't think this is an entirely accurate with respect to a typical CEO's role in a firm--the Board of Directors, for example, often has ultimate say on important capital issues.

But more importantly--so what? Providing management is an important job, but it is just another job, as is being the baker. So this still tells us nothing in particular about how the cookies should be split up.

But what if there is a person who would love to bake the cookies for less money, but due to a union artificially imposing conditions, that person cannot work?

I think you are stretching this analogy past its breaking point--unions only make sense in the context of a large number of employees collectively bargaining with a large firm, so an analogy with only one employee and one owner/manager is inherently unhelpful.

But this is a fact...not every job is of equal value to a company. You can close your eyes and point and find someone who can do manual labor/blue collar type of work, however there are very, very few people who can perform the duties of a CEO. Not every job is of equal difficulty. Labor is easy to find. Quality top level management is not.

As an aside, studies of CEO value-added call a lot of your assertions in this passage into question. Moreover, in the past U.S. CEOs have been paid less, and today in other developed countries CEOs of large countries are paid less. So there is good evidence that in the U.S. we are systematically overpaying CEOs.

Here is a quote from one, "And since Shedlock’s work indicated that our GDP is wildly overstated, so too is our productivity growth." Starting an article by making wild assumptions (such as productivity is the only contributing factor to GDP growth) is not the best way to go....

Um, there was no such assumption. If you understand the specific issue in question, it is that Europeans and the U.S. have been using different methods for comparing GDP in different periods, so if the U.S. was using the European method, its reported GDP growth would be lower (alternatively, if the Europeans were using the U.S. method, their reported GDP growth would be higher). Note nothing is actually changing in the real world--this is just adopting a common standard of GDP measurement to allow more accurate comparisons.

Measures of national productivity in turn use GDP as an input, so if you decrease your index of GDP growth, you will decrease your index of productivity growth. That's not an assumption about the factors contributing to GDP growth, that is simply applying the formula for calculating productivity.

Another quote: The dramatic drop in unionization in the United States from 1979 to 2005 did not lead to faster productivity growth than in the seven largest European countries with union density greater than 60% The author is claiming that as less unionization has not led to growth exceeding that of European countries, that unions are productive. That is a backwards inference which disregards lurking variables surrounding nuances of cultural factors.

I agree international crosstemporal comparisons are difficult, but prima facie, this evidence is not supporting the hypothesis that unions hamper productivity growth. If you want more rigorous studies, the metastudy I linked goes into all that.

So the authors look at total productivity differences between whole firms, without regard to industry, region, or market segment?

Of course not. I assume you didn't actually read the study if you are asserting that.

They weight purely based on sample size, of which they artificially generated 1000 iterations of statistical forecasting from 53 studies. A sample of 53 is not statistically significant, regardless of the compounding effects used.

I honestly have no idea what you are talking about here. This is a metastudy, and the studies in question can't be treated as "samples"--they aren't remotely the same thing.

So what happens when you simply take the lower wage worker and pay him/her more? They don't suddenly start working harder!

You seem to be ignoring your own point that setting wage terms necessarily has an effect on who actually gets employed. Moreover, as I pointed out before, unions do a lot more than just set wage terms.

When your overly simplistic model fails to explain the data, it is time to get a new model, not insist the data must be wrong.

If a person wants better health care, they are more than welcome to quit and apply for a job that already has that health care. Why is a union needed for that?

Are you asking why collective bargaining achieves different results than individual bargaining? It obviously does--you are admitting that yourself when you claim unions actually achieve higher wage terms, and there is no particular reason to believe they would fail to achieve different results in any other area. But precisely how collective bargaining achieves its observable results is a very complex question, and the academic models have varied over time. Personally, I don't have any firm convictions on that subject, and in fact I think it very likely varies considerably by case.

Maximizing personal compensation is NOT the goal of upper management. I haven't read a thing farther from the truth.

Um, there is a whole literature about observable agency problems in corporations for precisely this reason. The goals of upper management in some idealized model and how they actually behave in the real world are two very different things.

What is the point of a corporation? To maximize investor (shareholder) value. Upper management works towards THAT goal. Unions work to give someone who (though their work alone) only deserve one cookie three instead.

As an aside, this is an obvious double standard. In an idealized model, the point of the union is only to ensure fair working terms for its represented workers. In the real world, corporations have serious agency problems with upper management, and unions overreach, but you can't apply a real world analysis to unions and an idealized analysis to management.

But in any event, this is all beside the point. Maximizing equity value is not the same thing as maximizing firm productivity. So even if it were true that management was diligently working to maximize the return to equity investors, that still wouldn't mean they were diligently working to maximize productivity, and all the same analysis would apply: if they can shift enough income share from labor to equity, that could more than make up for losses in productivity.

And in fact, in the process of giving those two extra cookies to the worker, the union completely destroys half a cookie in the process. There is a net of LESS cookies in the system because of the union. How is that a good thing?

Again, you are asserting all this with no proof, and if anything the evidence is that unions lead to more cookies, not less. The problem is that upper management (or equity investors) may prefer to get 11 out of 12 cookies rather than 10 out of 13 cookies.

deadweight loss is an economic term. Not a vague concept.

I know, I just think you are misapplying it.

It shows how to calculate deadweight loss due to the minimum wage, however the calculation for minimum wage loss and loss due to collective bargaining is mathematically identical.

Whoa, that assumption is clearly erroneous. Unions set their wages through collective bargaining, and a minimum wage is set through an entity not party to the transaction. You clearly cannot use the same model for both situations.

I don't want to get too sidetracked since it should be quite clear you cannot use the minimum wage model in this context. But I would note a few things if you want to try to build a real model. For one, you can't treat workers as an undistinguished commodity. For another, you are going to have to model the firm correctly. For the last, you will need to account for the fact everyone has imperfect information, and the contracts in question typically last over significant periods of time.

By the way, as I noted earlier, there is actually a large and complex literature about how to model this. Way back, people tried to model unions as monopolies (which is essentially what you are trying to do), but those models have long since been abandoned as unworkable.

[BrianTH]

Originally Posted by ditchdigger

Then why do they take a paycheck?

Exactly. There is a job description for management, and then there are the actual human beings filling those jobs. In the real world, we recognize the difference, and equity investors struggle with making sure that their managers are faithfully fulfilling their duties--which frequently doesn't happen.

In fact, economic models of corporate governance typically start with the assumption that managers are as self-interested as anyone else, which is probably a safe assumption.

[soniqV]

Let companies that are too big to fail fail, stop the bailouts, let them be replaced as competition flourishes, and jobs will come back, the dollar will recover, and CEO compensation ratios will go back down. There's nothing inherently wrong with collective bargaining, but neither side should be able to use the force of government to achieve their goals. If a company has no problem finding "scabs" then labor is probably overreaching.