The Good. The Supper. The Supreme part the sixth.

It's good to be alive.
It's Super Tuesday.
It's impossible for the Rabid Right to get it's way with the Supreme court.

Regular reader may recall the proposition that motive and incentive always trump (hehe) data when they are in conflict. With the passing of Scalia, the RR immediately looked at the data, their control of the House and Senate, and pronounced that there would be no Obama replacement. They intended to get a Super Scalia, i.e. more rabidly right, next year.

My, how bad was that assessment? Earlier today I saw a headline that Obama is to meet today with Republicans to discuss appointment. At the time of the passing, I allowed (although not until now in print) McConnell was out of his mind.

Here's the situation. If the RR double down on the election, what's the likely outcome? We won't know until tomorrow morning, but the candidate will be either Trump, Rubio, or Cruz. None can win the election, whether against Sanders or Clinton. A bold statement, but Trump is a wacko billionaire, and the other two are paid for by wacko billionaires. Not only is the RR certain to not win the White House, if it pushes its KKK agenda, it will lose both House and Senate; the latter with greater certainty.

So, McConnell et al face a simple wager: get a "moderate", less than Super Scalia, now or risk another Ginsberg in a year. Such a deal?

[part 2, 3 March]
As any Ph.D. quant or desperate stockmarket plunger knows, odds on an outcome change as the event's time approaches. In the case of sports events, the bookies adjust odds to keep the amount of moolah going to each outcome about equal. The same might be said of political events. If events proceed as at present, Trump becomes more unstoppable toward the nomination. As that happens, the RR has to gauge whether they can get a Trump White House, and if so whether he'd nominate a Super Scalia. Both are highly unlikely.

Here's a bit of Kristof

Me: He has a reputation as a straight shooter, but he lies. When PolitiFact was choosing its "lie of the year," it found that all its real contenders were Trump statements -- so it collectively awarded his many campaign misstatements the "lie of the year" award. And in backing him, you're pretty much guaranteeing a Hillary Clinton presidency. Indeed, because of Trump, the betting markets are now predicting a Democratic Senate as well.

Voter: Come on! Trump proved all of you pundits wrong again and again, and he'll do so again. And even those betting markets you like to cite -- they show Trump with at least a one-in-four chance of being our next president, and that's while other Republicans are trying to rip him apart. Just wait until the party rallies around Trump.
[my emphasis]

And, as expected, Obama is floating a "moderate". Take a small chance now, or a bigger one later. Brinksmanship in action.

Naming her would escalate political pressure on Republican Chuck Grassley of Iowa, the chairman of the Senate Judiciary Committee, who spoke fondly of Kelly in 2013 before the Senate voted 96-0 to confirm her for the Court of Appeals.

Not game, set, match. Too soon, but it's political waterboarding at its best.

[part 3, 17 March]
Thanks to a raging flu, I've been bedridden, so this is a couple of days late in typing.

So, in sum, Obama has taken the predicted tack: here's an old guy, with a mixed/neutral record. If you don't take it, here's your options.
1) Trump wins, and may be you a Super Scalia. But you're just as likely to get a Sherman or Taft, if Trump is pushed to make good on his anti-hedgies rhetoric.
2) Trump gets trounced by Hillary. The most likely outcome. She'll nominate a 40-something anti-Scalia, and won't back down. Constitutional crisis, or as the Rabid Right likes to say, "elections have consequences". Moreover, Trump could (still too dim) take down the Senate. Which is the main reason the RR is so intent on getting rid of him. Cruz is no better, just sneakier.

[part 4, 22 March]
Thanks to some reporting from the NYT, we now know that Roberts, not the most truthful of appointees during confirmation called for a dialing down of partisanship.

Last month, Chief Justice John G. Roberts Jr. delivered some blunt remarks about the Supreme Court confirmation process. The Senate should ensure that nominees are qualified, he said, and leave politics out of it.

Still, only the fourth inning.

[part 5, 25 March]
Well, more pressure. Chinese water torture, sort of. Or, waterboarding, if you're of The Donald tribe.
the polls building pressure.

Senate Republican leaders have said they will not hold hearings on any Supreme Court justice nominated by President Obama because they want to wait until the next president is in office. Seventy-three percent of Americans think this is being done for political reasons, while just a quarter say it is because that's what Republican Senate leaders think is best for the country.

And the Blue State Republicans are getting squeezed.

A week ago, [Sen. Mark Kirk of Illinois], who faces a tough reelection fight, became the first Republican senator to break with the rest of his conference to call for an up-or-down vote on Garland, according to reports. Kirk told a Chicago radio station that his colleagues should "just man up and cast a vote."

One out, bottom of the fourth.

[part 6, 22 April]
In the news, GOP pulling the plug on state elections. Kind of a surprise, at least to me. The Rabid Right has taken over much of governance by the simple expedient of taking over state government. Why stop now?

The lede:

The Republican National Committee is scaling back its financial commitments to some of the most hotly contested states because of flagging fund-raising, the most concrete evidence yet of how the party's divisive and protracted presidential race is threatening the party's entire ticket in November.

Here's the pressure point: they're getting ever more worried that Trump will not only lose, but take the Senate, and worse the House, with him.

This sort of unease about Mr. Trump, along with the dislike many of the party's business-oriented donors have for the hard-line Mr. Cruz, has prompted the R.N.C. to begin privately assuring donors that it will create a so-called Senate Trust fund. Money earmarked for that fund will go entirely to initiatives aimed at retaining the Senate -- including hiring field operatives and opposition researchers and bolstering digital efforts.

If events keep following this story, Republicans will have to admit that Garland is the best they're going to get.

Drip, drip, drip.

Bayes At The Moon - part the third

The Bayesians base their argument on the notion that "prior beliefs" trump all. The FDA, among others, doesn't buy it. Good on them. The latest scam to be skewered by the folks in Silver Spring is Sarepta Therapeutics, which has a muscular dystrophy drug in development. Adam Feuerstein put it this way:

In many ways, the FDA's eteplirsen review posted Friday is more negative than the one issued in January. The FDA rebuts, sometimes in sharply critical ways, Sarepta's claims the agency's original review was flawed. Most notably, the FDA is deeply skeptical of Sarepta's attempt to use a data from a retrospective, historical control study of DMD patients to demonstrate the efficacy of eteplirsen.

The use of "historical" data as the control in a clinical trial always gets researchers' hackles up. Not just at FDA, since one can find some historical data which makes a sponsor's drug look like a miracle. There's good reason to require Phase III, approval, studies to be double-blind, placebo (or, better yet, standard of care) controlled. Wasting money on pointless drugs is, well, a waste.

The Lease You Could Do For Me

It's been mentioned in these endeavors before, so it oughtn't be a big surprise to find another bit of reporting on the successor to sub-prime house mortgage: auto sales and leasing. Once again, capitalists seek to find a way to soak up all that automated output by a dwindling middle class with little money.

Earlier essays pointed out that auto loans, once merely 36 months, now can stretch to 84:

Car companies are also offering auto loans over longer periods of time. Edmunds.com found that more than 27 percent of all new cars sold in the first quarter went to customers who took out loans lasting 73 to 84 months, up from 23.4 percent in the same period a year ago.

Long distance auto loans will have some side effects:
- total cost goes up
- more profit to the lender (often an arm of the car company)
- and the subtle one; as buyers of stuff do so on long length terms, they become less eager to be economically mobile

That last needs some explanation. If your monthly "nut" grows as a percentage of your monthly income, you become more cautious. Becoming more cautious means you're less likely walk from a job that's any of illegal, unethical, uncomfortable, and the like. The big corporations win. They get more cash flow from autos. They get to sell higher priced autos. They increase the fealty of workers. What more could you ask?

Luxury makers tend to lean on leasing more heavily than more middle-of-the-market brands. For example, more than half of all Lexus, BMW, Mercedes-Benz, Audi and Acura cars are leased. Still, all automakers use leasing to some extent, and all shapes and sizes can be leased.

If you read the fine print in the TeeVee ads (be quick about it, the fine print just flashes up), the per mile fee past the "low mileage" lease terms is eye watering. More money in the car company wallet.

One of the benchmarks of the status of the middle class is per capita new car sales. Not that you'd be surprised, but that number has plunged since 1970. Here's a picture:

[give it a click if truncated]

Is it any wonder that auto companies are trying the same trick used by CountryWide (and auto companies in the past)? Don't sell on the price, but on the monthly.

Cog in a Gear

This endeavor started out both as a reaction to my being rejected for my efforts to build Organic Normal Form™ databases at my then immediately former employer, and the fact that Fabian Pascal had gone missing from the innterTubes. I believed then, as now, that Codd had defined the sole data model. Previous, and au courant, data structures are merely that; they're not models of data. Moreover, the NoSql and xml and sundry flatfile offerings were just re-hashes of pre-Codd engines. In particular, the fascination with hierarchy as implemented in xml is just IMS without any engine. And IMS, with its horrid control structures and coding black hole was specifically the motivation to devise the Relational Model. Alas, Armonk wasn't pleased, coming only a couple of years into IMS's release, and allowed one of the IMS crowd to managed SQL development. Dr. Codd was right, damn it, and somebody should stand up and say so!

Over time, my interests wandered back to where I had spent my initial career: stats and quant. Not least because it was clear early on that The Great Recession had been caused by incompetent and corrupt quant. Some insist, still, that the quants were only following orders, and thus not to blame. The fact is, though, that Country Wide's quants had a major role in devising the toxic mortgages that came to be securitized. Within the last week, both Wells and Goldman have admitted fudging the disclosures regarding those securities (but some underwriters, nee quants, created the ratings), so one can argue that it wasn't just quants.

Then along came Watson, and what IBM now calls Cognitive Analytics. It seems, at this juncture, to be a branding effort by IBM, although other sites do appear in search. The Wiki doesn't yet have a page on that term. Go to it. (Oddly, the term was trademark, but is now listed as abandoned?)

What's of interest is that Date's statement is more true than ever:

I believe quite strongly that, if you think about the issue at the appropriate level of abstraction, you're inexorably led to the position that databases must be relational.
-- Chris Date/2009

There's some controversy, still, about the notion of relations in data. The RM, on the one hand, makes relations the province of the data designer to be specified a priori; an order line must have a resolved foreign key to an order table. However, if normality/orthogonality is followed from the start, changes to schema are transparent to existing data and code. The RM/RDBMS is the only data store which provides that flexibility.

The quant (Big Data maven), on the other hand, proposes to discover, in a probabilistic manner, correlations in the data. Whether those correlations are, de jure, relations is the crux. I say, not, but then I've always been a tad rebellious (but a Blue Yankee). All of the other data structures (whether network, or graph, or ...) follow from hierarchy: the structure specifies the one optimal access path through the data. Get the structure wrong, and you're up a creek. The ignorant complain that the RM is confining, not realizing (or refusing) that in the RM, connections among "records" is expressed in data, which is fungible, while in hierarchy the connections are explicit in the tree (not fungible without some off-line effort), which the designer has pre-specified.

Which brings us to Watson and Cognitive Analytics. What's the goal? Near as I can tell, it's needle hunting in haystacks higher than Everest. Those had better be very gold needles. After all, one benefit of the RM is to be able to infer new facts from the specified relations based on data.

Somewhat ironically, at least to me, is the following sentence in the Wiki article:

Are intelligent machines dangerous? How can we ensure that machines behave ethically and that they are used ethically?

As if humans routinely behave ethically! With the rise and vengeance of the 1%, ethics don't matter much. After all, it's just business.

The notion and practice of artificial intelligence, computerized division, is generally accorded to McCarthy and LISP. Prolog came a bit later. Neither has made much of a dent in IT, although Watson is reported to use some bit of Prolog. Will Watson itself? IBM is said to be betting on it. That link, if you don't go there, is from two years ago.

Watson's only real value-add is the ability to observe, then discard, 99.9967853% of text it "sees" in very short time spans. In other words, "Jeopard!" style infotainment and outright entertainment. In a clinical setting, that amounts to ER settings where diagnostic time does matter. The machine is far too expensive to install in hospitals, so the cloud version shared among hundreds of ERs might help. Humans, experts in their fields, know to not bother with the 99.9967853% in the first place. Watson is House on steroids, which is ironic I suppose since the whole point of "House" the TeeVee show was that House the character was entirely drug addled. For day-to-day research, not so much. Collaboration with real humans is more important.

Finally, IBM is co-opting its own history. Thomas Watson, Sr. invented IBM's signature meme, "THINK" at NCR. Externally, that was supposed to tell potential clients that IBMers used their heads for something besides a hat rack. In fact, the purpose internally, was to motivate IBMers to devise ever more clever and lucrative ways to separate clients from their money. Now the meme is "outthink". We'll see. May be Watson will succeed in being a very nice suit of Emperor's New Clothes?

Gordonian Not

Well, at long last I've finished my slog through Gordon's book. On the whole, recommended. With reservations.

First, if you're among those who still believe not only in American Exceptionalism, but in its perpetual exponential expansion, you should read every page. Slowly. On the other hand, if you're among those who understand the limits of the physical world, then a more cursory approach, at least on first reading, will do. Read the "Conclusion" section of the Part I and Part II chapters. Then finish the rest verbatim. Go back and sample the full I/II chapters at your leisure or interest.

The book is an odd duck, in economic treatise terms: while there is a tsunami of data, it's only graphed or tabled. Not a regression or other inferential derivation to be found. Some graphs have an undefined "trend" line, presumably regressed but not shown. Nearly all of the data is macro, and thus suspect. We've been over that before. Just read through the full BLS (or BEA, or ...) report for either weekly or monthly unemployment to see how wonky that data is; variance large enough to make your eyes water. I'm no fan of Big Data, but stratified random sampling is really rock science and data collection from entities that have no interest in being honest, well... And this is 2016.

So, what's wrong with Gordon? Let me count the ways.

1) He spends most of the book describing how the consumer experience has changed over the period of interest, 1870 to 1970. Whether that's progress is a real question. Some parts of his description qualify: the notion of "networked" housing; having gas, electricity, piped water, piped sewer, and telephone is significant observation. On the other hand for instance, he attributes "progress" to changes in the nature of TeeVee, from tiny, round CRT circa 1950 to today's 40" LCD Leviathan. This isn't progress, since it's still just TeeVee. The main effect of this change is to shift entertainment dollars from cinemas and such to TeeVee. Much of what he talks about comes under the rubric of "psychic income"; the new version of a widget is smaller, more expensive, uses more power, is more automatic, etc. so it must be "better". Tell that to the vinyl junkies who wouldn't touch a CD, even to use it as a Frisbee. The notion that the PC has more "power" than mainframes of 1970 is also apples and oranges: what makes a mainframe useful isn't its cpu, but the whole infrastructure. Today's IBM mainframe cpu is built, mostly, using IBM's microprocessor tech.

2) He buys into the Moore's Law meme. Here, too, is mostly psychic income. While the PC did change work life, it didn't really change it as much as most civilians believe. Yes, the Big Three white-collar office applications (word processing, spreadsheets, and word processing) were conducted differently. Again, not as much as one might think. In a footnote, he remarks that his mother had written an entire book with MultiMate from 1975 to 1977. That's physically impossible, directly. MultiMate (an application I taught in one of those vampire squid for-profit computer schools in the late 80s) was an unauthorized clone of the Wang (mini-computer/ChUI-terminal based) Word Processing System; it didn't exist before the IBM/PC. More to the point: PC based word processing didn't really become important until PC supporting networking was created, with NetWare. To make matters worse for Gordon's thesis, studies which demonstrate that GUI-fied office software make offices more productive are few and far between. None that I can recall. More than a few reflect my experience: WYSIWYG office automation applications lead to folks worrying about the sizzle and ignoring the steak. The London Whale did his damage running an Excel spreadsheet he didn't create nor understand. "I have no idea what this memo means, but I do like the font I chose."

(aside: also in the footnotes is one where he attributes the repeal of Glass-Steagall provisions to Clinton. Wrong. The legislation was the Gramm-Leach-Bliley Act (rabid rightwingers, all) which was not from the administration, but those three. Clinton signed it, only.)

3) He pays insufficient attention to physics. The underlying tech which has driven American Exceptionalism is filling out the periodic table, the understanding of the Bohr atom (which supports how molecules are constructed from each new atom), and an unmatched level of natural resources, chief among them fossil fuels. But iron ore and the soil of the Great Plains matter just as much. As the periodic table was filled in, new elements became available. Everything from light bulbs to microprocessors could only happen with a complete periodic table. And only with fossil fuels could aluminum, for example, even exist.

4) He ignores, for the most part, that growth comes from physical capital, but all of the data he cites is fiduciary capital. Most economists make the same mistake. They're just accountants with phat heads.

5) He ignores the forced cost increase for some of these "innovations". While cable/sat TeeVee offers up more variety, if "Swamp People" et al qualifies, the cost of access is not trivial. Same for cellphones.

6) He pays too little attention to the effect of distributional decline; not that he ignores it, but he doesn't make the connection between growth and distribution. An "Economist" article:

They reckon that when growth falters, inequality is often a culprit.

Growth, by definition, has to be demand driven. Say was an ass. Trump, trickle-downer in chief, keeps harping on 20% unemployment, but his agenda is to increase income and work inequality. That would help, yeah. Even some of the so-called Progressives insist that USofA should be more vigorous breeders. WTF??? How does having ever more, ever more poor, mouths to feed help in paying for keeping old foggies alive a few months longer? It won't, of course. Growth is driven by demand, and viewed in a long historical lens, growth appears to be driven by population. When most goods were consumed broadly, and made similarly, more bodies meant more demand. But with increasingly inequality, more bodies has much less effect. And the 19th and early 20th centuries were times when much of the world's resources were not even yet known. Now, said resources are under stress, not least water; although overloaded California reservoirs is mostly about skewed distribution of the El Nino effect. For most of that time, production of stuff necessitated more humans to do the work. Not so much now and in the future. Growth can only come from a vast Middle Class, spending money on stuff. Or a proliferation of goods and services only for the 1%. Which is to say: the end of mass consumption. Which is to say, further, the end of mass production. That would not be good for capitalists who've automated most labor out of production, no matter what they may think today.

7) He pays too little attention to why real interest rates remain so low. Only physical investment creates new value; fiduciary investment (and that includes real estate) is just stealing from Peter (someone's current consumption) to pay Paul (someone else today or tomorrow). The answer, of course, is that the CxO class can't find useful ways to turn moolah into machines. The knock-on effects of that are far more significant than all the historical data and inequality arguments combined. It means that individual savings can't earn anything resembling what the insurance and banking TeeVee commercials imply. Giving moolah to corporations that already have more moolah than they know what to do with is stupid. It's just hoarding money and reducing demand by reducing consumption. That will end badly for all except the .1%. Dee Feat is in Dee Flation.

So, on the whole, worthwhile. But, on the whole, doesn't get to the nub of why progress is slowing down. And those reasons are:
1) humans know about all there is to know of the macro world; there's no more elements to discover

We wanted flying cars, instead we got 140 characters.
-- Peter Thiel
[epigram to chapter 17]

2) the failure of progress too closely tracks the destruction of the middle class to be coincidental

As mass production has to be accompanied by mass consumption; mass consumption, in turn, implies a distribution of wealth -- not of existing wealth, but of wealth as it is currently produced -- to provide men with buying power equal to the amount of goods and services offered by the nation's economic machinery.
-- Marriner Eccles

Liar, Liar

The rabid right wingnut Republicans keep saying that the Damn Gummint caused the Great Recession. It was, in fact, caused by banksters just as the Great Depression was. Now, we get a bankster admitting it fudged the numbers. As usual, motive and incentive™ trumped the data.

According to the settlement, Wells Fargo "admits, acknowledges, and accepts responsibility" for having from 2001 to 2008 falsely certified that many of its home loans qualified for Federal Housing Administration insurance.

The bill?

Wells Fargo & Co admitted to deceiving the U.S. government into insuring thousands of risky mortgages, as it formally reached a record $1.2 billion settlement of a U.S. Department of Justice lawsuit.

My Argyle Socks

Overture:

The capitalist's dilemma is simple enough. Samuel Colt is generally credited with devising the interchangable part method of production. Being the USofA, he made guns, of course. Decades later, Henry Ford devised larger assembly lines to make the Model T. In both cases, the notion was to replace master/journeyman craftsmen with minimally skilled anonymous hands. In the case of Ford, the savings in labor expense was passed on (largely) to the consumer in a lower price. The Model T was the Yugo of its time, so it was not an apples-to-apples bargain vis-a-vis a Cadillac, say. Both shared one characteristic: the physical work didn't change much from past practice; a human attached a part to a unit-in-process. Little to no adjustment was needed, however, thus the skill substitution.

Modern robotic/automation/computerization/fairy dust production is very different. Rather than just moving the unit-in-process from worker to worker, now much of the process is carried out by machines. This has a profound effect on the business. Gordon goes off the rails big time when he repeats/re-uses the classic capital/labor ratio of Solow: 70% labor 30% capital. By far and away the stupidest idea in the book. We'll look at the numbers, to the extent that any are available, with the recent announcement that Ford (them, again!) will build an assembly plant in Mexico to exploit cheap labor and export cars to our Blue States.

Today's high machinery production poses a real problem for the capitalist: no matter how many units you produce, you still have to pay for the plant and equipment. You can't layoff a robot. High machinery production is fine and good, for the capitalist, when demand is at least steady. Any fall in output raises the imputed capital cost per unit. Oops.

My family was poor, but unhappy, before, during, and after I grew up there. La Madre did little to aid the poverty, preferring to sit and smoke Parliaments. She did, however, do a bit of sewing. Mostly, clothes for herself and my sister, much younger than I (my brother and I never figured out how that happened :) ). We got a shirt or two along the way. La Madre wasn't a master tailor, so she'd go to the fabric or department store (back when they had notions departments) and buy a pattern. Butterick is a brand I remember. They were packaged in an envelope about the size (H X W) of a standard textbook. Dress patterns are printed on the thinnest paper imaginable. Pity the poor child that should tear one.

Pattern recognition and pattern matching are the essence of intelligence, human or artificial, and mean that, given sufficient cycles and memory (you do know how many bytes can be addressed in a 64-bit space?; no, your i7/motherboard doesn't generally implement more than 36 lines) any problem can be solved by brute force. I give you Watson. And the FBI's success in finding the passcode to that iPhone.

One of the patterns that growth economists, especially the Freshwater sect, like to prance out whenever the income/wealth inequality issue is raised is the historical record. Which record demonstrates that poor, ill-educated blacks and whites migrated from The South to The North post WWI through post WWII to work in factories. The notion is that they traded low skill work for higher skill work, and thus better income. I've never bought that, since bolting on a tire to a Model T requires no intelligence at all. These days such a task is often done by robots. Being a farmer requires a much higher degree of skill, particularly when you're not getting vast subsidies from Uncle Sugar. The assumption that there will always be more higher skill/wage jobs available when/after capital destroys current ones is just an assumption. Convenient to the capitalist.

All of which brings us to a Sorkin piece today on the assault on finance by automation. More than once, in these endeavors, it has been mentioned that much of the employment of all those "high value" educated folks has been in FIRE, and not real science and engineering. Of course from a macro point of view, FIRE is pure overhead, not production. So, what does the future hold? If the major uptake in "high value" education has been in overhead occupations, and these occupations are under assault by computerization, isn't dystopia the likely result?

Some 800,000 people will have lost their jobs at financial services companies to some of the newly dreamed up software in a decade, the report said. "Roughly 60 to 70 percent of retail banking employees are doing manual-processing-driven jobs," the report explained. "If all the current manual processing can be replaced by automation, these jobs can disappear or evolve."

"Do you want fries with that?"

Silicon Valley has long shunned regulated industries, but having conquered so much of the landscape in other industries, it is now turning to finance.

I guess that's where all those now redundant MBAs will end up. The stories from years ago, snickeringly so from such MBAs, of Ph.D. sociologists driving cabs may well cut a bit closer to the bone. From a macro point of view, it's not a funny story. If we've been employing our best and brightest in overhead occupations, rather than productive ones, and even those overhead jobs go poof, what then?

Keep in mind the M&A fury of the last few years is fueled by all that trickledown money from TARP, QE, 0 interest rates. It went to corporations sitting on $$$trillions they haven't a clue how to invest in real physical capital. So they buy themselves and their competitors, better to extract consumer surplus.

On to Ford. Here's one report. And, of course, The Donald used the announcement to polish his populist bona fides:

Mr. Trump was quick to comment on Tuesday, saying in a statement that Ford's plan to build a new factory was "an absolute disgrace."

The capitalist zealots will counter with something along the lines of, "well, yes 2,800 Americans will be made redundant, but X00,000 new Fords will be cheaper for the Blue State folks who still have jobs. Seems a fair trade, don't you think?" Well, they may not be quite that honest, of course. Whether the Fords produced in Mexico will actually be sold at a lower price than if they were made in Michigan is an assertion with no evidence. The price of a Ford is determined by demand. If they make a hit (think: original Taurus), then they'll keep the cost savings for themselves and charge as much as the consumer will bear. If it's an Edsel, not so much.

Ford, the nation's second-largest auto manufacturer, said it would invest $1.6 billion to construct a plant that will employ 2,800 workers and begin making vehicles by 2018.

Now, let's see if we can ferret out the capital/labor ratio for such an investment. Here's an article. 300,000 units, eyeballing it, looks like an average. So, $1.6 billion over a five year straightline depreciation gives us $320,000,000 per year. 2,800 workers at $15/hr at 40 hours/week and 50 weeks/year yields a labor bill of: $84,000,000. How much of the $1.6 billion is plant and how much robots isn't said. We'll treat it all the same for this exercise. So, each Ford will cost: $1,066 in capital and $280 in labor. Add in material, of course, and electrics to run the lights and robots. And so forth. It certainly appears that labor hours/unit isn't all that much. Hell, the plant manager's annual bonus is likely more than the entire labor cost for the year. :)

Here's the bedtime story:

The difference is that back then, businesses were actually spending that cheap cash, buying equipment, building factories and hiring workers. Today they're just hoarding it. The pile of corporate cash on the balance sheets of nonfinancial companies has grown to $1.48 trillion, according to Moody's. That's an 81 percent increase since 2006.

So, more evidence that Big Capital is buying itself and competitors. But, you knew that anyway.