My extended family was all engineers and computer programmers, with no (successful) entrepreneurs to speak of. I tried to buck the trend. I got an MBA (finance and quantitative methods), worked in an accounting office, studied for the CPA exam, considered the CFP exam, and worked in a large corporate finance systems department. Heritage won out and I slipped back to EE and CompSci.
Then my daughter got a BS Econ, and I read her texts. That got me curious, and I picked up some texts on econometrics et al. So while I'm no expert, I can provide an introduction to the world of Economics.
Economics prides itself on being a science, with rigorous mathematical disciplines. This is all hogwash. The experimental specimens are human beings, their behaviors are mediated by individuals' real and perceived needs, and true experiments are immoral, unethical, and impractical. Economics has often served as mumbo jumbo to lull the masses into accepting one scam after another perpetrated by the big boys.
The problem is that econ deals with large numbers of units, each of which has extremely complex and often distinct behavior. Econ tries to model this with techniques appropriate to large numbers of identical units. What works for molecules doesn't always work for humans.
Like every other field, economics has been transformed by computers, both for data collection and for modeling.
Finally, once you have a good model of human economic behavior, you can use that to:
A physicist, chemist and economist are stranded on a desert island. They find a can of beans..
Physicist: Put it in the sun. The heat will make the contents expand and burst the seams.
Chemist: Put it in the sea. Corrosion will breach the walls.
Economist: Assume we have a can-opener.
A typical text:
David Besanko, Ronald R. Braeutigam. "Microeconomics: An Integrated Approach". John Wiley & Sons, 2002. ISBN 0-471-17064-X.
The idea is to model a rational unit's behavior in a well-understood environment. Each unit has various needs and desires. These can be obtained in various combinations. The unit's problem is to obtain the combination which best suits its needs, at costs that it can afford. That is defined by a utility function or optimization function.
To make the model workable you need to translate everything into common units -- such as money. You must assume everything can be bought/sold in those terms -- no "commons", no "spiritual needs", no love, altruism, hatred, etc., unless they have a monetary value.
If you put together several such units, you can have competitive strategies. You can do game theory. You can have monopolies.
Does microecon have any value? All else being equal, it does help a business position itself in the market, set a price for its goods, etc. It does not take account of lobbying for earmarks or rezoning, back-room deals for or against your firm, storm damage, etc.
A typical text:
N. Gregory Mankiw. "Macroeconomics", 5th ed. Worth Publishers, 2003. ISBN 0-7167-5237-9.
Somehow all those competing units get together and operate as entire nations or branches of industry. We have no idea how to make that leap, so we just start off fresh, assuming massive roll-ups of the data. Balance of payments, trade deficits, GDP, business cycles.
Again, you have to measure everything in terms of monetary units. When rolled up we get oddities like Gross Domestic Product as a measure of well-being. If we all live on garden-of-eden farms, eat our own produce, weave our own clothes, and use money only to buy peppermint candies at the holidays, then we have a low GDP and are poor. If we live in crowded tenements, work in sweatshops, buy expensive-but-unhealthy food and then pay more to repair our health, then we have a high GDP and are rich.
Does macroecon have value? All else being equal, it can at least let you know how you are doing as a collective body (e.g., a nation). That does not take account of vastly wealthy CEOs vs sweatshop laborers, or unpaid overtime, or undocumented workers.
An excellent text:
Frederic S. Mishkin. "The Economics of Money, Banking, and Financial Markets", 7th ed. Addison Wesley,2004. ISBN 0-321-12235-6.
You may have noticed that everything in micro and macro depends on a trusted common denominator -- money. What if money itself is in play? The whole micro/macro world would be in limbo. It could be manipulated by anyone who controlled the value of money. You would have interest rates, exchange rates, financial markets, derivatives, banking systems, etc.
Does financial econ have value? All else being equal, you can "clear markets" with only small arbitrage profits, and let micro and macro worlds continue on their way. That does not take account of inefficient markets where insiders make deals among themselves at the expense of everyone in the "real" economy.
It has been said that we have a new class warfare: financiers vs everyone else. Gordon Gecko. Masters of the Universe.
An excellent text:
Peter Kennedy. "A Guide to Econometrics", 5th ed. MIT Press, 2003. ISBN 0-262-61183-X.
"What distinguishes an econometrician from a statistician is the former's preoccupation with problems caused by violations of statistician's assumptions; owing to the nature of economic relationships and the lack of controlled experimentation, these assumptions are seldom met." [pg 1]
As noted above, the std approaches assume away a lot of problems. Econometrics tries to get the game back on a sound footing. In doing so practitioners have become quite rigorous in their technique. It has been said that econometricians use math more carefully than physical scientists precisely because they cannot be warned off a flawed hypothesis by experimental findings.
For that reason, and because this particular book is exceptionally well-written, this is a valuable text for anyone dealing in mathematical models.
On the other hand, this attention to method over getting good empirical data makes this akin to very-well-done astrology.
Clearly, everyone was having trouble getting real data. Companies relied on gross orders. "Hmmm, I see we sold 3,000 cases in December. Wonder if that had to do with the holidays or the snow storm". They reported sales and inventory to government clearing houses where economists used it to build models. "Snow storms increase use of soda pop."
Barcodes, RFID tags, and electronic payments have changed all that. A community goes shopping and the data starts flowing. Buried in the data is information that a redheaded woman got a pregnancy test at the same time she purchased a Guns & Ammo magazine. A man with prostate problems and a taste for porn is looking for a used Jeep.
"Buried" is the operative term. Corporations, esp, the credit agencies and clearing houses, have terabytes of data and more coming all the time. Making sense of it is the hard part. This is called Data Mining.
A good text:
Jiawei Han and Micheline Kamber. "Data Mining: Concepts and Techniques", 2nd ed. Elsevier, 2006. ISBN 978-1-55860-901-3.
Assorted Artificial Intelligence and indexing and natural language techniques are used.
Even without fancy algorithms, if you have the data, you can see there is a growing gap between rich and poor. As George Soros said, there was a class war, and my class won. On the political stage, David Sirota says there are two parties: the Money Party and the People Party. Basically all Republicans in the Senate and House (and local legislatures) are spokesmodels for the Money Party, as are some of the Democrats. As a practical matter, pretty much all legislators at the national level are members of the Money Party. You can determine this by checking their campaign contributions and "independent expenditures", and then by checking their voting records.
Why is this relevant to data mining? If you have access to that data, you can know everything about everyone. Wealthy people have access via their corporations. Poor people do not -- they can't afford it and don't even have legal standing to get it.
You can be darn sure data mining is used to advantage come election time. The only way to beat the system is to pay with cash. Which is why the Republicans and some Democrats are calling cash payments a "terrorist technique" and insisting on traceable cash.
Standard modeling, based on theories of aggregate behavior, is suspect even if we have better raw data. Humans are just too complex individually or in groups to be so simply modeled. What if we just put a bunch of individuals in a box and watch what happens?
Better yet, do it with computer models, with hundreds of thousands of "agents". Provide each "agent" with as much knowledge, emotion, intelligence as you think appropriate.
This is a fairly new world, so texts are really compendia of reports, journal articles, etc. E.g.:
Gerhard Weiss. "Multiagent Systems". The MIT Press, 1999. ISBN 0-262-23203-0.
Also, one needs the std texts on Artificial Intelligence and on knowedge bases.
Obviously, there is opportunity for a revolving door. The key is that a good agent system in academe might get a PhD or a grant for computer time. A good agent in the financial markets can make a few extra percent on billion dollar portfolios.
Once the markets actually had lots of agents moving real money around, the agents of course had to do game theory on other highly intelligent agents. The humans doing the programming ("rocket scientists", and then "quants") spend their days determining new algorithms. Automated agents do the actual trading (no emotions, potty breaks, sick leave, or slow reactions).
The "investor" is the guy who put up the trading capital and picks up the rewards.
Invent whole families, with needy babies, impetuous teens, hardworking adults, and failing elders. Let them work in shops and factories, get sick, play hookey, go to night school, etc.
Let it run a few simulated years and see how it looks. Tally up a company's profits when it provides daycare. Redo the experiment without daycare.
To tune your models, compare the results with the best real world data you can get -- provided via data mining. E.g., tune for:
Each agent lives its own life, according to its own needs and opportunities. The economist is interested in the emergent behavior -- the aggregate GDP or wage rates or car sales, etc.
DUAL USE ALERT Dual use is when a commercial, non-military technology can also be used for military purposes.
Suppose you have a community model that does a good job of modeling real communities and their reactions to threats of layoff, or to bonus payouts.
Now introduce "terrorists". Have them talk to neighbors about resisting "the system" instead of about stock tips. Estimate the number of additional toilet flushes needed if a terrorist is hiding in a home. Detect hidden females by the number of tampons purchased in a given month. Compare this model to real measurements taken in the sewers outside suspects' homes. ("Your honor, they flushed it and we snagged it outside their property so it is fair game. No warrant necessary.")
Suppose you have a good model for communities and terrorists. You have folded in real life data via data mining and PATRIOT ACT-approved checks of grocery purchases and library loans. Add in travel patterns picked up from traffic "stoplight cameras".
It works fine for tracking anyone who is even suspected of maybe thinking of becoming a terrorist.
An election is coming up. Your boss wants to know if there is a grassroots movement to dump a friendly (to him) senator. Your boss asks if you can detect a pattern or even specific people.
Your career, your family's safety, and the future of the nation rests on your answer.
With sufficient raw data, we begin to realize that the normal aggregate rollups are misleading. If we look at salary income, some of the wealthiest people on earth look quite modest. If we look at net worth, but do the rollup in deciles, the top 10% looking vaguely better off than the bottom 10%. At finer resolution, the top 1% is notably better off, and the top 0.01% are pulling away from the pack. If we narrow to the wealthiest 10,000 people on Earth, we are looking at the top 0.00001%, and we see they own or control or at least influence most of the world's resources and governments and armies.
At this point we can no longer apply statistics and models -- the "law of large numbers" is not applicable. Instead, we are looking at specific people, and their specific actions. As it happens, they don't want us to know what they are up to.
SEC filings theoretically provide that data to the business owners (stockholders). These are often misleading. Financial journalists make their careers ferreting out the truth. Taking a company private removes even that impediment.
Individual and corporate tax returns theoretically provide that data to the government. But unfunding tax auditors and moving headquarters offshore remove that impediment.
Net result is that anyone trying to seriously understand local, national, or world economics must take into account backroom deals among powerful players. If the discussion or the deal includes illegal components (e.g., enticing a nation to go to war, or getting a "spot" rezoning), then it is by definition a conspiracy.
Of course, anyone allowed anywhere near the table is likely well-vetted for willingness to keep quiet. If the deal stinks so bad someone does speak up, it may trigger career destroying multi-media attack, blacklisting, etc. Or an "accident".
What we have left are a few blips on the public radar (Jack Welch's divorce settlement, Enron's meltdown), plus the efforts of "conspiracy theory" muckrakers. The former get 15 minutes of fame; the latter write websites or books published by second-tier publishers. I consider these to be forms of alternative_media.
If theory, law and custom declare that only owned and monetized goods and services have value, then "common resources" such as air, open seas, and untamed wilderness are "externalities". Everyone can exploit them as best fits his/her utility function. If the resource is in fact limited, then extended exploitation will degrade the resource. The first exploiter gains an advantage; later exploiters make do with the dregs.
This dynamic drives the "Tragedy of the Commons", first recognized as such when sheep overgrazing destroyed the English "commons" (unowned grasslands). We have since learned that air, rivers, aquifers, oceans, and ecosystems -- despite appearing unlimited -- can be destroyed by humans equipped with powerful technologies.
Unfortunately, unlike the original commons, we can't just fall back on our own pastures. These global commons ("the environment") are essential for human-sustaining life. Therefore a branch of economics has developed in an attempt to measure its monetary value, and apply std cost-benefit analyses to public decisions.
Once the subject is broached, it is immediately clear that just about anything technology-aided humans do has an adverse impact on the environment. Laws and policies can attempt sustainability (the environment is hanging on by its fingernails) or restoration (pre-industrial-age conditions).
There is of course a backlash, with some people declaring "I can do anything I want on my own land". To which a biologist might answer,
As you can guess, economics has been enlisted on both sides of the debate. Here are some players:
Unfortunately, it appears monetizing common resources is a losing game. Under economic models, high-intensity exploitation will always beat the diffuse common need. Sometimes this comes in the form of campaign contributions, outright bribes, or high-priced lawyers. Sometimes it comes as massive PR campaigns ("Save our jobs...let us pollute just a wee bit more").
The only effective stance appears to be based on community religious values. Save the redwoods, not because of the tourist value, but because they are spiritually magnificent. Restore salmon runs, not because wild-caught is $2/lb more valuable, but because teeming streams are glorious to behold. Stop knocking mountaintops into streambeds to get at coal, not because hilly land is more valuable than flat, but because those wooded hills and valleys are part of America.
[IMHO=In My Humble Opinion]
The markets are quite efficient, except for:
This certainly goes on, but it is hard to be the insider on everything, so it isn't much of a life-long strategy. Sure, pay a spy network to get you a bit ahead of the game, but recognize that other players are also paying for spies. So there are two tiers: The big boys with private information channels, and everyone else who relies on public knowledge, such as newspaper stock reports or SEC filings.
Because true insider trading beats the market in non-repeatable ways, it beats even the big boys. It doesn't make them happy. So while it goes on, if it gets too profitable (i.e., too expensive to the other big players), the big boys will stomp on you.
2009-11-01 News report show that hedge funds really can get insider information (and profit from it) on a consistent basis. No one will come harass them -- they are the big boys.
This can be lobbying a government to rewrite laws, or supporting candidates who when in office will do your bidding. It can also be lobbying CEOs so they will buy your products. Or, if lobbying is too slow, then get photographs of the decision makers in compromising poses with inappropriate partners. Or get them hooked on a steady flow of bribes, and then threaten to call the cops on them. These are all std ploys for recruiting and handling an "asset in place".
The net result is that the decision maker institutionally trusted to make decisions for the good of the group does not do so. The very basis of market theory breaks down.
Or, as some economists have tried to do, you can count the bribes as part of the cost of business. It has been said the ROI on lobbying is 4000%. Certainly the number of registered lobbyists in Washington DC suggests the ROI is extremely high.
This works until the population at large catches on and realizes they are living in a plutocracy, not a democracy. They can then accept it or revolt. Avoiding or crushing democratic revolts is also a cost of business.
"Sunshine laws", "Freedom of Information Act (FOIA)" requests, "Open meetings", and a diligent and non-embedded free press are the main ways to detect these maneuvers. Honest elections (for governments and for stock-issuing companies) are the main way to turn this knowledge into reform.
It is no surprise that those profiting from favorable treatment are opposed to the mechanisms necessary for detection and correction.
The markets may be efficient, but they do not "clear" instantaneously. If some players have fully automated "agents" which react within a few seconds to any system shock, they will consistently outperform the vast majority of investors.
The financial markets follow a sawtooth pattern -- a slow rise with efficient clearing on the way up, and then a rapid fall on bad news with inefficiencies on the way down. A fast-reactor needs a continuous stream of those downward shocks.
Earthquakes and hurricanes, and droughts are good but not consistent. Manmade catastrophies such as breached levees and terrorist bombings are more reliable. It pays to keep a few terrorists around to shock the market whenever you need a new infusion of profits.
If you can parlay natural and manmade disasters into also clamping down on "detection and correction" of favorable treatment, then all the better. And if you can arrange for your buddies to do a bit of insider trading along the way, even better still.
Creator: Harry George