by MrBill » Mon 24 Apr 2006, 08:15:14
$this->bbcode_second_pass_quote('Doly', '')$this->bbcode_second_pass_quote('Steven Bratman', '
')Because it isn't possible to change just one variable in an economy, nor to try an experiment twice by starting an identical society under new rules, economic analysis is really the equivalent of observational evidence (economic records) combined with plausible reasoning (an economic model). In medicine we have seen that observational evidence combined with plausible reason can lead to conclusions that are the exact opposite of truth.
Don't insult medicine by suggesting that it's anywhere as unscientific as economics!

Medicine has produced some admirable results in the last century. Life expectancy has significantly improved. Can economists say that the economy is significantly better from following their advice? Let's just say that the evidence is doubtful.
What's even worse, when medicine uses scientific tools, such as statistics, they take care to use them properly (carefully choose subjects so they are representative, double-blind studies, etc).
What about economics? You'd expect that, money being essentially numbers, their maths would be sound. I've been reading lately economics textbooks and, as a mathematician, I'm scandalized. They use nineteen-century maths and generalize with abandon, ignoring well-known results that prove that those generalizations are invalid. Very few economists are even considering applying chaos theory, when it's the obvious choice.
To give a relevant example, I read in an advanced economic textbook a model of how dwindling resources would affect growth. The model was ultra-basic and was analysed using 19th century maths. The results were predictably lame: it couldn't explain how oil prices could have gone down during the nineties to lower levels than before the seventies adjusting for inflation. The funny thing is that the authors of the "Limits to growth" books, who don't know a lot about economy, but know quite a lot about modelling, give a perfect explanation of why this happened: because there are significant delays in the oil industry between the moment when oil prices go high and the moment that new projects start pumping. The high oil prices in the seventies caused lots of new projects that went on line during the nineties.
I could go on rambling, but that's enough for a start.
Sorry Doly, not to pick on you specifically because I think you are all full of shit!
So what are you doing with your high level maths background? IT support? Why are you not working for some investment bank in London doing mathmetical modelling of interest rates and foreign exchange, and outperforming all those know-nothing idiots that have not got a clue? At least writing some scientific papers pointing out these other authors flaws?
I cannot say? I have always struggled with maths myself. Any standardized test I have ever taken (SAT-equivalencies or GMAT), I have always consistantly scored in the bottom 50-percentile. Who cares? I have a Statistics for Management & Economics textbook, which with an Excel spreadsheet has all the computing power I need to churn out multi-regression analysis and back test my models. More than 90% of any nineteenth or twentieth century mathmetician and a lot faster too. You do not have to be an Einstein these days.
Interpreting the models and understanding the underlying assumptions is more important that being able to build them or to be able to speak in equations, especially if you happen to be visually oriented and prefer graphs and charts.
A few years ago I had to evaluate a few PhD's quantitative models for predicting foreign exchange rates. I did not look at their math I looked at their underlying assumptions. Too few observations and a readily apparent autocorrelation made their model prone to directional errors, and their use of historical quarterly data made their model hopelessly out of date for predictions about future rates. For example, they used 5-years worth of quarterly data, which is only twenty observation periods, and was not long enough to encompass both bull and bear trends. And they looked at currency pairs like USD/CHF, USD/ITL, USD/DEM, USD/FFR, USD/NLG, etc., which are not independent, but were serially correlated as during the observation period those currencies traded as DEM/ITL, DEM/FFR, DEM/NLG, etc. and therefore moved together along with USD/DEM or USD/CHF if you will.
Their model was junk. Not because their maths were wrong, but because they did not understand the underlying market they were looking at and how it operates. That is the problem with most academics and it takes a practioner to point out their errors. Even one as dumb as I am. However, until you have a PhD in academia you do not exist as far as professors are concerned. That is why when I publish it is always with another doctorate. Otherwise I cannot get into top tier journals.
Oh, and by the way, most medical doctors are also useless. They also cannot experiment on live patients and do double blind experiments. And they are the last to admit they made a mistake. Any research is usually done in research hospitals and by drug companies not your average GP off the street. My sister is a pharmacist and she catches so many prescription errors all the time. Like heart medicines and blood thinners taken together that would cause patients to internally bleed to death. Not to mention expensive medications that could be replaced by drinking more water, or controlling certain conditions through diet and excercise rather than by popping pills.
The best advice I can give you is be your own doctor, lawyer, accountant and banker, as if you cannot catch the errors of the professionals or specialists they will make them and you will be no better off and often a lot worse.
p.s. Levitt & co. Freakanomics was an amusing one-time read. Nothing new. Nothing revolutionary. The South Beach Diet for the non-initiated.
The organized state is a wonderful invention whereby everyone can live at someone else's expense.