Uncommon Sense. Pape Scott
Чтение книги онлайн.
Читать онлайн книгу Uncommon Sense - Pape Scott страница 8
The gold standard for experimental study is the controlled study. Two systems are set up that differ in only one respect – the variable being tested for. Both systems are put into effect so the scientist can observe and measure the outcomes resulting from differences in just that single variable. Data is then collected from a series of experiments and tested for statistical significance to establish whether there's a high probability of a cause-and-effect relationship. If other researchers produce similar results when they repeat the experiment, the hypothesis is considered likely to provide an explanation for the observed outcome.
You can't perform these types of experiments in economics. A simple example will help explain why.
Let's say you're testing for the pull of gravity on objects of different mass. To do this you release a feather and a lead ball from the same height at the same time. Which one hits the ground first? Answer: it depends. If you perform the test in your living room, there's no doubt the feather will hit the ground long after the lead ball has punched a hole in your shagpile. But that's because there are two forces here. As the Italian physicist Galileo demonstrated four centuries ago, you aren't testing for gravity alone. There's also the opposing force of air resistance, and air resistance slows down the feather much more than it does the lead ball. It hasn't been a fair contest.
Now perform the experiment in a vacuum, where the effect of air resistance is eliminated. This time the feather and the lead ball hit the ground at the same time. You are now testing for just one variable – their mass. And you have been able to determine the relationship between mass and acceleration under gravitational force. Apollo 15 commander David Scott performed this test in 1971 on the surface of the moon. He dropped a feather and a hammer, and, due to the lack of aerodynamic drag, they hit the lunar surface at the same time. But what if you hadn't been able to test in a vacuum or on the surface of the moon? The effect of gravity on objects of different mass wouldn't have been clear at all.
That's the problem in testing economic relationships. There are too many variables in the real economy – not just one or two but a near infinite number. And you have no way of isolating them in order to test the impact of changes in each. Simply put, you'll never be able to set up an economic laboratory to test for changes in trade flows, fiscal stimulus, population or any other economic variable you wish to study.
Which all means economics never gets much past the storytelling stage. The problem is that a story, repeated often enough, starts to sound like a fact. Economist Milton Friedman warned us of how many different stories can be fabricated to explain any set of circumstances when he said: ‘If there is one hypothesis that is consistent with the available evidence there is an infinite number that are.'
This must have been on Sir Isaac Newton's mind when he was describing the law of universal gravitation. Newton outlined his gravitational formula in Philosophiae Naturalis Principia Mathematica (‘the Principia'), first published in 1687. And in the second edition, published 26 years later, he added the declaration ‘hypothesis non fingo', which translates as ‘I feign no hypothesis'. Newton had observed gravity and proposed an equation enabling the prediction of how objects would move under its influence, but he had no explanation of what actually caused it. He chose not to guess what might be behind it all for risk of being proven wrong later on. He didn't want to risk having proverbial egg left on his professorial face.
What about the branch of economic study called econometrics? It applies statistics to economic data. That's a bit like trying to accurately measure the physical dimensions of smoke. Which reminds me of that old economic adage: there are two things you don't want to see in the making – sausages and econometric estimates. But for those of you looking to explore econometrics further, I refer you to Ed Leamer's 1983 article ‘Let's Take the Con Out of Econometrics'.
Science is by no means perfect either, but it's harder to get away with as much artistic licence as in economics. But, hey, I'm not the only one who thinks this way. If you want to do some further reading, look out for The Pretense of Knowledge by Friedrich August von Hayek or ‘Science as Falsification' by Karl Popper.
A SOCIAL SCIENCE
Another monumental difference between economics and science is the human factor. Scientific laws are unshakeable but economic outcomes are influenced by human sentiment and behaviour. The imagined future can affect the present, and thereby influence how the future turns out, which means economic researchers are studying themselves. I've heard it said that this makes economics a bit like studying the movement of ‘billiard balls with eyes'. The rules of physics describe how two billiard balls interact as they crash into each other and the cushions of the billiard table. But consider how a billiard ball would move out of the way if it was capable of seeing another ball approaching!
In deference to others, I acknowledge that my views are not held by all. Consider these words taken from The Next Great Bubble Boom by Harry Dent Jr, published in 2004, containing cover-to-cover stock market and economic predictions:
Today in economics and in many fields of politics, sociology, and science, there is an attitude that ‘nobody can predict the future past a certain point' … But this is obviously nonsense as more and more fields of science have become capable of predicting more phenomena for centuries as our knowledge of the universe has grown exponentially.11
Clearly this author is confused. He argues that our ability to make economic predictions has improved as a result of progress in the ‘fields of science'. Interestingly, this book was released not long before the GFC, yet it failed to predict it. (No doubt this will be fixed in the next edition.)
Famed US economist John Burr Williams was working as a security analyst when the 1929 Crash decimated stock portfolios around the country. He later told of the experience, noting the price movement in a stock called American & Foreign Power. Pre-crash it traded for 199¼1 (100 times historical earnings). Post-crash it plummeted to a low of 2. Like pretty much everyone else, Williams hadn't seen the Crash coming. He felt this was due to an inadequate appreciation of the forces driving the economy. It prompted him, in 1932, to enrol at Harvard to obtain a PhD in economics. Someone, he figured, must be able to explain to him what had caused the biggest collapse in stock market prices ever seen and the ensuing Great Depression. He later reported that he never found the answer he was seeking.
Want to summarise the whole macroeconomic question in a few lines? Economics presents an interesting study. It assists us in thinking about and articulating important issues that impact our lives. It measures and reports on historical information in an interesting way. But economic models fail abysmally in either describing or predicting the real world.
Chapter summary
• Economics doesn't lend itself to experimental rigour. Therefore most economic concepts can be neither proved nor disproved.
• Economics is largely a social science.
4
FORECASTING THE STOCK MARKET
I want to start this chapter by saying that everything I've stated so far regarding our inability to predict applies to the stock market. I'll say it again. Everything I have stated so far regarding our inability to predict applies to the stock market. I could repeat that sentence a hundred times. But most people would forget it the minute they closed this book. Why?
1
Stock prices used to be quoted using fractions of a dollar. In this example, ¼ means 25 cents.