Statistics Matter: Oil, Dollars, Euros & Gold
Great editorial today in the Wall Street Journal.
Only problem is… despite being a print subscriber, the WSJ still prevents me from accessing their content online. Bleh. Thank goodness for Rupert Murdoch, right?🙂 In any case, I am still scanner-equipped, so I can share the better points with you.
Check out this graph. Let it sink in.
Maybe I’m making too big a deal about this, but I found this chart incredibly fascinating. What this basically says is that if the dollar had stayed even with the Euro since 2000, then we’d have $57 Oil, not $100 Oil. So an increase, yes, but not nearly as shocking. More importantly, if the dollar was “as good as gold”, then literally the price of oil would have just barely risen at all, maybe to $30.
It makes you realize how much the topics of the day (peak oil, dependency on foreign supplies, etc) are controlled by economic perspective. I’m not saying anything about the quality of those issues, or the validity of those topics. I’m just pointing out the obvious – the sensationalist nature of seeing a high dollar value on oil is likely fueling the interest in those topics.
However, as I read this piece, it made me wonder, really, what does $100 dollar oil really mean? Does it mean that oil is dearer, or that the dollar is cheaper? Or both?
The reason I titled this post with the preface, “Statistics Matter”, is because I realized today that of all the disciplines and fields I have had the occasion to study and practice over the past 15 years, the fundamental concepts that underly the mathematics of statistics seem to always be valuable, if not essential. (In fact, Against the Gods is one of the books I recommend to people regularly). In fact, I’m probably going to blog on a couple other topics this weekend that all highlight the importance of understanding statistics.
The insight here, which is so common it’s almost trite, is the insight on correlation vs. causality. Correlation measures how often when one thing happens, a second thing also happens. The relationship between their occurrence. Causality is literally the measure of whether when one thing happens, it causes the second to happen. The confusion that normally happens is that people assume that correlation implies causality, when in many cases, it doesn’t.
In my Introduction to Statistics class, 15 years ago, they gave this example. Many people with yellow teeth also develop lung cancer. They are highly correlated. But getting your teeth whitened will not prevent lung cancer. Why? Well because there is a third thing, smoking, which actually causes both yellow teeth and lung cancer. Yellow teeth are positively correlated with lung cancer, but they don’t cause it. Seems obvious, but check out in your daily news how often you’ll see reports of studies that demonstrate nothing but correlation. Health fads are almost all started this way.
Back to Oil.
This article made me wonder – is the weak dollar the reason for $100 oil, as this article suggests, or is $100 dollar oil the cause of the weak dollar. Alternatively, is there a third cause, not mentioned, which actually is weakening the dollar and making oil more valuable?
The great thing about economics, of course, is that almost everything is inter-related. As a result, I’ve always found it very difficult to use macro-economic theory to identify causal factors, except in retrospect. (Hence the joke about economists predicting 19 of the last 7 recessions…)
I accept that one explanation, based on the data in the article, could be that oil hasn’t really become more expensive, in absolute terms. It’s the dollar that has weakened, and that makes it seem like oil is expensive to Americans.
Alternatively, it also seems plausible that since oil is a external good that is predominantly sourced from outside the US, and since there has been a historical shift from our oil-producing partners from being dollar-denominated to a more balanced basket-of-currencies, that the increasing demand for oil has shifted the marginal demand for currencies away from the dollar, and towards previously underweighted measures of value like the Euro and gold.
My bet here is that neither of the above really explains the whole situation. It seems likely that there are a large number of factors affecting the value of the dollar and the value of oil, and the end result has generated a falling dollar and rising value for commodities, including gold & oil.
This issue of causality really matters, however, because if it is in fact a weak dollar which is the causal factor, we have very limited policy options. Let me leave you with the summary thoughts from the article:
This piece of the puzzle really worries me quite a bit – if indeed the rising prices we see are a monetary phenomenon, then we are really stuck between a rock and a hard place with the mortgage/credit issues and the weak dollar. What we could actually be seeing is a magnification effect that has spanned across multiple business cycles, each time the liquidity “solutions” getting larger and larger. This time, the liquidity needed may be so large that it’s actually finally breaking the dollar. Not surprising, really, since it’s pretty easy to argue that the size of the US home mortgage market is actually big enough to really matter versus the aggregate net value and annual product of the United States.
It could be that the future has already been written in this regard – the price we’ll pay over the next 5-10 years from the housing bubble will be measured in a weaker dollar. And that will inflate everything, including our most dear commodities, like oil. We may have to face the fact that liquidity may solve market failures that surround frozen credit markets, but there will be a price to pay.
Ugh. Carter Era.
Here is a link to the full scan of the WSJ article.