[I started writing this last month but with the updated release it is even more topical.]

When I was but a financial market acolyte, GDP was by far my favorite statistic. I definitely would have invited it as my date to the Econ Ball, and would have probably even proposed except for the fear that it’d lead to an onslaught of constitutional amendments, further enlightening our political discourse. It was so fresh, so simple, so clean. When it was on the rise sun filled the skies, and the days of decline were days to malign. You could make a list of it through both time and space to prove progress and power, or the lack thereof. Even sprinkled with a few caveats — like GDP has to rise 2% to keep up with population increases, so really that part should just be subtracted all the time — it was a nice and comforting sign of being in the know.

But no more. We’ve had our ups and downs, and I tried to maintain some emotional connection even as I found out its true nature; all honeymoons have to end sometime. Yet the more I discovered, the more I suspected that instead of being “exciting,” a quality that is capable of papering up so many flaws, the GDP really was just bipolar and off its meds. It contains is no rhyme or reason, so all arguments and analysis are fundamentally pointless.

This feeling has been simmering for quite some time, but it’s finally boiled over. So today, I will make a promise to myself: no longer will the GDP toy with my emotions, good or bad. The relationship is over, and I don’t say this out of anger but out of apathy…the surest sign it’ll stick.

Now let me try to make you apathetic as well. But be warned, we are about to enter a Carrollian underworld, where up is down, down is up, and insanity is not the exception but the rule. Turn back now lest you descend into madness, for Nietzsche was all too right when he wrote, “When you look into an abyss, the abyss also looks into you.”

The first point to note is more on the philosophical level. GDP is merely supposed to be a measure of production, it is not a balance sheet. What does that mean? Well it’s pretty simple, it means that it doesn’t take into account increases in debt, future liabilities, social costs known as externalities (like having to clean up pollution at a later date or face an increase in health costs) or sustainability of growth. Even worse, as Joseph Stiglitz is fond of pointing out large components of the GDP don’t even deal with output at all.

For example, while GDP is supposed to measure the value of output of goods and services, in one key sector – government – we typically have no way of doing it, so we often measure the output simply by the inputs. If government spends more – even if inefficiently – output goes up.

Ditto healthcare. So based on the GDP, it’s a good thing to have bloated ineffectual government and health systems even if they don’t give good outcomes. It is one (accurate) thing to say that GDP doesn’t necessarily correspond with citizen happiness, but Stiglitz’s points are more brutal in that they point out that GDP doesn’t even reflect business sense.

OK, that was enough for me to drop my infatuation and no longer be a “GDP Fetishist,” but surely it has some value and the positives outweigh the negatives. Right? Actually no. Here are the reasons why.

Exports minus Imports leads to some very weird behavior.

The “D” in GDP stands for domestic, highlighting that it shows Exports minus Imports, not an endorsement of its cooking and cleaning abilities as I was sad to discover. This is supposed to account for only what the country is creating, but instead just leads to silliness.

Let’s suppose a hypothetical situation in which global trade is contracting at the fastest pace in 75 years. Well if you are a trade surplus country your exports will fall faster than your imports and GDP will correctly show a decrease. However, if you are a country that runs a trade deficit, your imports will fall faster than your exports and your GDP will add a positive contribution. That’s right, the trade portion will increase the GDP even though the total activity is plummeting, heck even as your exports plummet. Then as global trade starts increasing again, if you are a surplus country the output again makes sense, but if you are a deficit country then the GDP trade component subtracts even though overall trade activity and exports are soaring. Why? Because the imports will rise faster than exports. See look at the US during this not so hypothetical past few years:

This dynamic made the fall look much much shallower than it actually was, and has made the rebound look slightly less impressive.

GDP deflator makes no sense.

The GDP deflator is used to subtract out inflation from the nominal GDP to create the real GDP. The intent is sound since just flooding the economy with money will increase GDP, and you really want to try to eliminate this effect. However, the implementation makes no sense at all for several reasons, but for right now I will focus on trade again.

Import prices have the opposite effect on the GDP deflator that you would expect. If import prices are increasing, that actually decreases the deflator and masks inflation. When the imports are actually sold (either directly or as part of an end product) then theoretically the price increase should be included in the end price and the inflation is added back in. If the import is part of a product chain that takes a few months to process and sell, then this can cause really erratic GDP deflators. People started paying attention to this a lot in 2008 when oil was moving all over the place very rapidly. The deflator went from 1.1% to 4.2% because in one quarter there was an oil price spike, then the next quarter it leveled out or fell and the products made from the oil were sold.

Of course this assumes that there is any pricing power for the end products. Oil and other commodity prices have spiked far more in the past six months than companies can raise prices, squeezing their margins and making it appear that inflation is far lower than it really is. Of course, they fell far more than end prices fell during the recession, increasing the deflator and making it appear worse than things were.

Inventories…inventories…inventories?

Inventories make up a huge part of the GDP. I don’t know why…I mean I know the reasoning, that inventories are representative of output increases or decreases theoretically, but again this leads to craziness. The simple mantra goes that if businesses expect sales increases they increase inventories in anticipation and thus inventories are a leading indicator (I’m not sure how the prevalence of on demand manufacturing plays into this). On the contrary, when they expect sales declines then their inventories will decline as they don’t restock. OK that makes sense, if companies were omniscient beings that are never wrong…which is sort of not the case.

Let’s say that there is a huge collapse in customer spending that catches companies off guard. In that case their inventories will increase even as the sales and orders decrease. GDP is rising but output is in actuality falling. On the other hand, if everyone runs into the stores all at once and buys everything off the shelves, the inventories will decrease, lowering GDP even though there is obviously increased economic activity. Presumably companies would then restock the next quarter, at which point it’s properly accounted for. I question why they don’t just count points of sale.

OK that could be argued to be a minor quibble that has a hard to measure impact on GDP. I would partially agree, except that the inventories are actually measured on the second derivative. Why? I have no clue, but it leads to statements like this:

Inventories were also less of a drag than expected as they fell $16.9b…In total for Q4, inventories added 3.9 % pts to GDP, 66% of the gain.

On what planet do inventories falling $17 billion ADD nearly 4% to the GDP? I couldn’t believe these second derivative shenanigans when I first read about them and never have bothered to try to figure out the rationale, because it will be such tortured logic that I’d have to report the authors to the World Court.

Overly complicated aggregation of a set of simple to follow data sets.

Looking at the breakdown of GDP would have some utility if it contained hidden messages that could be gleamed if you only learned to divine the signs. Even with all this nonsense it could theoretically still have some utility, after all, there used to be a very strong correlation between one’s knowledge of astrology and one’s knowledge of astronomy.

This is far from the case, as everything that is in the GDP is released in separate reports every month. If you want to look at sales, then find the sales report, you don’t need to try to indirectly infer it from GDP. Same goes for imports, exports, whatever. Those individual reports have their own flaws that range from silly to Hardball (is that topical enough to joke about still?) but by contrast the GDP is straight, “We were always at war with Eurasia.”

GDP just doesn’t matter.

A lot of it is just made up out of thin air.

A full 33% of the GDP is fantasy anyway.

Also, I can’t find it right now (can’t think of the proper Google search time) but there is another thing that they do with imports of end goods. Since it is hard to determine origin of goods at the point of sale what they do is subtract the import cost but at point of sale count it as if it were domestic. Thus, if they determine that it costs $400 to make a computer domestically but it is imported for $300, there is a net gain of $100 to GDP. Obviously this makes no sense and boosts the GDP whenever a country eviscerates its domestic goods manufacturing base to buy imports nearly exclusively. This is one of the arguments that explains why our GDP has continued to rise even though wages have stagnated for decades and employment has been flat this past decade.

If we were smart we’d just stop producing everything and continuously pay more for our imports, the combination would send our GDP through the roof!

CODA:

And so there you have it: a quarter with end demand below what we need to stand in place and inventories still falling produces a huge GDP print that during expansionary times would need to have gigantic increases in employment, sales and inventory. Comparisons are basically useless across time, and depending on the variance between individual countries’ bean counters, they are useless between countries too. Best to just stick to the “raw” data.

Of course there is a reason why GDP is constructed the way it is, and is emblematic of a philosophical assumption about the economy of the whole. As you may have noticed, when I noted an absurdity I often said “but then would be absurd in the opposite direction on the other way” which is exactly the point. If you believe that the economy is a force of nature that always increases at a set equilibrium rate and deviations from that are just “noise” in the scheme of things then all the GDP madness does even out over time. However, if you see the economy as naturally in disequilibrium that is held together by a balance of positive and negative feedback loops, the characteristics of which may change over time and could snap, then the GDP is just insane.

See the GDP is really not special at all, it is just reflective of the dominant steady state economic theory that has dominated economics/finance the last hundred years in an attempt to make it a hard science. These theories and tools look great when things are going normally, but fall apart completely when there is a major bust — the time when economic theory and tools are most important! As Steve Keen says about Bernanke, he is called an expert of the Great Depression but according to Bernanke’s assumptions, the Great Depression is an impossibility in the first place. The world economy has a bunch of psychiatrists that say they can treat us, but don’t even believe in mental illness.