Monthly Archives: November 2012

It’s the most ubiquitous of economic metrics, gross domestic product, or GDP.  In one handy, easily accessible number lies revealed the health of the American economy.  And woe to any administration that doesn’t oversee a steady growth in the number.

But what does the number mean exactly, and where does it come from?  While probably not thinking about it too much, you might have felt a twinge of skepticism that a single, simple, number can accurately and thoroughly express the well being of a nation of 300 million people.  Turns out the organization responsible for calculating GDP pretty much shares that skepticism.


The GDP number is calculated and released by the Bureau of Economic Analysis, a part of the Department of Commerce.  Helpfully, the BEA has a list of methodology papers you can view that discuss what the BEA does and how they do it.

Simply put, GDP measures the value of finished goods and services produced in the United States within a certain time period.  “Finished goods” being those goods that aren’t used as input to produce another good (such as flour that’s made into bread), but those that are consumed by end users (the bread that was made form the flour).  These measurements were first developed by Simon Kuznets in the 1930’s, in the midst of the Great Depression, the goal being to give the government a better idea of how the economy was doing and thus give some indication of the impact their policies were having.

The BEA refers to the economic reports it puts out as the national income and product accounts (NIPAs).  Account 1 is the Domestic Income and and Product account, which is where you’ll find the GDP number.  There are 6 other accounts, which basically break the numbers down further, such as into the major economic sectors of businesses, individuals, and governments.

For the basic GDP measure from Account 1, the BEA uses the “expenditure method”.  The idea being to determine the value of all the finished goods produced by finding the total amount of money spent.  But not everything that is produced gets bought does it?  BEAs NIPA primer briefly addresses this:

GDP is a measure of current production, not sales.

In the NIPAs, output measures when a good or service is produced, not when that good or service is sold. For example, an automaker may produce a car in one period and sell it in a later period. In the first period, the production of the car is recorded in GDP as an addition to inventories, a component of investment. In the later period, the sale of the car is recorded as a consumer expenditure and is offset by the withdrawal of the car from inventories.

It’s not clear to me if this concept extends to more perishable goods, such as bread, but I presume in that case, if the bread isn’t sold, gets moldy and is thrown out, that it does not count towards GDP.  And even though the claim is that production is being measured and not sales, there is plenty of production that is not sold on the open market and thus not counted towards GDP, particularly services that people consume themselves, such as car repairs, cooking, or child rearing.  If these services were being paid for, they would be counted towards GDP.

Since market transactions have two sides, a buyer and a seller, GDP can also be calculated as a sum of incomes.  So Account 1 also contains a section calculating GDP using the income approach, the resulting number is sometimes referred to as the gross domestic income or GDI.  In theory it should equal GDP, but there is usually a small discrepancy due to the different data sources used.

Lastly, for the sake of comparison between years, GDP is deflated to account for inflation using a price index, so apples aren’t being compared to oranges.  GDP that is inflation adjusted to a baseline year is referred to as real GDP (as opposed to nominal GDP which hasn’t been adjusted for inflation).


Most of the data the BEA uses in its calculations come from Census Bureau surveys.  But a comprehensive economic census is only performed once every five years.  In between these less comprehensive surveys are done.  When recent data isn’t available the BEA engages in extrapolation and estimation.  Because of this, the quarterly and annual GDP numbers the BEA puts out are continually revised as more information comes in.  The process is described this way in the paper “Taking the Pulse of the Economy: Measuring GDP”

The initial monthly estimates of quarterly GDP based on these extrapolations are revised as more complete data become available— early tabulations of monthly data are replaced by more complete tabulations in subsequent months and later by comprehensive annual surveys that have larger sample frames and provide more detailed information. The successive revisions can be significant, but the initial estimates provide a snapshot of economic activity much like the first few seconds of a Polaroid photograph in which an image is fuzzy, but as the developing process continues, the details become clearer.

So the initial numbers that are put out are “fuzzy”, and as time goes by, as more data comes in, the numbers are revised to their ostensibly more accurate versions.  How long does this revision go on?  From the same paper:

During the summer of each year, the Bureau of Economic Analysis revisits the estimates for the most recent calendar year and the two preceding years, when annual data from the Census Bureau, Internal Revenue Service, and other sources become available. These data are based either on more complete survey —Census Bureau annual data collections are mandatory and the sample frames are much larger than those for the monthly surveys, which are not mandatory—or on comprehensive administrative data, which provide more detailed information by industry, by type of product, or by type of income

So the GDP number is open to revision three years after it is released.  While the same paper discussed accuracy and called the advanced GDP estimates “fairly reliable”, this fact is something important to keep in mind when GDP numbers are released and subsequently used as political weapons.

Revisions and incomplete data to the side, can GDP be used to determine the well being of a society?  The BEA thinks not, as it expresses early on in the GDP Primer:

While GDP is used as an indicator of economic progress, it is not a measure of well-being (for example, it does not account for rates of poverty, crime, or literacy).

Incidentally this is a view shared by Simon Kuznets, GDPs initial developer.  In 1959 economist Moses Abramovitz cautioned:

we must be highly skeptical of the view that long-term changes in the rate of growth of welfare can be gauged even roughly from changes in the rate of growth of output.

So, given the methodological and philosophical hedging being done, it’s a bit incongruous that the BEA brags about the influence of GDP on their Misson, Vision, and Values page:

The GDP was recognized by the Department of Commerce as its greatest achievement of the 20th century and has been ranked as one of the three most influential measures that affect U.S. financial markets.


If GDP isn’t a measure of well being, why is so much attention paid to it, why does it have so much power?  It’s not as if the problems with GDP are unknown.  French president Sarkozy in 2008 created a commission, headed by Columbia University economist Joseph discuss and address the inadequacies of current GDP based economic measurement.  The commission generated a report that is worth perusing and can be found here.

A significant portion of the report discusses “sustainable development”.  A common baseline for yearly GDP growth for a healthy economy is considered to be around 3%.  GDP growth below that is considered anemic.  But a constant growth rate represents exponential growth.  An economy growing at 3% a year will double in about 24 years.  Is that growth rate reasonable?  If so, is it even desirable?

I don’t have the answers.  So while the folks at the BEA undoubtedly work hard to provide numbers that are as accurate as possible, it seems to me we should approach the venerable number as we should approach most things, with healthy skepticism and a sense of proportion.  And always keep in mind these wise words from Stiglitz:

Our economy is supposed to increase our well-being; it is not an end in itself.

Early on in a paper evaluating IPsec (a protocol, now in wide use, developed to allow private and secure communication over the public and insecure Internet), authors Niels Ferguson and Bruce Schineier introduce what they call a rule of thumb:

The Complexity Trap: Security’s worst enemy is complexity.

They proceed to explain:

This might seem an odd statement, especially in the light of the many simple
systems that exhibit critical security failures. It is true nonetheless. Simple failures are simple to avoid, and often simple to fix. The problem in these cases is not a lack of knowledge of how to do it right, but a refusal (or inability) to apply this knowledge. Complexity, however, is a different beast; we do not really know how to handle it. Complex systems exhibit more failures as well as more complex failures. These failures are harder to fi x because the systems are more complex, and before you know it the system has become unmanageable.

Not only can complex security systems break down and fail to be implemented correctly, most of us have been complicit in circumventing or simply ignoring security protocols that we find onerously complex.

While Ferguson and Schneier were focused on one specific security protocol, viewing complexity as the enemy is useful as a more general rule of thumb.  In computer networks, more complexity introduces more moving parts that can break.  More complexity reduces the certainty of how devices will interoperate when plugged in.  More complex networks are more difficult to document properly, and when things do inevitably break, the problems take longer to diagnose and repair.  Complexity can introduce a cascade of costs that flow downstream.

Then there is the realm of finance.  In our highly financialized world, with its dizzying array of complex financial instruments, as we’ve recently seen complexity can become Death, the destroyer of worlds.  Satyajit Das, a derivatives expert, book ends his book “Traders, Guns, and Money” with an account of his experience being an expert witness for an Indonesian noodle maker that was being sued by their bank.  Their bank had convinced the noodle maker to engage in a financial transaction called a currency swap.  Their income was in Indonesian rupiah, and the swap would convert their debt into dollars, with the goal of saving them money.

Imagine a very simple world economy, with two currencies: sticks and rocks.  You make grog and sell it to people.  They pay you in sticks.  You take out a loan to expand your operations, so you have debt.  However, the debt is to be paid back in rocks.  You have sticks from customers’ payments, but how do you get rocks?  Well, there is a currency exchange down at the corner.  At the exchange there is a rate that determines how many rocks you can get for a stick, which can change over time.  Currently the exchange rate is 1/1, an even exchange for sticks and rocks.  Lets say at some point the exchange rate goes to 1/2, one stick is worth two rocks.  This is great for you, without doing anything you have effectively halved your debt because you can get twice as many rocks for the same number of sticks (and hence pay off twice as much of your debt).  How about the other way?  What is it takes two sticks to get a rock?  Your debt has been effectively doubled, because the same number of stick will only get you half as many rocks with which to pay off your loan.

That is called currency risk.  Early in his interaction with the noodle makers, Das asks them “What about the currency risk?  You have borrowings in dollars but no dollar income.  If the dollar rose against the rupiah, then your dollar borrowings would show losses.  Did you consider the currency risk?”  Das’ clients can only respond “No risk, no risk.” because they were told there was none by their bank and they didn’t understand the transactions.

But that was only the beginning, the bank lead the noddle makers through a labyrinth of increasingly complex financial transactions with names like ‘arrears reset swap’ and ‘double up swap’.  After the noodle maker got into serious financial difficulty (owing the bank a great deal of money), this all culminated in the bank setting up a new trade, in which the bank would get 4 million a month from the noodle maker, and the bank would pay the noddle maker a sum calculated from a complicated formula that hilariously always came to zero.

And it is complex financial instruments, with acronyms like CDO and CDS, and the unrecognized risk hidden within them that blew up, that lie at the heart of the global financial meltdown of 2008.  One bracing thing to come out of the post mortem of the catastrophe was how little regulators understood the extent of the interrelation of the large banks and the systemic risk that posed , and how little understanding senior financial executives had of the complex mortgage related securities that sat in their institutions, ticking away and waiting to explode.

Complexity as a weapon

In another of Satyajit Das’ books, Extreme Money, he excerpts an Email from Fabrice Torre, a French employee of Goldman Sachs who sold a complex financial instrument that later exploded, to his girlfriend (yes he refers to himself  as ‘the fabulous Fab’):

More and more leverage in the system.  The whole building is about to collapse anytime now?.?.?.?  Only potential survivor, the fabulous Fab standing in the middle of all these complex, highly leveraged, exotic trades he created without necessarily understanding all of the implications of those monstrosities!!!

We’ve all failed to read the fine print in any number of agreements we’ve entered into, our credit card terms hide any number of absurdly usurious clauses that will be activated at the slightest transgression.  The complexity hides just how poorly our interests are being looked after (and just how well the interests of others is being looked after).  Weaponized complexity, if you will.  And it is aimed at us.  Greg Smith, a Goldman Sachs alum, said recently in a 60 minutes interview, “The quickest way to make money on Wall Street is to take the most sophisticated product and try to sell it the least sophisticated client”.

Matt Taibbi put it this way in his book Griftopia:

Our world isn’t about ideology anymore.  It’s about complexity.  We live in a complex bureaucratic state with complex laws and complex business practices, and the few organizations with the corporate willpower to master these complexities will inevitably own the political power.

I don’t agree with the dismissal of ideology, but clearly our world is, and will increasingly be about complexity and those who can manage and exploit it.  Take regulatory capture.  Those with the money, will, interest, and focus can navigate the baroque, labyrinthine legislative structure and heavily influence the content of laws and regulations.  Those of us without the time or energy to lobby our congressman or review the 10,000 pages of some new proposed legislation are often left out.

But what gives complexity added potency as a weapon is our intellectual vanity.  Most of us have been asked if we understand after having something explained to us.  Even if we don’t, “Do you understand?” is rarely a question we answer no to.  We don’t want to look like an idiot in front of other people.  And if the questioner is condescending, we certainly don’t want to give the asshole the satisfaction of a ‘No, I don’t understand’.

One simple way to fight back against complexity requires little time or energy.  We must lose that fear of looking stupid, we must ask questions.  We must admit when we don’t understand something and get answers that we do understand.  For if we don’t understand something, it may very well be not because we are stupid, but because we aren’t meant to.