The Underlying Size of the US Economy

I would like to start this post with a hypothetical item of government spending, such as building an extension to a government building, and track some of the money that might be spent. The example uses the US, but the principles apply to any country. The reason for this example will become apparent later, and may be a little disturbing.

As soon as the money for the extension is approved, there will be a surge in economic activity. Project management groups will be formed, an architect firm hired, contractors will be tendering, and medical suppliers will be preparing for large orders. In this case, I will just track some of the spending as it passes to the firm of architects. Depending on the nature of the contract, money will eventually start being passed to the architect firm, and that money will arrive in the firm's bank account. Once in the account, the money will be used for a range of purposes, such as the cost of operations, cost of office space, and legal and other services. All of these economic activities will be partially supported by the government expenditure. For the next movement of the money, I will concentrate on the salary paid to the workers in the firm, and in particular one of the architects.

Let us imagine that one of the the less senior architects is a man who works on the project for 6 months, and all of his pay is therefore coming from the hospital extension project. We might imagine that the architect is an average type of individual in a professional occupation, and that he is married with a young family and earns $100,000 per year. We can therefore imagine his spending, with his income covering a mortgage, payment on a car loan, and fairly large expenditure on consumer items, and consumption of services like trips to restaurants. All of these will be paid with government money. If we just consider the example of expenditure on consumer items, we might think of a toy purchase for one of the architects. He will visit a store, and we can imagine that he spends $50 on a toy for his daughter's birthday.

The $50 originates in government money for the extension, and it now lands in the till of the retailer selling the toys. The retailer then banks the money, and the money will then be utilised for a range of purposes. If we imagine that the toy has a $25 mark-up, then $25 of the money will be used to reorder the toy. The toy is made by a company based in the US, but with manufacturing overseas in China. As such, some of the money will be used to pay for the manufacturing in China, and we will just imagine that $10 covers the cost of manufacturing a replacement. That means that £15 will go to the US company, and will pay for the operations of the company and profit, with $10 'leaking out' of the US economy into the Chinese economy.

As with the architect firm, the $15 that goes arrives in the US toy company will contribute to staff salaries, pay for the costs of the office, pay for business expenses and so forth. If we imagine one of the office staff, part of the pay of the person will come from the sale of the toy, and that money has come from the architect, and that money has come from the building project, and that money has come from government expenditure. The money for the building extension is rippling through the economy creating activity....Whilst the headline amount of the money diminishes at each level, the money is utilised many times over in multiple transactions.

What I am (I hope) showing, is the way in which the money spent by government generates economic activity way beyond the headline figure of the actual expenditure. Each of the activities that the money generates, as the money use multiplies throughout the economy, is recorded in the GDP figures. It is a very complex process to track, and economists have great difficulty in trying to work out how much government spending increases activity throughout the economy. For example, one review of academic literature (see note 1) identifies that different fiscal expenditure in different situations and countries can lead to multipliers of between zero and four.

It is actually quite easy to see why the multiplier effects of government spending are so difficult to calculate. For example, the spending of government money on infrastructure such as a steel bridge might have very different effects according to the individual economy. If the economy has a strong steel industry, then the multiplier will likely be much greater, as otherwise a large portion of the expenditure will rapidly 'leak out' of the economy to pay for the purchase of the steel from overseas suppliers.

What we are therefore left with is a big question mark over how much activity each $1 of government expenditure might generate within the US economy. In light of the estimates of between zero and four, I will proceed on the basis of a multiplier of three, as the study cited suggests higher multipliers occur in larger economies. Furthermore, a large proportion of the US economy is based upon services, which again suggests a high multiplier. This means that the US might even have a multiplier of four, but I will stay with three as a more conservative estimate.

Having come this far, you may be wondering what all of this means, and why I am discussing this in so much detail. The reason is relatively straightforward. I want to give a (very, very) rough estimate of the real size of the US economy - the size if there were no borrowing from overseas. When we see that way that government spending multiplies in the economy, it is apparent that borrowing money from overseas (translated into government spending) will inflate the GDP figures more than the headline borrowing figures suggest.

Starting with the basics, the US economy for 2009 will produce about $14 trillion of GDP, the federal deficit is expected to be something like 12% of GDP for 2009 overall, and according to the Congressional Budget Office the deficit for the fiscal year (just ended)was $1.4 trillion. In order to isolate the effects of overseas borrowing, it is necessary to find out how much debt has been sold overseas. I naively thought that I might find these figures relatively easily, but this has proved to be problematic (see note 2). For example, I searched the Treasury website, and only found an 'estimate' of the overseas holdings of US government debt. Likewise, the St. Louis Federal Reserve (an excellent resource normally) only offers the following chart, for overseas federal debt holdings:



According to the chart, overseas holdings of federal debt have increased in the first six months of 2009 by $305.3 billion, making an annualised rate of overseas accumulation of US debt of around $600 billion (assuming that overseas investors continued to buy US debt at the same rate). It might be noted that the fiscal year in the US does not match the calendar year, so there is some more fudge in the numbers. However, we might make an approximation on the total amount of GDP that is actually due to the multiplier effect, by multiplying the figure by 3, which gives us $1.8 trillion.

Now at this stage, it would be normal to then minus the $1.8 trillion from the overall size of the economy ($US 14 trillion). The first problem is that some of the borrowed money will be servicing interest payments on previous borrowing ($182 billion), such that it will not create activity. The second problem is that there is the question of how much private borrowing is sourced from overseas lenders, in particular consumer borrowing. I have tried to identify how much of this borrowing originates from overseas, but have been unable to come up with any firm figures. I will therefore just take a guess at the impacts of these two factors and raise the overall total to $2 billion, implying something like $127 billion of consumer credit provided by overseas lenders (remember the multiplier of 3), an extremely conservative guess.

With the current figure of $14 trillion, the correction to remove the activity resultant from overseas borrowing would mean that the real size of the economy is just $12 trillion. In other words, something like 17% of the US economy is funded by overseas borrowing. You may note that there is a large amount of 'fudge factor' in all of this, and comments and critiques are therefore welcomed (especially from those who are more numerate than myself - which is most people).

What does this mean?

As I have frequently pointed out, measurement of GDP is flawed in any economy that has a net accumulation of debt, for the very reason that the activity in the economy includes debt activity derived from overseas borrowing. The problem becomes particularly acute when measuring the ratio of debt to GDP, as the activity from debt is included in GDP. Even more curious is the idea of GDP growth, when a country is accumulating debt, but nevertheless economists are predicting 'growth' for 2010.

The real concern is the question of what would happen if the overseas lenders were to stop lending. In this event, the US economy would quickly revert to its real size, and that would be significantly smaller than today. The results of any halt in credit provision for the US are extremely worrying, as support for a large percentage of the economy would disappear. The result would be that unemployment would explode upwards, government could not operate at current levels, and the shock would likely precipitate a broader collapse in the economy.

The problem that is faced by the government is that for the years preceding the economic crisis, the economy was also being flattered by a combination of overseas borrowing for government expenditure, but in combination with consumer borrowing that originated in overseas credit. I am guessing that, if it were possible to strip out the effect of borrowing for the three years preceding the crisis, the same kind of results would emerge. In other words, I believe the problem is structural, meaning the current level of the economy can not be sustained. As such that there is no possibility in the near term of achieving anything near pre-crisis economic conditions unless creditors continue to fund the US at current levels.

This is improbable.

One reason is that the printing of money by the Federal Reserve and size of deficits are already alarming creditors such as China, with China giving implicit threats to stop lending. A devaluation of the $US might provide a solution to the problems by eventually return the US economy to its true size and wealth relative to other countries. However, ongoing $US devaluation will just accelerate the point at which overseas creditors refuse further lending. In other words, it will precipitate a crisis before the adjustment is completed. In the event of no further credit, the US would face the prospect of rapidly transitioning to the real size of the economy.

The only solution that I can see is for the US to first freeze the growth in borrowing, and to then implement a clear and binding plan to reduce borrowing aggressively over the short to medium term, with a clear plan for a return to surplus in five years time. This plan would give confidence to overseas creditors, and allow a more orderly transition to the real size of the economy. Such a plan would be very, very painful, as the economy would still shrink dramatically. However, it would be less painful than the sudden collapse that might take place if overseas lenders were to stop lending.

The problem is that, at present, there are no such plans. Instead, deficits stretch to the horizon, overseas creditors express their doubts, and the $US is steadily sinking. In other words, my very rough estimates may be put to the test.

Note 1: Spilimbergo, A., Symansky, S., Blanchard, O., Cottarelli, C., & Hall, W. (2009) 'Fiscal policy for the crisis', Centre for Economic Policy Research, Paper No. 7130, January 2009.

Note 2: I used the Treasury website here, and downloaded the document titled 'Ownership of Federal Securites'. Actually getting firm numbers on actual sales of new debt is very difficult, and news reports, for example, seem to focus on debt type and/or sales to particular countries. A fairly tpical example is the following report:

WASHINGTON (MarketWatch) -- Net foreign purchases of long-term securities increased to $28.6 billion in August from $15.3 billion in July, the Treasury Department said Friday. Net foreign purchases of long-term U.S. securities were $ 32.9 billion. Of this, private investors purchased $21.3 billion and foreign official institutions bought $11.6 billion. U.S. residents purchased a net $4.3 billion of long-term foreign securities. China's holdings of Treasurys slipped $3.4 billion to $797.1 billion in August. So far in 2009, China has increased its holdings of Treasurys by $57.5 billion

If you have a definitive source of sales to overseas, a link would be appreciated, or more accurate figures. I am surprised that this data is so hard to find....perhaps I am looking in the wrong places..