Borrowing is nearly always voluntary. Somebody might want to lend to me, but they can't force me to borrow from them if I don't want to. But there's an exception. A country that issues the reserve currency can be forced to borrow from other countries, if those other countries want to lend to it. This point helps us understand the relation between China and the US.
To understand this point, lets go back to the gold standard first. Suppose gold is the only form of international reserves. Suppose one country, let's call it "China", decides it wants to hold a bigger stock of gold reserves. If the world stock of gold reserves is fixed, the only way that China can hold more gold reserves is if other countries hold less gold reserves.
What happens if China wants to hold more gold reserves, but all other countries don't want to hold any less? The result is world deflation, as each country raises interest rates to either acquire more gold reserves or prevent an outflow of gold reserves to other countries.
Eventually this deflation will provide the solution. Deflation can't increase the tonnes of gold reserves in existence (except very slowly, by encouraging more mining), but it will increase the real value of that fixed stock of gold reserves. Each tonne of gold will be worth more in terms of other goods, and countries care about the total value of the tonnes of gold they hold, not about how many tonnes they hold. So deflation ensures that the real value of the world stock of gold reserves will adjust to equal the sum of the real values of gold reserves demanded by all the countries.
But world deflation is not a very pleasant solution to the problem of an increased world demand for international reserves. So the gold standard isn't a very good system.
Now suppose that the US dollar replaces gold as the international reserve currency. And again suppose one country, call it "China", wants to increase its real stock of reserves.
There are two differences between dollars and gold as the international reserve. First, the US can print more dollars at will; it can't print gold. Second, dollars are a liability of the US; gold isn't.
To avoid deflation, the US can increase the stock of international reserves by printing more dollars. Or, it can hold the stock of dollars fixed, and allow deflation to increase the real value of the existing stock of dollars. But either way, the real value of US liabilities to China would increase. The US is forced either to print more dollars to lend to China to avoid deflation, which directly increases the real value of US liabilities to China; or else the US must accept the deflationary consequences of not lending to China, which nevertheless still increases real US liabilities to China by the same amount. Either way, the US is forced to increase the real value of its liabilities to China if China wants to hold an increased value of international reserves.
"Blame China" is not necessarily the lesson to be drawn here. It is quite understandable that some countries might legitimately want or need to accumulate international reserves. "Blame the international reserve system" seems a more appropriate lesson.
For example, if China allowed the US to accumulate yuan, the US Fed could hold yuan reserves, just as China holds dollar reserves. Then, if China wanted to increase its dollar reserves but the US didn't want to increase its liabilities to China, then China could accumulate dollars, and the US Fed could accumulate an equal value of yuan.
Perhaps, if we want to blame China for something, we should not blame it for buying dollars. We should blame it for not allowing the US to buy yuan.
Worthwhile Canadian Initiative 
Nick:
You write “What happens if China wants to hold more gold reserves, but all other countries don’t want to hold any less? The result is world deflation, as each country raises interest rates to either acquire more gold reserves or prevent an outflow of gold reserves to other countries.”
Isn’t this true only if China is sterilizing its increase in gold reserves? If there is no sterilization, then China’s monetary base increases, Chinese prices eventually go up, and finally Chinese start buying cheaper foreign goods leading to an outflow of gold.
Regarding global deflation,wouldn’t the decline in prices elsewhere be offset to some degree by the increase in prices in China? And over longer periods, if the gold stock is growing at about the same rate as the global economy the reversal of the gold flows would tend to lead global price stability, right?
Nick:
I think an important question to ask is why China would want more international reserves in the first place. Guilermo Calvo argues that prior to 2002 the buildup of excess reserves was to acquire a stock of foreign reserves for self-insurance purposes given the regions experience with economic crisis in 1997-1998. Thereafter, Calvo says Asia’s buildup of foreign reserves has more to do with preventing their currencies from appreciating against the dollar given the Fed’s easy monetary policy at that time. In the former case, your story of being forced to borrow makes sense. In the latter case, however, it is the Fed’s own accommodative policies that are driving some of the reserve buildup in Asia. If so, then, had the Fed been less easy after 2002 the reserve buildup by Asia would have been less too. From this perspective, some of the “saving glut” is simply recycled U.S. monetary policy being sent back to the United States.
Nick wrote : “To avoid deflation, the US can increase the stock of international reserves by printing more dollars. Or, it can hold the stock of dollars fixed, and allow deflation to increase the real value of the existing stock of dollars. But either way, the real value of US liabilities to China would increase.”
Only in a very narrow interpretation?
If the U.S. only prints enough for China’s desire to borrow, you are likely correct, the real value of China’s holdings would increase.
If instead the U.S. prints enough for China’s desires as well as enough to fully employ the U.S.’s (or world’s for that matter) population, China’s real value of its liabilities may increase, stay the same or decrease depending on how wasteful the endeavors on which that labor was put to use.
A nice war would very likely lead to a decrease in the value U.S. liabilities.
David:
Yes, I think you are right about sterilisation. Essentially, China would need to reduce its own money supply, relative to its money demand, in order to attract more gold reserves.
Under the gold standard, where the price level is the inverse of the real value of gold, the world price level is determined by world demand and supply of gold. If the real demand for gold was proportional to nominal GDP, and if the world stock of gold increased at the same rate as world GDP, then there would be price stability. But the 1930s taught us that the demand for gold as a reserve can fluctuate.
Suppose Calvo is right (he may well be). If the Fed (post 2002) was pursuing what it saw as domestic US objectives, and other countries didn’t want their monetary policies set in Washington for US objectives, then they should have let their currencies appreciate against the dollar. Hell, that’s the Canadian argument for flexible exchange rates. We want Canadian monetary policy set in Ottawa to meet Canadian objectives, not in Washington to meet US objectives.
Winslow: I don’t think so. If China wants to hold an increased value of US$ reserves, and if it gets what it wants, then US liabilities to China must rise.
“Winslow: I don’t think so. If China wants to hold an increased value of US$ reserves, and if it gets what it wants, then [Nick says ‘real’] US liabilities to China must rise.”
I thought everybody understood this was wrong. Am I misinterpreting what you are saying?
Say China wants to hold an increase in the nominal value of US$ reserves. Say China increases U.S. reseves from $2 trillion to $3 trillion, nominally.
The U.S. determines whether there is an increase in the real value of those US$ reserves.
The U.S. can easily print out and spend $10 trillion on a ground war in the middle east with large inflationary consequences, depreciating the real value of China’s nominal $3 trillion holdings by at least 50%. Total U.S. reserves would more than double without any increase in goods and service production that China might wish to purchase with its holdings.
Winslow: I think we were both misinterpreting each other!
I had my head so wrapped around monetary policy I didn’t notice you were talking about fiscal policy! That was my misinterpretation of you.
Your misinterpretation of me is different. If China’s nominal reserves are N, and P is the price level, then it’s real reserves are R=N/P.
If the Fed increases the world supply of dollars it can increase P and so reduce R for any given existing N. But if China cares about R, not N, then if China wants to increase R it can do so, regardless of what the Fed does. If the Fed wants to keep P constant (avoid deflation) then it must print more dollars to let China increase R and N by the same percentage. If the Fed refuses to print more dollars, then we get deflation, and R increases partly through a decrease in P and partly through an increase in N.
“I had my head so wrapped around monetary policy I didn’t notice you were talking about fiscal policy! That was my misinterpretation of you.”
I think we agree monetary policy has no impact 🙂
The Fed normally only exchanges short-term government liabilities for long-term liabilities through purchases of U.S. government securities. China reserves can be either short or long-term and so monetary policy normally (except now) has no impact on the real or nominal reserve value of China’s holdings.
I can’t think of a scenerio where monetary policy leads to long-term inflation without the help of fiscal policy?
Winslow: “I can’t think of a scenerio where monetary policy leads to long-term inflation without the help of fiscal policy?”
Just have the Fed set the interest rate below the natural rate, and wait. In normal times that sort of monetary policy is necessary and sufficient for the Fed to create any level of inflation it wants. Or to control inflation. I’m just reflecting standard monetary/macro theory here.
“I’m just reflecting standard monetary/macro theory here.”
Nice disclaimer.
Nick, China doesn’t hold dollars, they hold T-bonds. So I don’t see how that is deflationary. Only if they held currency would it be deflationary. On the other hand, lots of foreigners do hold lots of Federal Reserve Notes (currency), and this is very deflationary. But the Fed offsets this by printing enough money so that our Currency/GDP ratio is much higher than Canada. (Fewer foreigners hoard Canadian dollars.) Hence we don’t get the deflation we would otherwise get. The “problem” isn’t China, it’s Colombian drug dealers. We could earn a lot of seignorage off of these people, except that we don’t seem to remember how to debase our currency. It has become a lost art, known only to the ancients.
I don’t see how we are forced to borrow. We could have no national debt at all. As long as there is enough cash in circulation, there won’t be deflation. If we had no national debt then foreigners could hold currency, or interest-bearing bonds from other countries like Japan and Germany.
Scott: I’m going to have to think about that one, to find the clearest way to respond. T-bonds and non-interest paying currency are both liabilities of the US in one sense, but not in the same way.
The current account deficit represents a decrease in national income, and an increase in national consumption. Less income + more goods –> deflationary pressure.
What asset the Chinese store that income in — e.g. currency or treasuries, or corporates — does not matter, provided that the Fed ensures that there is enough currency to maintain its overnight target. If the we did not have a CB, and the chinese “kept” the currency (e.g. stuffing airliners with cash and flying them out of the country), then indeed there would be amazing deflation and it would be impossible to achieve the type of imbalances we have now on the back of a 1 T monetary base.
Nick, Your best argument is that if foreigners want to hold a massive amount of currency, but only temporarily, then they are forcing us to make an interest free loan to them. (Here I assume the Fed is targeting inflation, and must therefore cash in foreign currencvy holdings when the demand is no longer there.) But we can use the seignorage from that foreign demand for currency to have the Fed buy bonds, so it’s an odd type of “loan.” The bigger the loan the more interest we earn on the loan, not the more interest we pay on the loan. Again that is if foreigners hold currency. They would probably choose hold T-bonds, but they in no way force the Treasury to issue T-bonds.
I find this post a bit confusing. You should define what you mean by “China”. Assuming that you mean the monetary authority, then if it wants reserves of any kind, it must sell either base money or bonds. If China had a floating exchange rate and an open capital market, the monetary authority could acquire reserves by simply exchanging renminbi debt for, say, dollar debt, with only the transient involvement of money. In such a case, no-one need borrow more. At a high level, this is exactly what the Chinese authorities are doing – exchanging renminbi sterilisation bills for treasuries. The effect of this operation on prices (including exchange rates and interest rates) and quantities, however, depends on how the transaction is initiated – in China’s case, driven by export receipts exchanged at a fixed rate, rather than by a desire to acquire reserves per se – and how markets respond – in the case of the US, with an elastic supply of dollar debt. The point is that it is necessary to consider what the Chinese private sector and the US authorities and US private sector are doing as well as the Chinese authorities.