Ben Southwood is Head of Macro Policy at the Adam Smith Institute


In the popular business and economics press, as well as among interested laymen, there is a notion that countries can compete, or even fight a war, with their exchange rates. Cheaper sterling makes the stuff Brits produce cheaper to foreigners without reducing the amount of pounds firms can get for it, while it makes foreign-produced goods more expensive to import. This raises exports and cuts imports, and it’s understandable that people see imports as money flowing out of the country, and exports as money flowing in.

Thus, there is the idea that exchange rates are zero-sum: one country can raise their net exports (and economic wellbeing) only at the expense of other countries. One devaluation will provoke a series of ‘competitive devaluations’ – a currency war. But this is wrongheaded.

I will explain in general why there cannot be such thing as a currency war, and then explain specifically why this general idea applies to the situation in China. There are two principal points: firstly, so long as the macroeconomy is being managed well, exports are a cost and imports a benefit; secondly, real exchange rates cannot be controlled and are set in the market – governments cannot manipulate them.

Exports are a cost in equilibrium

If all prices were perfectly flexible, then markets would not suffer macroeconomic downturns, slumps, sluggish recoveries and so on. There would be no bouts of unemployment, few debt crises, or costly adjustments. But in the real world many prices, particularly the prices of debt and labour (ie. wages) are fixed in nominal terms, not correcting for inflation or other factors.

The total amount of demand, which is also measured in nominal terms, is split up amongst all the goods, and the prices make this split fit. If we see the economy as a game of musical chairs then prices clear markets by making sure there are exactly enough chairs to seat everyone. If the total amount of demand falls but price-setters don’t realise the importance of this, then prices are too high and there isn’t enough demand to pay them all. During this time there will be unemployment – workers are setting the price of their labour too high – and not just of human labour power, but also of all other factors.

Slowly, workers and debtors and creditors and investors will realise the changed macroeconomic circumstances and adjust their plans, offers and the wages they’ll accept. But during macroeconomic disequilibrium like this, any extra demand is helpful and will help deal with what economist Bryan Caplan calls “the grave evil of unemployment”. Things are so bad that in John Maynard Keynes’ view even paying workers to dig holes and refill them would be beneficial overall.

In the same way, these extraordinary economic times mean that extra exports – which add aggregate demand to the system – are good, and extra imports – which send demand out of the system – are bad.


But it is not usually the case. In macroeconomic equilibrium, where there is sufficient demand to employ all capital and workers, extra capital or workers devoted to foreign demand (ie. exports) means less capital and fewer workers devoted to domestic demand. This means fewer goods for residents to consume and more for foreigners to consume.

In equilibrium, exports are the cost you pay for past, current and future imports. What we want is more goods to consume – more and better healthcare, more and better education, more and better housing, more and better clothes, more and better transport. If we do more exporting we can produce fewer of these things for ourselves, and if we do less importing we can bring fewer of these in from abroad.

Extra exports can in the short-run bring more workers into the labour force but this disequilibrium can only last for a short time, and only then by tricking workers into thinking that their wage offers are really worth more than they really are in real terms. This exploits what is known as ‘money illusion’ to pay workers less in real terms than they would be willing to accept.

The upshot of all this is that if countries could devalue their currencies to make their goods cheaper, and simultaneously make foreign goods more expensive, then what they would really be doing is making their citizens work harder to consume the same amount of goods, giving the extra chunk to foreigners on the cheap. Before we start worrying about this though (China is still a very poor country, and Yuan-holders probably shouldn’t morally be subsidising the rest of the world’s consumption) we’ll note that adjustments cancel out exchange rate policy before it can have much effect.

Swimming against the tide

Consider: the Bank of China wants a cheaper Yuan/renminbi. How does it achieve this? Suppose it decreed that a dollar would now buy 10 instead of 6.25 (the going rate as I wrote this piece). Suppose the BoC said it would supply infinite Yuan at the price of $0.10. What we would expect to see is a huge flow of dollars to the BoC. But each dollar that the BoC buys means an extra flow of 10 into the world economy.

These renminbi will all be spent in China, and more money, we know, means higher inflation and higher asset prices. Assuming that China had a stable macroeconomic picture at the beginning of the peg, the extra aggregate demand will not translate into any more real output or income. People will buy Yuan until they are worth only $0.10 each, with inflation and asset price rises the result. This is borne out empirically, and across the world inflation and a cheaper nominal currency go hand in hand.

Essentially, the point is that trying to fix a currency is swimming against the market tide and can’t be achieved in real terms. If you unexpectedly print lots of extra money to devalue your currency you will, for a short time, make your goods cheaper to foreigners, but soon inflation will erode away those terms of trade gains. To sustain this into the medium term you would have to continually surprise the market with new devaluations – which is very hard.

So exports are a cost, outside a recession, and imports are a benefit. And in any case it’s impossible to fix a real exchange rate for long. So currency wars shouldn’t bother us too much. In fact, they even have benefits. Consider the monetary-macroeconomics disequilibrium scenario raised earlier. In this scenario, a deficiency of demand – money –means that certain prices 9particularly wages0 are too high relative to nominal demand, and the time they take to adjust to the new situation is a ressesion

A currency war involves printing lots of money and raising the total amount of nominal demand. And many economists, including Danske Bank chief analyst Lars Christensen and Bentley College Professor Scott Sumner, have pointed out that this is exactly how currency devaluation works its beneficial effects. It raises the total amount of demand to allow markets to clear even despite wage and price stickiness. This is why, they argue, Japan imported more after starting the massive money printing and devaluation that was Abenomics.

From this analysis, devaluations, far from being zero-sum – where a benefit to one country is a cost to another – are positive-sum. A devaluation in one country helps bring them out of recession, allowing more trade with other countries, in turn allowing broader and deeper specialisation and more goods for everyone to consume.


There are no such things as currency wars, and devaluations are not competitive. We cannot swim against the overwhelming tide of the market, and we wouldn’t want to. In equilibrium exports are a cost we pay to get more imports in the past, present, or future. But that doesn’t mean devaluation is useless – it may help get countries out of demand-side recession by increasing the amount of nominal demand in the economy.