Tyranny of numbers

The latest data from the Statistical Center of Iran (SCI) for inflation show that prices rose faster last month (the Iranian month of Dey ending on January 20) compared to the month before. At that time, inflation was on its way down from its peak of 59% annual rate in September to 22% in November. Last month it was back up to 32% last month. This is potentially bad news for the Raisi government, which is keen to show that his government can control inflation and dispel endless rumors that Iran’s economy is on the brink.

Curiously, the news of inflation moving back up can be good or bad economic news. We won’t know until the data on output and employment for the fall quarter become available. Is it too much money chasing too few goods, as the common economic refrain in Iran is these days, or is it a sign that demand (domestic and external) is steady and can sustain the economic recovery of the recent 6 quarters into the future?

In the meantime, a few interesting facts about inflation can be gleaned from the past inflation behavior alone. First, the most recent increase in inflation is different from the exchange-rate driven high inflation episodes of recent years. Since the exchange rate has been steady, rising inflation points to rapid increase in spending and inevitable increase in liquidity, which new Central Bank data (link in Persian) seem to suggest.

Second, inflation is in the downward phase of a negative Covid shock, which has a different origin than rising liquidity. The latter is clear from the Figure below, which presents the three-month moving average of monthly inflation rates on an annual basis. The two large episodes of inflation, marked with sanctions and devaluation, one starting in 2012 following the tightening of sanctions by Obama and the other in 2018 caused by Trump, are well marked by the collapse of the rial in the beginning of these episodes. It is difficult to attribute these high-inflation episodes to a sudden rush of new liquidity. They are, in my opinion, the result of external shocks that the monetary authorities wisely accommodated. If they had not, output would have had to contract.

The lesson for Iran is therefore to manage its foreign relations to reduce the likelihood of external shocks or to make its economy more resistant to them. Until then, tightening the money supply when they occur is the wrong response.

There are also two episodes I have marked as Covid-19 related, which follow a similar cost-push pattern as those related to sanctions — a quick rise followed by quick decline. These shocks obviously call for different policies, but Iran’s record in this respect is not any worse than some other countries with much better resources, like the US.

I am not going to make this post into a test of the quantity theory of money, not the least because it is not an area of my expertise. But I know enough about the effect of external shocks when prices are rigid downwards to draw the above conclusion. As Nicholas Kaldor used to say, quantities move faster than prices. In the case of a negative shock following a cut in oil exports, output and employment are more likely to contract than local prices to fall to adjust relative prices across the economy. That is, prices of traded good must rise faster than nongraded goods ratter then falling non-traded prices. This is how Iran’s economy has adjusted to external shocks, positive and negative in the past, by experiencing faster increase in one set of prices than others. As I mentioned on this blog two years ago, after devaluation, Iranians bought Iran-made boxed cereals not because their prices were falling but because the price of the imported brands were rising with the price of the dollar.

Finally, with the year 1400 almost over, it is safe to predict inflation for the year. If in the next two months prices move as they have in the last two, the average CPI for 1400 will be 40% higher than 1399. This is much higher than the expenditure increase in the proposed budget for next year, as I noted in my last blog post. It is also the rate of inflation I assumed in this piece for Project Syndicate, which was published before new inflation data were out.