The debate about NGDP targeting is mostly focused on US monetary policy and the focus of most of the Market Monetarist bloggers is on the US economy and on US monetary policy. That is not in anyway surprising, but this is of little help to policy makers in small-open economies and I have long argued that Market Monetarists also need to address the issue of monetary policy in small-open economies.

In my view NGDP level targeting is exactly as relevant to small-open economies as for the US or the euro zone. However, it terms of the implementation of NGDP level targeting in small open economies that might be easier said than done.

A major problem for small-open economies is that their financial markets typically are less developed than for example the US financial markets and equally important exchange rates moves is having a much bigger impact on the overall economic performance – and especially on the short-term volatility in prices, inflation and NGDP. I therefore think that there is scope for thinking about what I would call exchange rate based NGDP targeting in small open economies.

What I suggest here is something that needs a lot more theoretical and empirical work, but overall my idea is to combine Irving Fisher’s compensated dollar plan (CDP) with NGDP level targeting.

Fisher’s idea was to stabilise the price level by devaluing or revaluing the currency dependent on whether the actual price level was higher or lower than the targeted price level. Hence, if the price level was 1% below the target price level in period t-1 then the currency should devalued by 1% in period t. The Swedish central bank operated a scheme similar to this quite successfully in the 1930s. In Fisher’s scheme the “reference currency” was the dollar versus gold prices. In my scheme it would clearly be a possibility to “manage” the currency against some commodity price like gold prices or a basket of commodity prices (for example the CRB index). Alternatively the currency of the small open economy could be managed vis-à-vis a basket of currencies reflecting for example a trade-weighted basket of currencies.

Unlike Fisher’s scheme the central bank’s target would not be the price level, but rather a NGDP path level and unlike the CDP it should be a forward – and not a backward – looking scheme. Hence, the central bank could for example once every quarter announce an appreciation/depreciation path for the currency over the coming 2-3 years. So if NGDP was lower than the target level then the central bank would announce a “lower” (weaker) path for the currency than otherwise would have been the case.

For Emerging Markets where productivity growth typical is higher than in developed markets the so-called Balassa-Samuelson effect would say that the real effective exchange rate of the Emerging Market economy should gradually appreciate, but if NGDP where to fall below the target level then the central bank would choose to “slowdown” the future path for the exchange rate appreciation relative to the trend rate of appreciation.

I believe that exchange rate based NGDP level targeting could provide a worthwhile alternative to floating exchange (with inflation or NGDP targeting) or rigid pegged exchange rate policies. That said, my idea need to be examined much closer and it would be interesting to see how the rule would perform in standard macroeconomic models under different assumptions.

Finally it should be noted that the there are some clear similarities to a number for the proposal for NGDP growth targeting Bennett McCallum has suggested over the years.