Handbook of Exchange Rates

7.5 Aggregating from the LOP to PPP: What Can We Infer?

We are now in position to discuss what happens empirically to the properties of the LOP as we aggregate from individual tradable goods and services to sectors, and then to all tradables, and finally to all goods (tradables and nontradables). The reading of the literature on the LOP applied to individual goods and individual sectors indicates that the LOP holds in the long run. To the extent that the LOP holds for individual goods, it ought to hold for sectors of similar tradable goods. Then, if the LOP holds for each sector of tradable goods, it ought to hold for all tradable goods, that is, the PPP hypothesis for tradable goods must hold, given that the LOP holds for individual goods.

Moving one further level up in the aggregation process, and therefore assessing PPP applied to national (e.g., consumer) price indices as opposed to baskets of tradable goods only, is less straightforward from a theoretical standpoint. The inability to extend the arguments about arbitrage in goods markets to nontradable goods and the potential presence of HBS effects does not allow us to infer that PPP must also hold for national price indices. However, we also know from the extensive literature on PPP and on HBS effects that (i) PPP holds in the long run across a number of currencies even when using price indices that include nontradables and, consistent with this, (ii) HBS effects are generally small, albeit varying over time. It is not entirely ...