Notes on the "Nominal vs. PPP exchange rates" debate

I. Introduction and Background:

Mr. Jayant Bhandari is an institutional investor and consultant, active in the mining industry, an international traveller and an avid socio-economic thinker. He has written an article titled “Purchasing Power Parity (PPP) or Nominal exchange rates” on his site acting-man.com (see here: http://www.acting-man.com/?p=52263) partially in response to the points raised by the author of this blog during his programme on “Indian Economy, Culture and Investing”. This was undertaken at the AURO University, Surat. The programme was quite good with considerable discussions and debates on several key issues of macro and micro economic importance. The speaker provided considerable anecdotal insights into various growth related issues with a special emphasis on the popular India vs. China debate. Given his vast international exposure, it was no surprise that his take on various issues was much more cosmopolitan and enriched compared to the uncritical “nationalist” propagandists.

   However, the conclusions that were reached (even if correct) were based on wrong and fundamentally incorrect methodologies and tools, which were employed to derive the speaker’s a priori beliefs. The author of this article found that the speaker was using arguments whose conclusions could not be reached from their premises without overhauling the very premises that were used. Mere tentative correctness of conclusions does not lend an argument, whether inductive, deductive or abductive, to logical truthfulness. The conclusion must logically flow from the premises employed.

   Keeping aside these matters, the author made three critical observations on the speaker’s approach and methodological strategies:
1. Use of nominal GDP was a very weak way of comparing any two nations, including India and China.

2. The speaker’s lack of full understanding about convergence dynamics, which has been an intensely debated area in Growth and Development Economics.

3. The unawareness of the speaker about some core and key economic issues in macro and growth economics.

   The speaker Mr. Bhandari, recently published an article as mentioned above and this post provides the view of this author on the same. It is not meant to be a criticism or refutation. Such exercises are generally futile. Given Mr. Jayant’s large international exposure, the author of this article is not in a position to criticize his conclusion that India lags behind China on many accounts and that the interaction between Indian culture, economy and investment practices are producing sub-optimal socio-economic outcomes. Hence, the discussion below will be confined to issues of theoretical and conceptual importance on which this author has some minimum understanding. The discussion is kept at as elementary and common-sense a level as possible so that others can jump in and provide their perspectives too. Names are given to each premise to make them easier to refer. These names are the author’s own choice and others can ignore them if needed.

II. Core Issues:

Underlying Mr. Bhandari’s article are some key premises, which might be stated as below:

1. There is something like “the nominal exchange rate” (The Ontological premise).

2. This is what the market has to offer and hence its discovery is subjected to the forces of price mechanism (The Representativeness premise).

3. Nominal exchange rates are better suited for growth-related comparisons and analysis (The Suitability premise).

4. PPP exchange rates are an alternative to Nominal exchange rates, and hence can be clearly differentiated (The Alternatives premise).

5. Qualitative differences between economies are better captured by Nominal bilateral exchange rates compared to PPP rates (The Qualitative premise).

6. Convergence is probably a wrong idea or at best not applicable to the current context of India vs. China debate that was touched in the said programme in Surat (The Convergence premise).

7. Price differences do not capture quality differences (The Price-Quality premise).

8. Comparison of GDP across a “nation” is a correct representative of the true economic space over which actual economic activities are undertaken (The Space premise).

   As is clear from above, the writer of the said article is undertaking an analysis on issues of great significance based on some questionable notions, concepts and theories. He was honest enough to accept his lack of exposure to formal Economics. But because he has written on a critical economic issue, now the discussion must be made clearer for all stakeholders involved. It is to be noted that the point of disagreement was on nominal vs. real per capita GDP. However, the speaker wrote an article on a quite different issue. Accordingly, the following section tries to clarify why the above held premises are probably not as correct as they seem prima facie.

1. The Ontological premise: There is something like “the nominal exchange rate”

In reality there are “bid” and “ask” rates as far as bilateral nominal foreign exchange rates are concerned. The so-called “the nominal exchange rate” is at best an average of the both. It is quite different from the purer market-given bilateral foreign exchange rates that embody at least two dimensions, one emerging from the buyer-dealer side of the foreign exchange market and second emerging from the seller-dealer side of the same. The kind of information content contained in such average values is also different than their constituents. Any average is characterized by information loss and appealing to an average value as a true representative of its constituents without having sufficient information/approximations about the underlying population is untenable. Using the average bilateral nominal exchange rate to analyze such profound issues as differential growth dynamics is highly questionable. In terms of Economic Ontology, the nominal bilateral exchange rate that is generally employed for economic analysis, including the conversion of the amount of domestic GDP into other currencies, is a representative notion and presuppose that the underlying reality is closely approximated by that representation. In case of bilateral exchange rates, holding such a belief for long-run and probably even short-run comparison without detailed research on the issue is highly questionable. More issues can be debated here but space does not allow it.

2. The Representativeness premise: Nominal bilateral exchange rate is what the market has to offer and hence its discovery is subjected to the forces of price mechanism

Markets don’t offer “a” nominal exchange rate. As pointed above, different rates are offered depending on the side of the market and the structure of the market being focused upon. Moreover, basic economics suggests that nominal values are also governed by underlying real forces and the foreign exchange markets offer bid and ask rates which continuously fluctuate (even vary throughout the day). Such variability and the fact that the bid-ask spread is also volatile make it difficult to rely on bilateral representatives for analysis beyond probably an immediate horizon. Moreover, it is not possible to locate the determination and offering of the nominal exchange rate purely to the market-price system. Distortions generated by central bank interventions and other external shocks also make it difficult to rely on just the observed bilateral rates. Literature in international macroeconomics suggests that the exchange rate determination process is very complex and simplistic measures such as bilateral exchange rates are not the best or true representatives of the actual underlying exchange rate. Better measures are available and can be used instead.

3. The Suitability premise: Nominal exchange rates are better suited for growth-related comparisons and analysis

Given that nominal exchange rates are marred by several fundamental problems as pointed above, they already are questionable for use in medium to long run comparisons. Discussions on economic growth related issues presuppose a stable long-run conversion factor that can be used to compare GDPs across different countries. Given that any two economies are governed by different underlying structural characteristics and, different real forces (e.g. real investments, savings, etc.) and nominal forces (e.g. general price determination process, inflation, monetary policies, etc.), the role of the conversion factor and hence the implicit weighting factor becomes critical in establishing some degree of similarity between the variables of interest (e.g. GDPs of China and India here). PPP rates were historically developed to address such concerns. By accounting for price differentials, at least one crucial differentiating factor between two countries (differences in cost of living and price-structures) could be tackled with. Establishing some degree of similarity between two structurally different economies is needed to avoid comparing apples and chairs. PPP rates are also representative measures and cannot be blindly relied upon. But avoiding them in growth analysis is at best a dangerous error.

4. The Alternatives premise: PPP exchange rates are an alternative to Nominal exchange rates, and hence can be clearly differentiated

Juxtaposing nominal bilateral and PPP rates as two different alternatives implies that both are based on different underlying determinants. However, PPP theory is also a theory of the long run equilibrium behaviour of nominal bilateral exchange rates and is also bilateral as well as nominal in nature. Considering them as two alternatives for the same problem is probably not correct. PPP rates are not real exchange rates. Real exchange rates are an alternative to nominal exchange rates. Moreover, the construction of PPP rates requires nominal exchange rate itself. For example converting a local price of a good into its foreign currency price to construct a PPP rate will require the use of nominal exchange rate (generally the spot rate). However, given that nominal bilateral rates are a component of PPP rates themselves, it is difficult to rule out something called a “two-way causality” at least statistically (what causes what?, do nominal rates cause PPP or it is the other way round or both cause each other simultaneously?). Economists do use them as alternatives but strict adherence to such beliefs has always been criticised. For long-run analysis including comparisons of GDPs and their growths, PPP rates provide more meaningful measures by accounting for “purchasing power” differences and to some extent tackling the problem of differences in local price structures and the aggregate cost of livings.

5. The Qualitative premise: Qualitative differences between economies are better captured by Nominal bilateral exchange rates

It is not possible to rely on nominal bilateral exchange rates to account for qualitative improvements and differences between two nations mainly because the behaviour of nominal rates is governed by many factors including randomness. It has been found to be erratic and accounting for qualitative improvements requires a stable trend in the proxy variable that is closely related to the slow and gradual evolution of an economy’s goods, services and human capital. On this account, even the PPP rates fail, and comparing the economic outputs of two economies by taking into account their qualitative differences requires much more sophisticated quality-adjustment models. Recent literature in International trade theory has taken up this issue intensely and it is at least clear that nominal or PPP rates are not even remotely qualified to do the task. They are not at all meaningful representatives of qualitative changes and thus separate indexes and other such tools are usually employed. Large amount of literature is available on the New Goods-related trade models which are generally analyzed in a dynamic framework. Full-sized articles are needed to explain these issues.

6. The Convergence premise: Convergence is probably a wrong idea or at best not applicable to the current context

Convergence is an inherently dynamic (i.e. time-dependent) phenomenon. Comparative-Static approaches such as comparison of two countries’ GDPs, per capita GDPs, etc. using any kind of exchange rate is not applicable to such a process. Every Indian, Chinese, etc. at least has an ambition or hope to have a better lifestyle both materially and otherwise. Generally, there are benchmarks that we all have to use in order to define what a “better” economic condition is assuming basic needs are met. These notions and belief about what is a “good”, superior” or “better” life are different across individuals and societies. Hence, when two economies are compared, two different underlying social preference functions and social welfare functions are compared. It is thus important to have an idea as to what extent and with how much time can the two economies be expected to have similar preference functions so that comparison becomes meaningful. Not only that, but having an idea on “how long will it take me and at what rate of growth to reach some desired level of well-being?” is also important to my economic decisions. Such logic can be inflated to whole nations too. Convergence does exactly that. It allows an analyst to understand the path, time and lags needed for an economy or some section of it to reach some desired level on an economic parameter. It does not imply simply the equalization of the growth rates of two nations. Rather, over a period of time, two nations might happen to reach similar paths of growths and not necessarily growth rates and hence the actual amount of GDPs for example might be expected to come nearer to each other say between India and China and probably become equal. Such equalization does happen and economies also often overtake each other in terms of GDPs, per capita GDPs, growth rates, etc. Now what path of convergence is focused upon depends on the context. For example, Economic literature talks of long run steady state of growth and convergence would imply the path required for a given economy to reach that steady state level.

7. The Price-Quality premise: Price differences do not capture quality differences

Prices are not pure measures of quality. Qualitative changes in economies require a lot more information and sole reliance on price is incorrect. However, when higher degrees of aggregation are used, there are hardly any measures of quality differences and given the law of central tendency (meaning actual values tend to converge to their mean values as observations become infinitely larger) so that nominal variations in prices cancel out each other (such assumption is open to further debate and research), aggregate price indexes are the only simple measures available to understand the changes in qualities of life, goods and services between two nations. A higher price in one country implies, among other things, a better quality of product, given that the goods are homogenous. If two goods such as a bus trip in India vs. a bus trip in U.S. are not homogenous (one can refer literature in the choice-theoretic tradition), then many other factors could be causing these observed price differences. Quality differences are only one component. Despite these problems, however, removing relative aggregate price dimension from GDPs is needed to locate the changes in the quality and competitiveness of outputs between two nations. Otherwise the price movements will dominate the behaviour of GDP. Hence, nominal bilateral exchange rates fail on this account too.

8. The Space premise: Comparison of GDP across a “nation” is a correct representative of the true economic space over which actual economic activities are undertaken

Increasingly, there has been utter dissatisfaction with the very concept of State and its corresponding measures of economic actives primarily on account of the breakdown of the traditional economic spaces and increased integrations of erstwhile national boundaries in the economic realm (Harvey, 2001: Globalization and the “Spatial Fix”). Economic geographers in particular are discontent with the way macroeconomists treat economic spaces and hence there has been a burgeoning literature in the field of Geographical Economics. It is now no more possible to use GDPs as produced by nation-states given that economic activities are increasingly becoming inter-dependent. This inter-dependency dictates the need for better measures of aggregate outputs. Such developments are making the use of national GDPs unreliable, no matter how much adjustment is done for international linkages, because the underlying concept of space (the “nation”) is based on the notion of a state and it no more encompasses the full proportion of aggregate economic activities that societies undertake today.

III. Conclusion:

Alfred Marshall conceived of Economics as the science of the ordinary business of life of humans. What is less known is that he was interested not only in the economic dimensions but also the sociological dimensions of day-to-day life. He stressed on the need to incorporate dynamic economics in common-sense understanding. Yet, people in general use a lot of static and comparative-static concepts to understand essentially dynamic processes going around them. The above article has stressed on the need to move beyond traditional measures of economic activities as generated by the state. If it is the state that is the problem, then the factual repertoire produced by it either needs to be complemented with better measures of economic lives of societies or must be adjusted accordingly. It is quite clear that adjustment of the existing database on economic activities is a much more realistic exercise. It is hoped that some meaningful and mutually beneficial ideas come out of this rough write-up for all readers.

   Please feel free to criticize, reject, question, discuss or even appreciate what is written, but do think on the matters discussed and share an opinion.

   With warm regards.