Hi Chris,

This is Marek Kolodziej.

I usually don't comment on TOD articles due to the lack of time (I'm studying for my PhD qualifying exams right now), but I would like to make a couple of clarifications. First, Doug Reynolds and I never said in our papers that the Soviet economic decline was chiefly due to the oil production slump - at the very end of the USSR's existence, the military was appropriating something like 85% of the country's budget (I don't remember the actual statistic, please refer to Thane Gustafson's "Crisis Amid Plenty," available e.g. via Amazon). The country had to dissolve due to its military expenditures sooner or later, as well as to other inefficiencies like the inability to produce an adequate amount of food. It is noteworthy that e.g. the food imports had to be paid for with hard cash, and the ruble was not convertible, so the USSR needed to earn pounds/dollars/DM via resource exports - so the Soviet peak in oil production should have and did hasten the Soviet demise. What we did say in our papers was simply that there was "Granger causality" between oil output and GDP - that is, the dynamics of GDP (i.e. first differences of GDP data, e.g. GDP(t)-GDP(t-1)) were better explained statistically by adding the lagged dynamics of oil output (i.e. Q(t-1)-Q(t-2), Q(t-2)-Q(t-3)) to the regression. Since the causality worked only from oil to GDP and not vice versa (the lagged oil production was helping explain the GDP series but the lagged GDP data didn't help explain the oil production data), we concluded that the oil production decline did CONTRIBUTE to the GDP decline - measuring the precise degree of contribution would require a much more detailed specification than just this simple Granger causality test. Unfortunately, Soviet economic data were meaningless - for instance, the farm machinery output was often measured in TONS of machines, not in value added - how do we compare tons of tractors with tons of lasers (the latter being a high value added good), for instance? So, all in all, peak oil did precede and hence HELP cause the GDP decline, but it was by no means a decisive factor. However, given the trouble with converting rubles into foreign currencies, the decline in the amount of exportable resources did hasten the collapse, given that the USSR was forced to import food from the West due to its inefficient agriculture, among other things.

Hope this helps.

Chris,

One more thing - I disagree that coal and gas were exclusively for domestic consumption. Germany, Poland and many other countries were heavily reliant on Soviet natural gas even under communism. Eastern Europe generally traded their industrial products for gas, and at below-market prices - however, a significant percentage of gas was exported to the West. Similarly, the USSR, Poland and other countries that were communist pre-1989 were exporting coal outside of the CMEA (communist) trading bloc - I know quite a bit about that, because I grew up in Poland and have spent the first 19 years of my life there. Apart from the fixed cost nature of the gas infrastructure that we discuss in the paper, it needs to be noted that the European gas market was disconnected from the rest of the world due to the still nascent LNG market - hence, a contributing factor to the oil decline was the collapse of its price in 1986, which did not happen in the case of the exportable Soviet gas due to the relative constancy of European gas demand.

Marek

The question is, no doubt, complex. My personal interpretation is that the slowdown and the peaking in oil production was a direct cause of the fall of the Soviet Union simply because it reduced the amount of foreign currency that was available to the government to purchase food and indispensable items from the West. That was helped by other inefficiencies of the Soviet government that generated plenty of negative feedbacks. Anyway, all that is of course debatable.

A good point that Chris makes is how to explain the precipitous drop in production after the peak. That is, indeed, much steeper than what you can see in the case of the US. I think it is another result of the negative feedbacks that occurred in the US but not in the SU (nice symmetric names!). Just one of these negative feedbacks could have been the "brain drain" that led a large number of Russian engineers and scientist to migrate to the West.

One last point - by "meaningless" Soviet economic data I mean that while GDP was a figure that could still be somehow recovered despite the price distortions (e.g. check out the work of the Penn World Table folks at UPenn who have spent their entire careers calculating PPP adjusted figures for about 120 of the world's countries), most disaggregate data were not available to augment the statistical analysis. So, the model may well be subject to omitted variable bias. Even so, the result agrees with our assumption that stems from the timing of the 1987 peak and the 1991 collapse of the USSR - also, the time series data were extensive enough that the conclusion of the Granger causality test indicates the direction of the causality both way before and after the dissolution of the USSR. As always, one wishes there were more data available.