A review of the graph in this article of per capita energy consumption demonstrates that the thesis that that increased investment in efficiency R&D results in reduced energy consumption is not supported by historical experience.
Almost the the entire reduction in per capita energy consumption of 7,000 kwh observed in the U.S. from 1980 through 2010 is attributable to two events. The first event was a decrease of approximately 9,500 per capita from 1980 through mid 1983, and the second was a decrease of approximately 8,600 per capita in the period from mid 2007 through mid 2009. The per capita energy consumption in the remaining 24.5 years out of the 30 year study period actually increased by 11,100 kwh per capita.
The first period of decline was associated with a severe recession that was a consequence of a spike in oil price that resulted from the Iranian Revolution. As the graph indicates, similar declines in per capita energy consumption occurred in Germany and Japan as well. This decline had nothing to do with R&D or energy innovation.
The first decline event ended in mid 2003, when oil prices stabilized, and per capita energy consumption increased steadily until 2000, when oil prices began to rise.
The second decline event began in mid 2007, when the increase in oil price accelerated, culminating in a record high price of $134 (in March 2013 dollars) in June of 2008. The decline continued until mid 2009, as a result of the recession caused by the credit collapse of 2008. Here also, similar declines occurred in Germany and Japan, although they were less severe, because the oil consumption per capita in thos e countries is lower than that of the U.S., due to geographical and cultural factors.
In fact, the curves of per capita energy consumption over the study period are almost identical for Germany and Japan, except for the period of the 1990's, when per capita energy consumption increased in Japan, but decreased in Germany. This is likely a result of the economic consequences of the cost of reunification in Germany, and had nothing to do with energy policy.
The bottom line is that there is absolutely no evidence that the variations in the trends in per capita energy consumption over the study period in the three countries has anything to do with the magnitude of R&D expenditures or energy innovation.