In some cases, one can observe absolutely clear evidence of the presence of long-term forces in economic variables, which produce sustainable linear trends over decades. There is no way to interpret these forces as random ones.
The price index for computer equipment in the previous post demonstrates a sustainable (almost linear) trend relative to the overall PPI since the very beginning. Many components of the PPI show changing trends - they deviate from the PPI, return to the PPI, and deviate again. The conservation law demand the existence of some indices which deviate linearly from the PPI in the direction opposite to that of the index of computers - the weighted average of all components makes the PPI. As with the CPI components, for example communication and medical care, the best candidate among the PPI components is the index for drugs and pharmaceuticals, dpPPI. Figure 1 displays the difference between the PPI and the index for drugs. Between 1960 and 1980, the difference was growing from -19 to +7. After 1980, the difference is characterized by the presence of a sustainable linear trend: from +7 in 1980 to -140 in July 2008. There are several short-term fluctuations - the late 1990s and 2000s. But these fluctuations do not bias the trend slope. So, one can expect that the producer price index for drugs and pharmaceuticals will evolve along the same trend.
Pharmaceutical companies will retain their pricing power relative to other goods and services in the PPI.
Notice that the figure shows the difference between two price indices, i.e. the difference between absolute levels of corresponding prices. In relative terms, the rate of price growth of drugs and pharmaceuticals, dcPPI/PPI, approaches that of PPI in the long run, since annual increment in the difference, d[dcPPI-PPI], is constant and the denominator, i.e. the PPI itself, grows over time. So, in the rates of growth of the dcPPI and PPI asymptotically converge.