01.02.2015 Views

Producer Price Index Manual: Theory and Practice ... - METAC

Producer Price Index Manual: Theory and Practice ... - METAC

Producer Price Index Manual: Theory and Practice ... - METAC

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

1. An Introduction to PPI Methodology<br />

proach in itself does not determine the form of the<br />

index number. There are several stochastic indices<br />

to choose from, just as there are many possible universes.<br />

However, as already noted, the elementary<br />

prices from which most aggregate price indices are<br />

constructed usually have to be based on samples of<br />

prices, <strong>and</strong> the stochastic approach may provide<br />

useful guidance on how best to estimate them.<br />

E. The Economic Approach<br />

1.87 The economic approach differs from the<br />

previous approaches in an important respect: quantities<br />

are no longer assumed to be independent of<br />

prices. If, for example, it is assumed that firms behave<br />

as revenue maximizers, it follows that they<br />

would produce more of products with aboveaverage<br />

price changes in, say, period 1 compared<br />

with period 0. As a result, the revenue shares in period<br />

1 from such products will increase, <strong>and</strong> therefore,<br />

their weights. This behavioral assumption<br />

about the firm, as it switches production to higherpriced<br />

products, allows something to be said about<br />

what “true” indices should be <strong>and</strong> the suitability of<br />

different index number formulas. For example, the<br />

Laspeyres index uses fixed period 0 revenue shares<br />

to weight its price relatives <strong>and</strong> ignores the substitution<br />

of production toward products with higher relative<br />

price changes in period 1. It will thus understate<br />

aggregate price changes—be biased downward<br />

against its true index. The Paasche index uses fixed<br />

period 1 weights <strong>and</strong> ignores the initial revenue<br />

shares in period 0. It will thus overstate aggregate<br />

price changes—be biased upward against its true<br />

index.<br />

1.88 The economic approach can be seen to be<br />

very powerful, since it has identified a type of bias<br />

in Laspeyres <strong>and</strong> Paasche indices not apparent from<br />

other approaches: substitution bias. Laspeyres <strong>and</strong><br />

Paasche indices ignore the change in weights as<br />

producers substitute their production toward products<br />

with above-average price increases. Yet the nature<br />

of the bias arises from an assumption about the<br />

behavior of producers—that they are revenue<br />

maximizers. Consider an alternative assumption:<br />

that producers respond to dem<strong>and</strong> changes<br />

prompted by purchasers buying less of products<br />

with relatively high price changes. Products whose<br />

price increases are, for example, above average will<br />

see a falloff in dem<strong>and</strong> leading to a falloff in production.<br />

In this case the revenue shares or weights<br />

of products with above-average price increases will<br />

fall in period 1, <strong>and</strong> the fixed period 0 weighted<br />

Laspeyres will overstate aggregate price changes—<br />

it will be upward biased. This compares with the<br />

Paasche index, which will understate aggregate<br />

price changes—it will be downward biased. It is<br />

shown in Chapter 17 that Laspeyres <strong>and</strong> Paasche<br />

indices can under certain conditions act as bounds<br />

on a more generally applicable “true” economic<br />

theoretic index. The axiomatic approach in Section<br />

C led to an index number formula that used an average<br />

of the Laspeyres <strong>and</strong> Paasche indices, <strong>and</strong><br />

even at this early stage in the discussion, the economic<br />

approach seems to provide further support.<br />

1.89 The economic approach also identifies the<br />

circumstances under which the conventionally used<br />

Laspeyres index is appropriate. This would require<br />

that the firm does not change its production configuration<br />

in response to relative price changes, at<br />

least over the short term of the price index comparisons.<br />

Economic theory thus argues that the<br />

Laspeyres index may be appropriate for industries<br />

in which quantities are known not to respond to<br />

relative price changes over the period of the price<br />

comparisons. But it is more likely that this will be<br />

the exception rather than the norm, <strong>and</strong> the theory<br />

points to a requirement for a more generally applicable<br />

index number formula.<br />

1.90 The PPI indices considered here include<br />

output, input, <strong>and</strong> value-added price indices (deflators),<br />

<strong>and</strong> different assumptions arise in their formulation<br />

from economic theory. In the output case,<br />

an assumption is made that firms act to maximize<br />

revenues, from a given input base. Firms substitute<br />

toward products with relatively high price increases.<br />

For the input price index, the concern is to<br />

minimize the costs of purchased intermediate<br />

goods. Firms substitute away from input products<br />

with relatively high price increases. For the valueadded<br />

deflator, the unusual use of negative weights<br />

for the inputs is considered. The economic approach,<br />

as shown in Chapter 17, demonstrates that:<br />

• A substitution bias can exist when using<br />

Laspeyres <strong>and</strong> Paasche formulas.<br />

• The nature of the bias depends on the behavioral<br />

assumptions of the firm, which will vary<br />

between industries <strong>and</strong> the type of PPI index<br />

required—input or output PPI.<br />

• Laspeyres <strong>and</strong> Paasche indices act as bounds on<br />

their true indices <strong>and</strong>, under certain conditions,<br />

17

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!