Demand Shocks and Real GDP Measurement

sizzledlickforkManagement

Oct 28, 2013 (3 years and 10 months ago)

62 views


Demand Shocks

and Real GDP Measurement


by Emi Nakamura

Discussion by Susanto Basu

Comments on

SSHRC Conference, Vancouver, June 29, 2004

Demand & Real GDP

CRIW Conference, June 29, 2004

2

The really quick summary

S

D

S’

D’

Q
0

P
0

Q
1

Demand & Real GDP

CRIW Conference, June 29, 2004

3

Index number theorist’s reaction


This is a novel and striking result


Also very simple

once explained!


Explains some puzzling facts


Demand shocks create interpretation problems for
Fisher and Paasche indices


But can attach an utility
-
based interpretation to
Laspeyres index




Demand & Real GDP

CRIW Conference, June 29, 2004

4

Macroeconomist’s reaction


One more reason to not attach a structural
interpretation to real GDP


Are problems worse than, e.g., aggregation across
consumers with differing MU’s of wealth?


Macroeconomists will ignore measurement issues
unless they matter for particular applications


Needed: Example(s) where macroeconomists will
draw wrong/misleading conclusions by using chain
-
weighted real GDP when there are demand shocks

Demand & Real GDP

CRIW Conference, June 29, 2004

5

One example of a macro application*


In a multi
-
sector economy, suppose a transitory, persistent
technology improvement in, say, sector 1 raises Q
1
(t) by X(t)


Now suppose instead a transitory demand shock that has the
same effect on the time path of Q
1
(t)


The increase in total GDP due to the technology shock will
appear smaller (because P
1

falls), but will also appear more
persistent

since P
1

is rising as the shock fades away


By contrast, the demand shock will appear less persistent,
because GDP will rise by more on impact (since P
1

will rise with
Q
1
), but then also fade away more quickly, since P
1

will fall as
Q
1

asymptotes back to zero.


Thus the persistence of GDP, which is of great interest to
macroeconomists, depends on the type of shocks


* Inspired by a comment from Carol Corrado