PDA

View Full Version : Changes in Household-level Demand for Postal Delivery Services from 1986 to 2004



darin
08-17-2007, 04:40 PM
I found this Gem - some of your more economics-minded folk may enjoy the 49-page read.


http://netfiles.uiuc.edu/hyunhong/www/postal.pdf


BLUF:


Our estimation results are not encouraging for future financial viability of the USPS. Our model
predicts that the recent trends in postage rate increases, the growing penetration of personal computing
technology, and the declining cost of telephone services will all reduce the amount of revenues
the USPS will receive from US households in the future. Unless the demand for USPS services from
other segments of the economy picks up this slack, the USPS is likely to experience revenue reductions
in the future even with significant rate increases.
The own-price elasticity estimates from our model indicate that rate increases are likely to significantly
reduce USPS revenues from the household sector at an even greater rate than the results
reported in Wolak (1997). Moreover, these large own-price and cross-price elasticity estimates are
even more precisely estimated than those reported in Wolak (1997). The impact of computer ownership
is also much more precisely estimated and larger in absolute value than the results reported
in Wolak (1997) for the period of rapid expansion of the Internet usage by computer users from
1994 to 2000. These estimation results predict revenue reductions from the US household sector
from future postal rate increases, telephone services price reductions, and adoption of personal
computing technology by US households, although the computer ownership elasticity for the post-
2000 period implies the future computer adoption at the household-level will not lead to as large
of revenue reductions as the adoptions that occurred during the period 1994 to 2000.
The analysis of the welfare impacts shows that postal rate increases impose greater burdens on
older, higher-educated households, located in urban areas. Although the absolute burden is greater
for higher income households, the relative burden is lower. In terms of relative impact, the age of
the head of household is a major determinant of the magnitude of both the absolute magnitude of ln(CV) and the value of ln(CV/M).
Households located in the South Census region appear to
suffer less harm than households located in the Northeast, Midwest or West census regions. In
summary, the absolute and relative welfare losses associated with postal rate increases tend to be
distributed among US according to observable demographic characteristics in a manner consistent
with our demand estimates.
Clearly, there are many caveats associated with these results. One obvious direction for future
research is to investigate alternative functional forms for both the demand for postal delivery
services and the frequency of purchase model. As noted above, we experimented with a large
number of more flexible functional forms involving squares and cross-products of the logarithms of
the prices of postage, telephone services and non-durable goods and total non-durable expenditures
and obtain similar values for the sample means of the household-level elasticities. Nevertheless,
there are other directions worth exploring to determine the robustness of these estimation results,
given their dire predictions about the future demand for USPS services from US households.
We now explore the implications of these results for future postal policymaking. First, operating
statistics from the USPS show trends that are consistent with our estimation results. According to
USPS Revenue, Pieces, and Weight Reports available from the USPS website, the volume of single
piece First-Class mail has declined every year from 2000 to 2006 at an average annual rate of close
to 4 percent per year. A substantial fraction of the postage expenditures by the household sector
is for single piece First-Class mail, so this steady and significant decline in single piece First-Class
mail volume is broadly consistent with our estimation results.
Despite this precipitous decline in single-piece First Class mail volume from 2000 to 2006, total
First-Class Mail volume has fallen at approximately 1 percent per year over this same time period
because of a steady increase in the volume of Presorted First-Class Mail.