From: michael perelman (michael@ECST.CSUCHICO.EDU)
Date: Thu Dec 23 2004 - 13:15:22 EST
Sorry, I put up the wrong sections.
_Replacement
Investment Patterns_
Because the installation of long‑lived capital goods is likely to
disrupt production, one might expect that long‑lived investments would
be installed or significantly upgraded during contractions when plant
and equipment is used less intensively (S. Moss, 1984, p. 297). Because
of the British capital shortage in the years immediately following the
Second World War, the ratio of capital disposals to capital acquisition
stood more than 50 per cent lower than it was in the 1950s (Butler,
1960, p. 261). More dramatically, 'in the panic of 1920‑1921, Ford
closed down for six weeks, cleared his plants of obsolete equipment, and
then proceeded to improve working methods and layout and to install
modern equipment. Also in 1920, the Lelands spent $4,249,000 for
special machine tools to produce postwar Lincolns' (Wagoner, 1968, p.
126; see also p. 136). Most firms did not respond to the depressed
economic conditions in this fashion. Commentators took notice of Ford's
policy only because it seemed so exceptional.
Firms do most of their replacement when competitive pressures
compel them to do so (Boddy and Gort, 1971, p. 182; S. Moss, 1984, p.
297). As a result, productivity increases tend to be associated with
business failures (Montgomery and Wascher, 1986). For example, during
the Depression, the quantity of machine tools shipped in the US dropped
sharply from a peak of 50,000 units in 1928 to a low in 1932 of 5500 but
the equipment that was shipped during the Depression was predominantly
for replacement rather than expansion (Wagoner, 1968, p. 137). Prior to
the 1930s, machine tool manufacturers used to purchase the best
available equipment. In the wake of the Depression, 'machines were
replaced only when they clearly could not do the work required'
(Wagoner, 1968, p. 136), perhaps because the intensification of
competition on the demand side was compensated by reduction in wages and
a higher value placed on liquidity.
During the Depression, firms weeded out inefficient plant and
equipment, creating a much newer capital stock (Staehle, 1955, p. 124).
By 1939, one‑half of all manufacturing equipment in the US that had
existed in 1933 had been replaced (Staehle, 1955, p. 127). Thereafter,
business produced as much output as a decade before with 15 per cent
less capital and 19 per cent less labour (Staehle, 1955, p. 133).
French productivity also improved noticably during the Depression
(Aldrich, 1987, p. 98, citing Carr'e, Dubois and Malinvaud, 1972).
Similarly, in the recessionary period of 1982‑4, only 20 per cent of
West German manufacturers replying to IFO's investment survey gave
capacity expansion as their motive for investment; 55 per cent cited
rationalization (Anon., 1985a, p. 69).
Once the economy begins to prosper, scrapping returns to its
normally low level. During expansions, firms tend to increase the
proportion of investment devoted to long‑lived capital of the sort that
expands capacity (Boddy and Gort, 1971; see also Mairesse and Dormont,
1985). During these periods firms feel little pressure to replace
obsolete plant and equipment (Bleany, 1985, p. 77). Costs are less
concern than the opportunity to expand capacity. For example Schmenner
reports that space constraints and poor plant layout are much more
important than labour costs in explaining why plants move to new
locations (Schmenner, 1980; see also Boddy and Gort, 1971).
In recent times, fashionable theory explains investment by
factors that affect the supply of capital, especially taxes. Supposedly
the interaction between inflation and the tax structure theoretically
made long‑lived investment relatively less economical (Auerbach, 1979).
Tannenwald emphasized the combination of the tax code and the rising
cost of structures, along with the changing mix of economic activities,
as an explanation for the declining share of investment used for
structures (Tannenwald, 1982; Ott 1984). This result was based on the
official construction cost index for industrial and commercial
structures derived from the price structure for single family homes,
which rose almost 50 per cent faster between 1974 and 1981 than the
corresponding deflator for producers' durable goods (Eisner, 1982, p.
106). However, industrial and commercial structures' costs were rising
much more slowly than the costs of single family homes (Allen, 1985).
Moreover, inflation could make industrial and commercial structures
relatively more attractive as an inflation hedge, just as was the case
with residential structures during the early 1970s. Finally, recall
Eisner and Chirinko's finding that firms are insensitive to tax
incentives (see Ch. 3).
The tax argument cuts both ways. Although taxes might make
consumption more attractive than investment, tax shelters might tend to
channel a greater portion of investment in office buildings and other
structure‑intensive activities, such as oil and gas drilling (Garner,
1986). The pattern of resale prices for structures suggests the
importance of the benefits which the tax codes offer to investment in
structures. The estimated depreciation rates of assets, based on the
market prices for second‑hand capital, implies that the aggregate
equipment stock is quite close to that of the Bureau of Economic
Analysis. In contrast, a similar estimate for the market value of
non‑residential structures was more than 30 per cent greater than the
official estimates (Hulten and Wykoff, 1981a; and 1981b).
These estimates might suggest that favourable tax treatment of
office buildings promotes a buoyant resale market for real estate
compared with the market for other second‑hand capital goods, which are
less easily transferred to other uses. For example, in 1984, commercial
real estate investment was more than 50 per cent higher than its 1979
level. Construction by the industrial sector, which is presumably more
specific than commercial real estate, was less than 74 per cent of its
1979 level (Magdoff and Sweezy, 1985, p. 7; see also Garner, 1986).
Other explanations are possible. The divergence between the
results for equipment and for buildings could be explained by the
peculiar estimating procedures used (Harper, 1982). It could also be
the result of the lemon effect for equipment (discussed above), but this
phenomenon, if it is operative at all, would probably not be sufficient
to explain the full difference.
_The Contradictory Nature of
Demand Management_
Notwithstanding the complications associated with the empirical
application of the q‑theory, the q‑ratio approach suggests a promising
way to analyse replacement investment which throws a great deal of light
on the post‑war history of the US.
Common sense suggests that a buoyant economy with high asset
values encourages the purchase of new capital goods, thereby promoting
economic efficiency. Further reflection reveals that the same policies
which lifted the q‑ratio also impaired efficiency in the long run by
inhibiting the replacement of existing investments. Such policies are
especially dangerous when combined with a weakening of the wage
pressures which might otherwise compel business to replace capital.
Much obsolete plant and equipment is operated, or at least
stands ready to operate, alongside modern operations. In the US
automobile industry during the late 1960s the productivity per employee
of the top quarter plants was 2 1/2 times as productive as the lowest
quarter (Melman, 1983, p. 184). Such disparities in productivity became
even more extreme during the 1970s as a result of the continued deferral
of scrapping.
Carter's input‑output study of US industry suggests the same
phenomenon. She set up a linear programming problem to minimize the
total factor requirement in order to produce the 1958 vector of output
using either 1947 or 1958 technology in each industry. In 14 of 76
sectors, the 1947 technology was superior (Carter, 1970, p. 171‑2).
Sato and Ramachandran note replacement was less vigorous in these 14
industries (Sato and Ramachandran, 1980, p. 1007). Much of the 1958
technology in these industries might simply be the 1947 technology after
11 years of operation. Indeed, the rate of productivity of the US
economy seems to have fallen despite evidence that the rate of
improvement in new capital goods has accelerated in recent decades
(McHugh and Lane, 1987).
The average capital good is now being used more intensively
than it was in the past. A 1976 follow up survey of a 1929 US Census of
Manufacturing query about the hours worked by fixed capital found that,
from 1929 to 1976, the average work‑week of fixed capital rose by about
25 per cent, or 0.475 per cent per annum. During this same period the
ratio of fixed capital to manufacturing output fell by 45 per cent
(Foss, 1981 and 1981a and 1985). In part, this trend is associated with
more extensive shift work, rising from 22 per cent of the workforce in
the 1960s to 28 per cent in 1972‑5 (Bosworth, 1982). The US data for
1973‑8 are unchanged (Hedges and Sekscenski, 1979). The rise in shift
work in the US is far milder than in France, where the per centage of
French workers in shift work rose from 28 to 77 per cent in metal
production between 1957 and 1974; from 8 to 39 per cent in machine
building; and from 34 to 50 per cent in textiles (Lipietz, 1982, p.
223). Even after adjusting for plant hours, the ratio of fixed capital
to manufacturing output ratio dropped by 30 per cent.
Other things being equal, one would expect a trend toward a
rising capacity utilization, as shift work became more common. Yet the
evidence does not support the existence of such an increasing trend
because of a changing distribution of the utilization patterns of
capital. The most modern plants are used intensively, while many
obsolete plants are used little, if at all. A similar pattern exists in
manufacturing in the Less Developed Economies, but there it is explained
by the difficulty in importing spare parts, a problem that does not
affect plants in the Developed Economies (Kibria and Tisdell, 1984, p. 64).
Many of the less utilized plants represent what I will
describe later as phantom capacity. Again, the data are not clear
enough to prove my point decisively, but they are consistent with the
idea of an economy carrying an increasing load of obsolete equipment
during the post‑war period.
The perverse long run relationship between the q‑ratio and
productivity will figure largely in what follows. When the q‑ratio is
high, firms tend to install new investment alongside the old facilities
instead of replacing existing investments with more efficient plant and
equipment. More recently, a growing scepticism about the efficacy of
expansionary economic policies has been responsible for a perception of
uncertainty that makes older, depreciated capital goods more attractive.
This process is relatively new, dating back to the decline in
competition, which began in the late nineteenth century. Previously the
US economy was characterized by rapid replacement brought on by high
wages and rapid technological advance (see below). Beginning with the
deferred replacement policies of the twentieth century, scrapping became
limited until slowdowns of crisis proportions caused massive waves of
scrapping such as occurred during the Great Depression, when much of the
excess capacity from the First World War boom was wiped out. After the
Second World War, the process of accumulating excess capacity began
anew. Renewed scrapping began only after a long period of stagnation,
extracting a dreadful human toll in the process.
Consider the relationship between the q‑ratio and capital
replacement. All q‑ratios do not move uniformly. With the onset of
prosperity, capital goods devoted to some specific cyclically sensitive
uses might become relatively more valuable. Their q‑ratios would also
rise relative to others.
Why would firms not take advantage of the high market prices
for existing capital goods by selling them rather than maintaining them
in operation? In many cases, firms did just that. At other times, they
did not.
In the first place, existing capital goods can serve as the
basis for loans or the issuing of equities. Additionally a firm that
sells its capital goods earns a fixed amount on the second‑hand goods
market, but it loses the option to profit from the use of that good
should some new opportunity arise. The more uncertain the environment
is, the more likely that such an opportunity will actually occur.
Consequently firms might rationally hold on to capital goods even when
their salvage values exceed the expected present value of the cash flows
earned from their operation by a considerable amount (McDonald and
Siegel, 1986, p. 711).
Since managers, unlike shareholders, are more concerned with
expansion than the rate of return on investment, they may be inclined to
hold on to existing capital goods. Donaldson's survey of Fortune 500
companies led him to conclude: 'the financial objective that guided the
top managers ... [was] the maximization of corporate wealth. Corporate
wealth is _that wealth over which management has effective control_ ...
Corporate wealth differs considerably from the shareholder value which
is central to much financial theory' (Donaldson, 1984, p. 22).
Management might want to hold on to plant or equipment, even if it were
technologically obsolete, because it appears to add to the value of the
firm.
Despite my earlier insistence on the existence of active
second‑hand capital goods markets, transaction costs introduce a wedge
between the value of capital goods to a firm and its value on the resale
market. These transaction costs discourage the marketing of existing
capital goods.
Delivery lags may explain why firms might not replace highly
priced used capital goods. Additionally the market for some second‑hand
equipment becomes weaker during prosperous times. For example, old
tractors are traded in for new models when times are good, thereby
diminishing second‑hand prices relative to new capital costs (Reid and
Bradford, 1983, p. 329). Lacking sufficient incentive to sell such
equipment, old tractors may be kept on for later contingencies. During
downturns, farmers, like other managers, frequently conserve liquidity
by continuing to operate existing equipment. In the process, the
tractor stocks age. Once boom times return, sales of new tractors
expand once again. The potential tax burden of good years reinforces
this pattern. In reality, tractors are atypical capital goods since
their prices are pro‑ rather than counter‑cyclical, partly because the
pride that farmers take in their machinery makes tractors into a
quasi‑consumption good. Recall the high second‑hand capital prices of
the 1960s. Most other types of obsolete capital goods can only be
operated at high levels of demand because the capital services that they
provide are not as homogeneous as those provided by tractors. A huge
number of outdated adding machines are not equivalent even to a modest
computer.
I will turn to a more important explanation for the continued
use of old and seemingly obsolete plant and equipment. I will frame
this analysis in terms of a revised interpretation of the q‑ratio.
--
Michael Perelman
Economics Department
California State University
michael at ecst.csuchico.edu
Chico, CA 95929
530-898-5321
fax 530-898-5901
This archive was generated by hypermail 2.1.5 : Sat Dec 25 2004 - 00:00:02 EST