(COND(X)) indicate that multicollinearity is not a se-
rious problem.
If the errors of the two equations are mutually in-
dependent, direct regressions would be preferred in es-
timating equation (1). Under these conditions, both
OLS and 2SLS yield unbiased estimates, but OLS es-
timators will be more efficient. However, if the errors
of the two equations are correlated, the coefficients,
from direct regressions will be biased while 2SLS still
yields unbiased estimators (Malinvaud). Since there
was reason to suspect correlation in the errors of the
two equations, the 2SLS coefficients were used in
computing the elasticities reported in Table 3.
Preliminary investigations considered QINPUT as a
determinant of the adoption rate, but the hypothesis that
its coefficient was equal to zero could not be rejected,
so this variable was not included in equation (2). An
alternative formulation of the model with the price
variables entered separately, as in equations (1) and (2),
found the coefficients of PMDP and PSUB to have the
expected positive (own-price) and negative (cross-
price) signs, respectively, with both significant at the
.05 level. Distributed lag relationships on the relative
price variable were also examined for equations (1) and
(2), but none proved as satisfactory as the more sim-
plistic formulation reported in Table 2.
REGULATORY POLICY ANALYSIS
Mechanical deboning is not unique to the poultry-
processing industry. The technique was first devel-
oped for the Japanese seafood industry over 20 years
ago and was later adapted for use on poultry and red
meat bones. In the United States, the supply response
to the technology by the red meat and poultry indus-
Table 3. MDP Supply Response Model Elasticities
Elasticity Elasticity
Measures Valuea
Elasticity of QMDPfc
with respect to:
(PMDP∕PSUB)t~1 .20
QINPUTfc .25
ADOPTfc 1.51
Elasticity of ADOPT
with respect to:
(PMDP∕PSUB)t~1 .09
a) Calculated at scaled sample means of
QMDPt = 22,203, (PMDPZPSUB)t.ɪ = 166,
QINPUTt = 1,084, ADOPT, = 7 X IO2, TREND = 125
tries has been quite different. In 1979, almost 330 mil-
lion pounds of MDP were produced, compared to less
than 3 million pounds of mechanically deboned red
meat (MDM). Nearly 10 percent of the federally in-
spected poultry-processing establishments had adopted
the technology, compared to an insignificant number
of red meat processors.
Some of the disparity between production of MDM
and MDP may be due to differences in the physical
characteristics of red meat and poultry carcasses, which
affect the extent to which bones can be utilized, as well
as yields (Fields). However, USDA calculations tak-
ing into account the potential supply of useable inputs
as well as their expected yields estimated the potential
supply of MDP to be about twice the amount actually
produced in 1979. The potential supply of MDM in that
year was estimated to exceed 100 times the amount ac-
tually produced.
The large disparity between the ratio of actual pro-
duction to potential production for MDP and MDM has
led most researchers to attribute the difference in sup-
ply responses to differences in the federal regulatory
environment [CFR]. The markets for MDP were de-
veloped in an environment virtually free of regulations
governing ingredient standards, use limitations, and
labeling requirements.1 But the potential producers of
MDM were confronted with a more constraining reg-
ulatory environment with provisions that they believed
created a negative image of MDM and products that
might contain it.2
The possible effects of federal regulations on the
supply of MDM have been analyzed elsewhere (Bul-
lock and Ward; McNiel; Williams). The contrasting
relatively free market in which the MDP supply re-
sponse to the deboning technology evolved has been
virtually ignored by policy researchers. The MDP sup-
ply-response model developed here facilitates our
ability to assess the implications of a recent report to
Congress [GAO] recommending that institutional dis-
parities in the marketplace between the two ingredi-
ents be removed by extending the recently promulgated
MDM regulations to MDP (USGSA).
While a more rigorous impact analysis must await
specific regulatory proposals, a general approach for
that analysis can be suggested. If nutrient standards
should reduce either the type or amount of potential
poultry inputs that may be used for producing MDP,
the amount of MDP marketed will be reduced. De-
pending upon the extent to which blending is allowed,
imposition of maximum fat and minimum protein
standards, or restrictions on the use of MDP from fowl
may eliminate some potential raw material sources for
adopting firms. Since actual production of MDP is
presently less than potential production, however, reg-
ulations that restrict the supply of potential inputs for
producing MDP will not result in a proportional re-
1 The only ingredient standard for MDP is a 1 percent maximum on-bone content (9 CFR 381.117 [d]). MDP may be used in any product in which poultry or poultry meat is allowed,
including some processed red meat products such as bologna, knockworst, and frankfurters (9 CFR 319.180). MDPmay comprise upto IOOpercentof the poultry portion of a meat or poultry
food product, and no special labeling is required. Products are simply Iabled by kind (chicken or turkey) if meat, skin, and fat are present in natural proportions, or as (kind) meat if no fatty
tissue or skin are included (9 CFR 381.117).
2 Duringtheperiodanalyzed, regulations governing MDM as an ingredient included standards for bone content and particular size, calcium, fat, protein quantity and quality (9 CFR 319.5).
Use of MDM was limited to no more than 20 percent of the meat portion of any final meat food product (9 CFR 319.6 [b]). MDM could not be used at all in products such as ground beef and
hamburger (9 CFR 319.6 [c]). The name on the label of a final product containing MDM had to be qualified by the phrases, “With Mechanically Processed (Species) Product,” “Contains
Up to____% Powdered Bone,” and the ingredients statement had to list MP(S)P in order of predominance by weight (9 CFR 317.2 [j] [13]).
136