- Energy expenditures, perceived family well being, and energy
conservation actions were analyzed by family composition and level of
income. Data were from a three state subsample (Arizona, Colorado,
and Oregon; N = 2,633) of a larger stratified random sample of
households in the Western United States. Data were collected in the
spring of 1981 by mail survey.
Self-reported total annual energy expenditures were correlated
with scores on an Index of Well Being, a measure of the perceived level
of cut-backs in areas of consumption other than energy, due to rising
fuel prices. No significant (p ≤ .05) correlation was found between
energy expenditures and scores on the Index of Well Being. The
proportion of income spent on residential energy (energy budget share),
however, was significantly (p = .001) correlated with the Index of Well
Being (r = <.247).
Families were classified according to the age and marital status
of the head of household, and the number of dependents in the family.
One-way analysis of variance was used to test differences in energy
expenditures, the energy budget share, and scores on the Index of Well
Being between family types and families at different income levels. All
variables differed significantly (p ≤ .001) between groups. Families past
retirement age had an average energy budget share of almost twice the
amount of other families. The form of relationship was tested by fitting
linear, quadratic, and cubic contrasts to the group means. Energy
expenditures varied across stages in the family-life-cycle in the form of
an inverted U. Interactions between the two grouping factors, family
composition and income, were tested in a two-way analysis of variance.
Only in the case of the energy budget share a significant (p ≤ .05)
interaction was found. There was no significant effect of climate on
Log-linear analysis was used to find differences in the probability
with which families at different stages in the family-life-cycle and at
different income levels had taken various energy conservation actions.
Models were fitted to five-way frequency tables of energy conservation
actions by age and marital status of the head of household, family size,
and income. Conservation actions were classified as energy efficiency
improvements or curtailments. The probability of efficiency
improvements generally increased with age and income, while no clear
trend existed for curtailments. A stepwise logistic regression procedure
was employed to find socioeconomic and housing factors associated with
the differences in conservation actions between family types and
income groups. Interfamily differences in energy conservation actions
were largely determined by differences in the built environment. The
probability for having taken conservation actions was lowest in rented
multi-family dwellings, built before 1975. Weatherization programs
should, therefore, be targeted to these dwellings. Since loans were
equally used by all income groups, but tax credits more often by high
income families, a loan program would be a more equitable way to
encourage energy conservation than tax credits.