Table 3. Determinants of Recreation Participation | ||
Variable___________ |
Coefficient Estimate______ |
Asymptotic |
Constant |
-3.684“ |
-2.00 |
1 n (Trip Cost) |
-0.779* |
-6.16 |
Gender |
0.643“ |
2.09 |
Age |
0.139“ |
2.07 |
Age Squared |
-0.001“* |
1.89 |
Education |
-0.065 |
-1.13 |
Children |
0.328" |
2.23 |
Hourly Wage |
0.036*** |
1.93 |
Urban |
-0.676*** |
1.73 |
Conservationist |
1.355* |
3.70 |
Z2 |
90.50 (9 d.f.) | |
McFadden’s R2 |
.232 | |
Sample Size_____ |
477__________ |
*, **, ***, indicate significance at the a = .01, .05, and
.10 levels, respectively.
pants. The participation equation performed well
statistically according to the Chi-square statistic and
McFadden’s R2 statistic (Amemiya 1981).
Empirical results showed that there was a negative
relationship between recreation participation and
trip cost. This result was consistent with economic
theory: as the cost of an activity increases, participa-
tion in the activity declines. Coefficient results on
the standard explanatory variables were consistent
with previous studies as described by McConnell
(1985). Participation was more likely if the survey
respondent was male and did not live in an urban
area. Participation increased at a decreasing rate
with age, and increased with number of children and
income (WAGE). The membership in environ-
mental and conservation organizations variable
(CONSERVATIONIST) was included to account for
leisure activities that may be complementary with
recreation participation. For instance, reading
magazines, newspapers, or organizational literature
will increase information about recreational area
availability and wetlands-related activities. It was
expected that this type of behavior will increase
recreation participation. Empirical results showed a
strong positive relationship between leisure behavior
(measured by organization membership) and recrea-
tion behavior.
Use Value Estimates
Current and forecast use values were estimated by
the one-step method and are presented in Table 4.
Use values were weighted to account for the over-
sampling of coalfield households.6 Six percent of
the sample was predicted to participate in wetlands-
related recreation during 1990 and have positive use
values.7 The average use value per trip during the
1990 season was $5.16, ranging from $0.12 to
$25.64. The median use value was substantially less
than the mean suggesting a skewed distribution of
use values. Use value per season was found by
multiplying use value per trip by the sample
weighted average number of trips (trips = 4.63, n =
27). Each expected participant is expected to enjoy
a use value of $23.89 each season.
Three forecasts of use value for the year 2000 were
made. The Kentucky population was forecast to age
by 2.5 years and real household income was forecast
to increase by 3.7 percent (WAGE increases by
$5.83) between 1990 and 2000.8 Each forecast and
a combination of both were examined. All forecasts
had mostly neutral effects on the probability of par-
ticipation. The aging of the Kentucky population
will leave the number of participants about the same.
The income increase will increase the number of
participants, but this effect will be reduced by the
increased trip cost from the increased opportunity
costs of time. Average use values per trip ranged
from $5.93 to $7.49 for the three forecasts. Use
values per trip increase as the expected participants
change in the future. Again, the distribution of use
value was nonnormal with the median less than the
mean for each group. Use value per season ranged
from $27.44 to $35.96.
Aggregate use value was estimated by multiplying
participants as a percent of the sample by the Ken-
tucky household population to get the number of
forecast participants. The number of forecast par-
ticipants was multiplied by use value per year to get
aggregate use value. The forecast use values were
calculated using a population projection for 2000.
Use value during 1990 was estimated as $351,183.
With increasing participation rates and increased
population, use value is expected to increase from
1990 to 2000. If both age and incomes increase as
expected, aggregate use value will increase by 73
percent to $609,090.
6Households in the three-county recreation region and outside Kentucky represented 54 percent of the sample and 10 percent of
the population. Households in the rest of Kentucky represented 46 percent of the sample and 90 percent of the population. The
weights were equal to the percent of the population divided by the percent of the sample for each group.
7This number is less than one-half of the observed participants. This small number may be a result of the use of logistic
regression. Logistic regression tends to Imderestimate the number of recreation participants when participation is low.
8These forecasts were made using data from the Kentucky Statistical Abstract, 1988.
117