CONTEMPORARY ECONOMIC POLICY
amendment to the Wire Act was rolled into H.R. 4411, but when the UIGEA was added to the SAFE Port Act, the amendment was dropped (Doyle 2006, 3). Other provisions of H.R. 4411 that were dropped from the UIGEA included an increase in the maximum prison term for vio- lation of the Wire Act from 2 to 5 years and an authorization of $40 million in appropriations spread over 4 years for enforcement of the Wire Act (Doyle 2006, 5).14
We use probit regressions to investigate the relationship between representatives’ votes on H.R. 4411 and factors that measure the presence of gambling in the district of each representative, relevant economic and demographic informa- tion, campaign contributions, and the represen- tative’s party affiliation. Probit is an appropriate estimation method in this case, since the depen- dent variable is a binary choice variable (Greene 2003, 665 – 89). Probit has been used in previ- ous voting studies; for example, Lopez (2002), Abetti (2008), and Broz (2008) all use probit to examine congressional voting patterns.
The variables used in our regression models are as follows:
AYE = 1, if the representative voted for H.R. 4411; zero otherwise.
REPUBLICAN = 1, if the representative was registered Republican; zero otherwise.
GAMBLE_CONTRIB = the dollar amount of contributions given to each congressman by commercial and tribal casinos, racetracks, raci- nos (racetracks with casinos), and other pro- gambling interest groups (and their employees) from November 6, 2002 to July 10, 2006 (thus the contributions measured start with the previ- ous election cycle and end the day before the bill was voted on).
RELIGIOUS_CONTRIB = the dollar amount of contributions given to each congressman by religious groups (and their employees) from November 6, 2002 to July 10, 2006.
SPORTS_CONTRIB = the dollar amount of contributions given to each congressman
14. In addition to these extensions of the Wire Act that were included in H.R. 4411 but absent from the UIGEA, there are a few other more minor differences between the two. Thus, for example, H.R. 4411 allows the seizure of funds from an account maintained by an insured depository institution if the account is owned or controlled by a gambling business that violates the Act. The UIGEA, however, does not provide for injunctions on financial transaction providers, except in cases of fraud. H.R. 4411 prohibits “information assisting in the placing of bets or wagers” by a gambling business, whereas the UIGEA prohibits only information pertaining to the movement of funds to or from an account used for gambling purposes.
by professional sports organizations, arenas, and related sports service industries (and their employees) from November 6, 2002 to July 10, 2006.
GAMBLE_NUM = the total number of com- mercial and tribal casinos, racinos, horse tracks, and dog tracks in the congressional district in 2006.
EVANGELICAL = percentage of adults in the state who consider themselves to be asso- ciated with an Evangelical Christian denomina- tion, 2007.15
INCOME = median household income in the congressional district, 2006.
Data sources are listed in the appendix.
IV. EMPIRICAL RESULTS
Descriptive statistics for the variables defined above are presented in Table 1. The sample size is 408. There were 410 votes recorded for H.R. 4411, and EVANGELICAL is miss- ing for Hawaii’s two congressional districts. Table 2 shows the results of three probit regres- sion equations, all of which have AYE as the dependent variable, for the sample of 408 (the probit coefficients, as well as the slope estimates at the mean values of the independent variables, are shown). The first set of results in Table 2 is for a probit regression which includes all of the variables listed above.16 There are four coef- ficient estimates that are statistically different from zero at a 5% significance level or better, namely, the coefficient estimates of REPUBLI- CAN, GAMBLE_CONTRIB, GAMBLE_NUM, and EVANGELICAL. Below, for ease of inter- pretation, the marginal effects, represented by the slope estimates at the means, are used. (The levels of significance for the probit coef- ficients and the estimated slopes at the means are equivalent.)
The estimate of the slope for REPUBLI- CAN when all of the independent variables are assumed to take their mean value is positive and statistically significant at a 1% level. This indi- cates that Republican congressmen were more likely to vote for the bill than their Democratic Party counterparts. In fact, 201 of 218 voting Republicans (92.2%) voted for the bill, com- pared to only 115 of 191 Democrats (60.2%).
15. District-level data for 2006 are unavailable. 16. When all 410 observations are included and EVAN-
GELICAL is excluded, the results for the remaining inde- pendent variables and the overall fit of the regression model are similar to the first regression shown in Table 2.
HALCOUSSIS & LOWENBERG: INTERNET GAMBLING 23
TABLE 1 Descriptive Statistics
Variable Mean Standard Deviation Minimum Maximum
AYE 0.774 0.418 0.000 1.000 REPUBLICAN 0.534 0.499 0.000 1.000 GAMBLE_CONTRIB 13,610 40,626 −2,500.0a 463,793 RELIGIOUS_CONTRIB 693.89 1,745.2 −200.00a 15,200 SPORTS_CONTRIB 1,857.9 3,633.9 0.000 24,800 GAMBLE_NUM 1.319 3.419 0.000 36.000 EVANGELICAL 25.755 11.362 7.000 53.000 INCOME 50,119 13,331 21,088 97,753 N 408
aNegative values represent returned contributions. These values were included as they are likely to represent an unfavorable reaction on the part of the congressman to the contribution. When negative values for contributions are replaced by a 0, the regression results are very similar (see Table 2, Regression (3)).
TABLE 2 Probit Results by Congressional District (Dependent Variable Is AYE)
(1) Probit Coefficients
(2) Probit Coefficients
(3)a Probit Coefficients
(t -stats) Variable Slopes at Means Slopes at Means Slopes at Means
Constant −0.640 (−1.37)
REPUBLICAN 1.17∗∗ (6.79)
0.297** GAMBLE_CONTRIB −7.54×10−6∗∗
RELIGIOUS_CONTRIB 5.49×10−5 (1.11)
SPORTS_CONTRIB −2.23×10−5 (−1.06)
GAMBLE_NUM 0.096∗ (2.41)
0.0235* EVANGELICAL 0.0314∗∗
0.00772** INCOME 3.40×10−6
8.37×10−7 McFadden R2 0.242 0.237 0.242 % Predicted correctly 81.1% 79.7% 81.1% N 408 408 408
aFor this regression, all negative values indicating a returned contribution have been replaced with a 0 (negative values were present for GAMBLE_CONTRIB and RELIGIOUS_CONTRIB but not SPORTS_CONTRIB).
∗Significant at 5% level; ∗∗Significant at 1% level.
To interpret the 0.297 value of this estimate, consider an imaginary “typical” congressional district that has mean values for all of the independent variables. For convenience, call
the district “Seahaven.” Seahaven would be in a state where 25.76% of the constituents are Evangelical Christians. Pro-gambling groups, religious groups, and sports entities would
24 CONTEMPORARY ECONOMIC POLICY
give Seahaven’s congressman the campaign contributions shown as the means in Table 1, namely, $13,610, $693.89, and $1,857.90, respec- tively. Seahaven has one gambling establish- ment, a Republican congressman, and the median household income in Seahaven is $50,119.
Suppose that Seahaven had elected a Demo- cratic congressman instead of a Republican. The fitted value from the probit results shown in the first column of Table 2 would be 0.33. Recall that fitted values for probit regressions can be interpreted as z-scores, and then the cumulative normal probability distribution can be used to find the estimated probability of a district’s rep- resentative voting for the bill.17 The correspond- ing probability for the 0.33 fitted value is 0.63, meaning that there is a 63% chance that Sea- haven’s Democratic congressman would have voted for H.R. 4411. Since the slope estimate at the means for REBUPLICAN is 0.297, if Sea- haven’s congressman were Republican instead, the probability that the congressman would vote for H.R. 4411 would increase by 0.297 ceteris paribus so that there is now a 93% chance the congressman would support the bill. This shows the importance of party affiliation in this vote.
Here, and in the discussion below, keep in mind that the mean of AYE is 0.774, so that there is an overall unconditional 77% probability that a representative votes for H.R. 4411. The probability that Seahaven’s representative votes for the bill does not come out to 77% because we assumed that Seahaven’s congressman belongs to a particular party, not that he is somehow 53% Republican and 47% Democrat, which would be the average but would make no sense in this context. Below, we assume that Seahaven’s representative is a Republican, since there were more Republicans than Democrats in the House of Representatives at the time that H.R. 4411 was voted on. Likewise, we assume that Seahaven has one gambling establishment, not the mean of 1.34.
The estimated slope at the means for the gam- bling industry’s political contributions given to each district’s representative (GAMBLE_CON- TRIB) is significant at a 1% level with a neg- ative sign, meaning that the greater the amount of contributions a representative receives from pro-gambling groups, the more likely he would be to support online gaming by voting against H.R. 4411. This result is plausible since these
17. See Becker and Waldman (1989).
groups might see online gambling as a way to get people interested in gambling in gen- eral and promote traditional casinos and race- tracks. Industry groups dominated by land-based casino companies might therefore be opposed to H.R. 4411. Recall that the value of political contributions by gambling groups in Seahaven, our typical district, is the mean value, $13,610, and that there is a 93% probability that Sea- haven’s Republican congressman would vote for H.R. 4411. Now suppose that these contribu- tions increased by $1,000. Given the −1.85 × 10−6 slope at the means estimate, the probabil- ity that this district’s representative would vote for H.R. 4411 would decrease by approximately 0.185%.18 (The slope estimate is small because GAMBLE_CONTRIB is measured in dollars, and a one unit or one dollar change would have a small effect on the politician’s vote.)
The estimated slope at the means for the number of gambling establishments in a district, GAMBLE_NUM, is positive and statistically significant at a 5% level, indicating that the more places there are to gamble in the district, the more likely that district’s representative would be to vote in favor of H.R. 4411. Although, as we have seen, contributions from the gambling industry are associated with opposition to H.R. 4411, the presence of gambling establishments in a district is correlated with support for the bill. While national organizations may have per- ceived online gaming as a way to promote the gaming industry, it is reasonable to suppose that local casinos, racetracks, and similar establish- ments viewed H.R. 4411 as a way to eliminate competition from online casinos, in which case these constituents would have lobbied their local representative to vote for the bill in order to protect land-based businesses and employment within the district. We have assumed that Sea- haven has one gambling establishment within its borders, but suppose now that Seahaven had two gambling establishments instead. Using the slope at the means estimate of 0.0235, the prob- ability that Seahaven’s congressman would vote
18. It is approximate because the slope at the means estimate is most accurate for a small change right at the variable’s mean. In addition, for substantial changes in the independent variable, the effects are not symmetrical with regard to the change in the representative’s probability of voting for H.R. 4411. For example, a $10,000 increase in GAMBLE_CONTRIB would decrease the probability that the congressman votes for the bill by more than a $10,000 decrease would increase the probability that he votes for it. This asymmetry occurs because the z-scores behind the probabilities that are relevant here are located toward the right-hand side of the normal probability distribution.
HALCOUSSIS & LOWENBERG: INTERNET GAMBLING 25
for H.R. 4411 would increase from 93% to approximately 95%.
EVANGELICAL also has a statistically sig- nificant slope at the means estimate. This pos- itive and highly significant result (1% level) indicates that voters associated with Evangeli- cal churches generally do not approve of online gambling and that their representatives are more likely to support H.R. 4411. If the percent- age of adults in Seahaven’s state who asso- ciate themselves with an evangelical Christian denomination increased by one, the probabil- ity that Seahaven’s representative would vote for H.R. 4411 would increase by the slope at the means estimate, 0.00771 or 0.771%. Over- all, the regression correctly predicts 81.1% of congressional votes.
Clearly, party affiliation is critical here. Polit- ical contributions from the gambling industry, the number of gambling establishments in the district, and the percentage of evangelical Chris- tians in the state are also relevant. However, median household income, as well as two other variables which measure political contributions, were not close to having statistically significant coefficients at a 5% level. Variance inflation factor tests as well as an examination of the correlation coefficient matrix do not reveal any multicollinearity problem that would be gener- ating this lack of results. As a test of robust- ness, the second regression equation reported in Table 2 shows probit results with the indepen- dent variables that had insignificant coefficients removed. The results are very similar in terms of the probit coefficient estimates and slope at means estimates, significance levels, and the percent predicted correctly (79.7% instead of 81.1%).
As pointed out in a footnote to Table 1, the variables that measure political contribu- tions contain some negative values represent- ing returned contributions. The third regression model in Table 2 reports the results when these negative values are replaced by zeros in the data set; these results are very similar to those that include the negative values as reported in the first regression.
We value our customers and so we ensure that what we do is 100% original..
With us you are guaranteed of quality work done by our qualified experts.Your information and everything that you do with us is kept completely confidential.
You have to be 100% sure of the quality of your product to give a money-back guarantee. This describes us perfectly. Make sure that this guarantee is totally transparent.Read more
The Product ordered is guaranteed to be original. Orders are checked by the most advanced anti-plagiarism software in the market to assure that the Product is 100% original. The Company has a zero tolerance policy for plagiarism.Read more
The Free Revision policy is a courtesy service that the Company provides to help ensure Customer’s total satisfaction with the completed Order. To receive free revision the Company requires that the Customer provide the request within fourteen (14) days from the first completion date and within a period of thirty (30) days for dissertations.Read more
The Company is committed to protect the privacy of the Customer and it will never resell or share any of Customer’s personal information, including credit card data, with any third party. All the online transactions are processed through the secure and reliable online payment systems.Read more
By placing an order with us, you agree to the service we provide. We will endear to do all that it takes to deliver a comprehensive paper as per your requirements. We also count on your cooperation to ensure that we deliver on this mandate.Read more