قمار خانه های جدید و بازارهای کار محلی: شواهد از کانادا

The local labor market effects of new casinos are examined by comparing the employment and earnings growth in areas with new casinos to the growth in areas with existing casinos and without casinos, exploiting numerous casino openings across multiple locations in Canada over several time periods. The opening of a new casino directly doubles the employment and earnings of the local gambling industry within five years, while this growth does not appear to continue beyond this period. Indirect positive spillovers are limited to the related local hospitality and entertainment industries. For every job created in the gambling industry, roughly one to two additional hospitality jobs are created. Increased gambling employment does not appear to increase employment in any other local industry.

New casinos have the potential to generate both positive and negative impacts on the local economy, including tangible benefits such as local economic development and increased tax revenues, and negative consequences like increased problem gambling, crime, bankruptcy, and traffic (Eadington, 1999). Because these benefits and costs are of public concern, governments closely regulate the supply of casinos and typically require some positive economic benefit to outweigh any of the negative costs when expanding access to legal gambling. One common justification for a new casino is that it will lead to new job creation and enhanced earnings in the local labor market, due to a casino-induced increase in local labor demand, which may occur only in the gambling industry or in other local industries as well.1 While it has been claimed that new casinos can generate significant spillovers outside of gambling, there is little consistent evidence to support this claim. For these reasons, it is important to properly assess these local labor impacts, in order to quantify just how many jobs may be created, or perhaps even lost, both inside and outside of the gambling industry. The identification of the impacts of a new casino on a local labor market can be difficult.2 Proper identification must rely on both the location and the timing of new casinos, as well as on an appropriate counterfactual addressing what would have happened if a new casino had not been built. Without proper identification, these impacts could be overstated or understated, such as where underdeveloped locations with relatively low employment and earnings are targeted for a new casino or when the timing of a new casino opening is more influenced by the conditions of the overall economy rather than by the conditions specific to the locality. Previous research examining these effects had focused on the United States (Rephann et al., 1997, Evans and Topoleski, 2002, Garrett, 2004 and Cotti, 2008), but this may not be an ideal setting, as casinos there tend to be clustered in either highly-agglomerated tourist destinations or in remote areas, and are likely to be privately operated, making it difficult to disentangle the impact of new casinos from other effects. This paper analyzes the effects of new casino openings on local labor markets in Canada. Canadian casinos are distributed more uniformly across the country, are more likely to be located in populated areas, and are more likely to be government run than casinos in the United States, all of which enhance the identification of these impacts. In addition, the restricted-access data from the Canadian Census of Population contain detailed geographic and industry identifiers which allow for the precise definition of the local labor markets and the separation of the direct impacts in the gambling industry from the indirect impacts in non-gambling industries. Under the quasi-experimental identification strategy of this study, treatment areas with new casinos are paired to two unique comparison areas, areas with existing casinos and areas without casinos, which are used to fill in the counterfactual for the treatment areas. All of the estimation techniques generalize over three five-year time periods, so that each estimate is interpreted as the average impact of a new casino within a locality, making it less susceptible to the influence of any single time period. First, the direct growth within the gambling industry is estimated separately for each of the three area types. Second, the indirect growth is measured for the related and other local industries, by estimating the differential growth in labor outcomes between the treatment areas and each of the comparison areas. A nearest-neighbor matching estimator is additionally used to further restrict the comparison sets. Third, the local job multipliers are estimated, which have not been previously used in the literature to address the impacts of new casinos. This is done while taking into account the endogeneity of the employment relationship between the gambling and non-gambling industries through instrumentation. The results of this study confirm the existence of positive local labor market gains following the opening of a new casino. The local gambling industry experiences the direct impact, which is a doubling of its employment and earnings in areas with new casinos, within one to five years after the casino opening. These effects were insignificant in areas with existing casinos, however, suggesting that the local effects of a new casino are short-lived. The indirect spillover effects were also positive and significant but mainly limited to differential employment growth in the closely related local industries of hospitality and entertainment, specifically accommodation, food, and beverage services and other amusement and recreation services. For every job created in the gambling industry due to a new casino, one to two additional jobs are created in the hospitality industry. Contrary to some previous findings in the literature, there are no significant employment or earnings effects in the other local industries of construction, retail trade, or all other services.

The local labor market effects of new casinos are examined in this paper in order to answer whether the opening of a new casino generates greater employment and earnings within the locality. The paper contributes several improvements to the literature in the identification and measurement of these potential impacts. First, it is the only paper to comprehensively investigate these effects for Canada, which offers a more uniform distribution of casinos located within more populated areas relative to the United States. Second, it uses detailed geographical descriptors available in the restricted-access Census data to establish proper local labor markets, which then define the treatment areas (with new casinos) and two unique sets of comparison areas (with existing casinos and without casinos). Third, the detailed industry descriptors in this data allow for the targeting of specific industries where these effects might take place, either directly in the local gambling industry or indirectly in local non-gambling industries. Fourth, all of the estimation techniques used in this paper are generalized over three consecutive five-year periods, diluting the influence of any particular period. The direct labor market growth in the gambling industry shows that areas with new casinos experience large, positive employment and earnings growth within one to five years following the opening of a casino. However, this growth was insignificant for areas with existing casinos, suggesting that the local effects of new casinos do not extend beyond five years. The indirect differential growth for the related and other local industries is estimated between treatment areas with new casinos relative to the comparison areas with and without existing casinos, using a matching estimator for further refinement. These indirect growth effects were mainly confined to differential employment growth in the related local industries of accommodation, food, and beverage services and other amusement and recreation services. Contrary to the previous evidence, no significant employment or earnings effects were found in construction, retail trade, or all other services. In order to properly identify any spillover or crowd out effects, a local job multiplier analysis is performed which links the magnitude of the direct employment effect in the gambling industry to the magnitude of the indirect employment effects in local non-gambling industries. The results ignoring potential endogeneity suggest that there are negative crowd out effects for jobs in other amusement and recreation services, as well as positive spillover effects for jobs in construction and retail trade. However, once endogeneity is accounted for through the use of two different instrument sets, the only significant multipliers are found in the closely related accommodation, food, and beverage services industry. More specifically, for every job created in the gambling industry due to a new casino, one to two additional jobs are created in this hospitality industry, results which are robust to the choice of instrument and comparison set. No other local industry experiences job gains or losses related to the creation of casino gambling jobs. Communities around the world continue to struggle with the question of how much casino gambling should be allowed or whether casino gambling should be available at all. Much of this debate focuses on the potential economic benefits that casinos can generate within a local economy, and whether these potential benefits can outweigh any of the potential social costs. Only through the proper measurement of these benefits can policy makers determine the right amount of casino gambling to allow. The evidence presented in this paper suggests that a skeptical approach be taken regarding the use of employment and earnings gains to justify the legalization or expansion of casino gambling within a locality. Any expectations of new jobs or earnings enhancement should be considered short-term and narrowly-focused within the gambling and hospitality industries. Broad employment and earnings gains in other local industries outside of gambling and hospitality should not be expected.