Judgment, Supreme Court, New York County (Alice Schlesinger, J.), entered May 6, 2011, which, to the extent appealed from, granted the petition pursuant to CPLR article 78 to annul respondent Comptroller’s July 1, 2010 prevailing wage schedule for building service employees engaged as furniture movers and remanded the matter to the Comptroller to determine a new prevailing wage schedule, unanimously affirmed, without costs.
*597Petitioners, a group of companies providing moving services and a not-for-profit advocate for the New York City moving industry, brought this article 78 proceeding against respondent John C. Liu, Comptroller of the City of New York, seeking to set aside his July 1, 2010 prevailing wage schedule for furniture movers engaged on building service contracts with the City. Petitioners maintained that the methodology used by the Comptroller to set the wage schedule was irrational, arbitrary and capricious, and resulted in wages inordinately higher than those prevailing in the industry. The Comptroller opposed the petition, asserting that his method for determining the prevailing wages was sound and within his broad discretion. The motion court granted the petition to the extent of annulling the wage schedule and remanding the matter to the Comptroller to determine a new schedule. The Comptroller appeals and we now affirm.
Article 9 of the Labor Law sets forth the prevailing wage requirement for building service employees, including furniture movers (see Labor Law § 230 [1]). Specifically, Labor Law § 231 (1) provides that £<[e]very contractor shall pay a service employee under a contract for building service work a wage of not less than the prevailing wage in the locality for the craft, trade or occupation of the service employee.” “Prevailing wage” is defined as “the wage determined by the fiscal officer to be prevailing for the various classes of building service employees in the locality” (Labor Law § 230 [6]).1
Each year, the Comptroller publishes a schedule setting forth the prevailing wage rates for various trade classifications engaged in work on city contracts. The purpose of the schedule is to establish the wage rate that employers must pay their employees when contracting to provide services to the City. On April 28, 2010, in order to determine the July 2010 prevailing wage schedule for furniture movers, the Comptroller mailed surveys to all known New York State licensed commercial moving industry employers operating in the city. The surveys sought information about how many employees worked with the company, the hourly wages for those employees, and the employees’ labor union affiliations, if any.
Approximately 95 moving companies responded to the survey, representing about 2,241 employees. According to the petition, all statistical indicators in the data collected by the Comptroller showed that the wages that actually prevailed in the region were between $10 and $20 per hour. The average wage received *598by movers surveyed was $19.19; the median wage was $15; and the mode (i.e., the most commonly paid wage) was $12.00.
On July 1, 2010, the Comptroller published the final prevailing wage schedule for building service employees for the 2010-2011 fiscal year. The hourly wage rates for furniture movers in the Comptroller’s schedule are significantly higher than those reflected in the survey results. They range from $30.63 to $38.90, based on the worker’s status as a “driver” or “helper,” as well as the worker’s seniority.2 Instead of looking to the survey results to determine the prevailing wage, the Comptroller simply used the wages earned by workers covered under the collective bargaining agreement of Local 814 of the International Brotherhood of Teamsters (Local 814). The Comptroller maintains that he used Local 814’s agreement because the union represented at least 30% of the moving industry’s work force,3 and that this methodology was permitted under the prevailing wage provisions of the Labor Law.
The motion court properly granted the petition to annul the Comptroller’s prevailing wage schedule. In an article 78 proceeding, an administrative action can be set aside if it was affected by an error of law, was made in violation of lawful procedure, or was arbitrary, capricious or an abuse of discretion (CPLR 7803). An action is arbitrary if it “is without sound basis in reason and is generally taken without regard to the facts” (Matter of Pell v Board of Educ. of Union Free School Dist. No. 1 of Towns of Scarsdale & Mamaroneck, Westchester County, 34 NY2d 222, 231 [1974]). Although deference is normally given to an administrative agency’s determination, where such determination “runs counter to the clear wording of a statutory provision, it should not be accorded any weight” (Roberts v Tishman Speyer Props., L.P., 13 NY3d 270, 285 [2009] [internal quotation marks omitted]).
We find that the Comptroller’s use of Local 814’s collective bargaining agreement as the sole basis for determining the prevailing wage schedule was arbitrary, capricious, and lacked a rational basis. There is no question that the survey results show that workers in the moving industry received much lower wages than those listed on the Comptroller’s schedule. The data compiled by the Comptroller indicates that wages of between *599$10 and $20 per hour were prevailing in the moving industry, roughly half of the $31 to $39 range that he adopted. According to the survey results, over 70% of the workers received less than $20.63 per hour, a wage that is a full $10 less than the lowest wage picked by the Comptroller as “prevailing.” More than 80% of the workers received hourly wages lower than the low end of the Comptroller’s schedule. Virtually all — 96%—of the workers surveyed were paid wages lower than the high end of the Comptroller’s schedule. And, as noted earlier, no matter what statistical method is chosen — average, median or mode— the survey results reveal figures substantially lower than the wages chosen by the Comptroller.
By ignoring the data from his own survey and instead blindly adopting Local 814’s rates, the Comptroller failed to comply with the statutory mandate to determine the wage “to be prevailing” (Labor Law § 230 [6]), meaning the actual prevailing wage (see Matter of Action Elec. Contrs. Co. v Goldin, 64 NY2d 213 [1984] [annulling Comptroller’s prevailing wage determination because it was based on arbitrary and irrational interpretation of statute]). The Comptroller’s exclusive reliance on a labor union agreement that does not reflect wages that are actually prevailing was arbitrary and capricious (see Matter of Louis v New York City Employees’ Retirement Sys., 26 Misc 3d 1236[A], 2010 Slip Op 50426[U] [2010] [agency decision set aside where agency “considered only those tests and reports that supported its denial and ignored those tests and reports that contradicted its position”]).
The Comptroller concedes that he did not base his prevailing wage schedule on the results from the survey. Nevertheless, he argues that in setting the schedule, he had the discretion to use the “30% rule” set forth in Labor Law § 220 (5) (a). That provision defines “prevailing rate of wage” as “the rate of wage paid in the locality ... by virtue of collective bargaining agreements between bona fide labor organizations and employers of the private sector . . . provided that said employers employ at least thirty per centum of workers, laborers or mechanics in the same trade or occupation in the locality where the work is being performed.” Thus, the Comptroller argues, since Local 814’s workers constituted 31% of the industry’s workers, he could rely exclusively on the union’s labor agreement to set the schedule.
The primary problem with the Comptroller’s position is the statutory provision containing the 30% rule (Labor Law § 220 [5] [a]) is found in article 8 of the Labor Law, which applies to laborers, workmen, and mechanics, but does not cover building *600service employees, such as the movers herein. As noted earlier, prevailing wage schedules for movers are governed by article 9 of the Labor Law, which has its own definition of “prevailing wage” that is different from the definition in article 8. Indeed, section 220 (5) (a) explicitly states that the 30% rule applies “for the intents and purposes of this article (i.e., article 8)” (emphasis added). Thus, in the face of survey results showing much lower wages than those in the collective bargaining agreement, it was improper for the Comptroller to graft onto article 9 the definition of “prevailing rate of wage” contained in article 8 (see Matter of Jones v Berman, 37 NY2d 42, 53 [1975] [“Administrative agencies can only promulgate rules to further the implementation of the law as it exists; they have no authority to create a rule out of harmony with the statute”]).
Finally, there is no merit to the Comptroller’s contention that Labor Law § 234 (1) (a) gives him the unbridled discretion to use the 30% rule. That section provides that in determining the “wages prevailing,” the Comptroller “may utilize wage and fringe benefit data from various sources including, but not limited to, data and determinations of federal, state or other governmental agencies.” However, merely because the Comptroller has the discretion to use various methods does not divest him of his statutory responsibility to determine the wage rate “to be prevailing” (Labor Law § 230 [6]). Where, as here, the union contract contains wage rates grossly disproportionate to the other data collected, the Comptroller cannot blindly use the 30% rule while ignoring the other data.
In light of this conclusion, we need not address petitioners’ alternative grounds for affirmance.
We have considered the Comptroller’s remaining arguments and find them unavailing. Concur — Andrias, J.E, Saxe, Acosta, Freedman and Richter, JJ. [Prior Case History: 32 Misc 3d. 175.]
. For the City of New York, the term “fiscal officer” means the Comptroller (Labor Law § 230 [8]).
. The prevailing wage rate has two components: the “hourly cash rate of pay,” and the “supplements,” which are fringe benefits, expressed as an hourly amount (see Labor Law § 230 [5]). These figures represent the sum of these two components.
. According to the Comptroller, the survey revealed that 31% of the workers belonged to Local 814.