*52OPINION OF THE COURT
Graffeo, J.In this CPLR article 78 proceeding, a contractor challenges the methodology used by the Department of Labor to classify work for purposes of determining the appropriate prevailing wage rate to pay workers on public projects, arguing that the Department erred because it did not consider survey evidence of local contractor practices in making a trade classification. We hold that the Department need not conduct such surveys in order to classify work and conclude that the Commissioner’s determination was neither arbitrary nor capricious.
Petitioner Marke Lantry, doing business as Thorough Job Architectural Construction Systems, subcontracted to install preglazed windows for the Ichabod Crane Central School District in Columbia County in 1997. Lantry paid his employees (who were not represented by a trade union) the glaziers’ wage rate for their labor on the project. The Department’s Bureau of Public Work conducted a random audit of the project and concluded that Lantry had underpaid some of his employees. With respect to the workers engaged in window installation, the Bureau asserted that they should have beén compensated at the higher rate of pay for ironworkers.1 The Bureau assessed Lantry with an underpayment of approximately $5,400 and directed the school district to withhold certain payments from him.
Lantry responded by requesting a hearing to contest the Bureau’s findings. Presuming that the Department would use a “prevailing practice in the locality” approach, which relies on surveys of local contractors to determine the trade classification of workers for wage rate purposes, Lantry sought to establish that the prevalent practice in the Capital District area was for contractors to pay their employees installing preglazed windows at the glaziers’ rate, rather than the ironworkers’ rate.
At the hearing, the Department’s principal witness was the senior public work wage investigator, who had audited Lantry’s project. He testified that in classifying a particular task into the appropriate trade or occupation, the Department generally analyzes the specific nature of the work, together with collective bargaining agreements, jurisdictional agreements and past Bureau recognition. In his 16 years as a Department investiga*53tor, he indicated that he never examined the “prevailing practice” in a locale or elicited survey-type information from contractors. The witness explained that he classified the work done by Lantry’s employees as ironwork after considering several factors, including the nature of the task and an agreement between ironworkers and carpenters indicating that ironworkers retained jurisdiction over the installation of metal window frames into masonry.2 The Department also presented testimony that revealed that the glaziers’ union had a minimal presence in the Capital District area in the mid-to-late 1990s.
In contesting the Bureau’s determination that the window installation work should have been classified as ironwork, Lantry submitted the testimony of various union and nonunion glazier contractors who claimed to have consistently treated the installation of preglazed windows as glaziers’ work and had paid the glaziers’ rate for such labor.3 A market research consultant who had been hired by the New York State Glass Association to survey the scope of work undertaken by glazier contractors also testified. Based on statistical evidence developed from this survey, he stated that most of these contractors used glaziers to install preglazed windows, although some employed ironworkers or carpenters for such tasks.
After the administrative hearing, the hearing officer issued a report and recommendation accepting the Bureau’s ironwork classification and finding of underpayment.4 Recognizing that the Department could rely on collective bargaining agreements in making classification determinations, the hearing officer noted that in this case the relevant agreements for glaziers and *54ironworkers both claimed jurisdiction over preglazed window installation. He therefore needed to examine other indices that the Department historically relied on to classify work, such as the nature of the work and pertinent jurisdictional agreements. The hearing officer accorded Lantry’s survey evidence little weight, believing that the classification of work should not be dictated by statistics.
The Commissioner subsequently issued a determination and order adopting the hearing officer’s findings of fact, conclusions of law and recommendations in their entirety. Lantry commenced this article 78 proceeding seeking to vacate the Commissioner’s order. The Appellate Division confirmed the determination. We granted Lantry leave to appeal and now affirm.
Labor Law § 220 implements the constitutional mandate that contractors engaged in public projects pay their workers wages and supplements which “shall be not less than the prevailing rate for a day’s work in the same trade or occupation in the locality within the state where such public work ... is performed” (Labor Law § 220 [3]; see also NY Const, art I, § 17). The Commissioner is responsible for determining the prevailing wage rates for the various trades in each locality (see Labor Law § 220 [3]).5 Calculation of the prevailing rate entails a two-step process. First, the Commissioner must classify the work by assigning the task performed by an employee to a specific trade or occupation (see Labor Law § 220 [3-a] [a]). Second, the Commissioner must ascertain the prevailing rate for that trade or occupation in the relevant locality (see Labor Law § 220 [3], [5] [a]).
In the early 1980s, the Commissioner fixed prevailing wage rates by performing industry surveys of the actual wages received by trade workers in each of the state’s localities (see Matter of Liquid Asphalt Distribs. Assn. v Roberts, 116 AD2d 295, 296 [3d Dept 1986]; see also Labor Law former § 220 [5]). This process, which required the Department to survey over 1,500 localities, was costly and burdensome and the Department’s experience demonstrated that the prevailing wage rates that were adopted would invariably equate to collectively bargained wages. Thus, the Legislature amended the Labor Law in 1983 to authorize the Commissioner to dispense with the survey process and instead adopt the rate paid in a locality by *55referring to collective bargaining agreements between labor organizations and private sector employers, provided that such agreements covered at least 30% of the workers in that trade or occupation (see Labor Law § 220 [5] [a], [c]; see also L 1983, ch 447).
Nothing in section 220 of the Labor Law mandates that the Commissioner consider actual contractor practices in a locality when classifying work. To the contrary, while the statute provides a specific process for calculating the “prevailing rate of wage” (see Labor Law § 220 [5] [a]), it does not require a specific procedure for the Commissioner to use in evaluating the appropriate trade or occupation to assign to particular work.
Lantry argues that the Commissioner’s classification of the work involved in installing preglazed windows into masonry as ironwork rather than glazier work was arbitrary and capricious. Specifically, Lantry contends that a “prevailing practice in the locality” test, based on industry survey information and contractor testimony regarding what type of workers are actually performing the task in dispute, should be the standard the Commissioner follows to classify work. Hence, Lantry asserts that the Commissioner erred in failing to credit the survey data and the testimony of the contractors that he claims confirmed that glaziers do more preglazed window installation jobs than ironworkers. We disagree.
In the absence of a statutory directive, this Court will uphold the Commissioner’s methodology as long as it is not unreasonable (see Matter of Chesterfield Assoc. v New York State Dept. of Labor, 4 NY3d 597, 604 [2005]). Furthermore, it is well settled that trade classifications “are a matter given to the expertise of the Department and courts are strongly disinclined to disturb them, absent a clear showing that a classification does not reflect ‘the nature of the work actually performed’ ” (Matter of General Elec. Co. v New York State Dept. of Labor, 154 AD2d 117, 120 [3d Dept 1990], affd for reasons stated below 76 NY2d 946 [1990] [citations omitted], quoting Matter of Kelly v Beame, 15 NY2d 103, 109 [1965]).
Here, the Commissioner outlined the analysis that the Department generally follows in classifying work. Although a number of factors may be relevant, where a union has obtained jurisdiction over a certain category of work through the collective bargaining process, the Department will typically adopt the classification assigned in the collective bargaining agreement. *56But where a dispute arises between two unions that each claim jurisdiction over the same type of work, the Department necessarily expands its review to encompass other factors, such as jurisdictional agreements, past precedent, and the inherent nature and characteristics of the work in question. In neither scenario, the Commissioner observed, would the Department rely on “a survey or comparison of sheer numbers to determine the appropriate classification of work.”6
Courts have consistently held that the Commissioner may rely on collective bargaining agreements in making trade classifications (see e.g. Matter of Otis E. Serv. v Hudacs, 185 AD2d 483, 485 [3d Dept 1992]; Matter of Sierra Telcom Servs. v Hartnett, 174 AD2d 279, 283 [3d Dept 1992], appeal dismissed 79 NY2d 1039 [1992], lv denied 80 NY2d 757 [1992], cert denied 507 US 972 [1993]; Matter of Naftilos Painting & Sandblasting v Hartnett, 173 AD2d 964, 966 [3d Dept 1991]). The Commissioner’s use of collective bargaining agreements to classify work is consistent with the 1983 amendments to section 220 of the Labor Law, which authorized the Department to rely on collective bargaining agreements rather than survey data to determine prevailing wage rates and supplements. Although those amendments did not specifically address the Commissioner’s classification prerogative, they evince a recognition by the Legislature of the administrative burden of compiling and updating surveys and therefore support the Commissioner’s conclusion that the survey approach urged by Lantry is impractical.
Likewise, we do not find the Commissioner’s consideration of other factors, including the nature and characteristics of the work at issue—to the exclusion of survey evidence of contractor practices—to be irrational. Although underpayments sometimes occur, necessitating reimbursement of workers after the fact, the Department’s mission is to ensure that workers are paid the correct wage in the first instance. Thus, the Department and contractors need to know the trade classification for an item of work before commencement of a public project so that the appropriate prevailing rate of wages and supplements can be *57ascertained and workers compensated accordingly. It would be difficult for the Commissioner to classify work in a timely manner if she had to depend on survey data of local practices with respect to each of the more than 100 trades in the construction industry.7 Additionally, the Commissioner’s emphasis on examining the nature of the work for trade or occupation classification is reasonable given this Court’s pronouncement that the nature of the work performed is “[t]he pivotal question” in reviewing the Department’s trade classifications (Matter of Kelly v Beame, 15 NY2d 103, 109 [1965]). In sum, contrary to Lantry’s position, the Department need not compile survey evidence to determine the appropriate trade classification for a particular type of work.8
In this case, both the applicable glaziers’ and ironworkers’ collective bargaining agreements claimed jurisdiction over the installation of preglazed windows. The Commissioner therefore looked to other factors to classify the task. Primarily, the Commissioner relied on evidence of the process of installing preglazed windows into masonry openings, which involves the attachment of aluminum clips to each window with fastener screws, noting that no glazing was necessary because the window manufacturer already performed that function. The Commissioner also considered the jurisdictional agreement be*58tween carpenters and ironworkers, under which the ironworkers claimed jurisdiction over the installation of metal window frames into masonry, as well as evidence that the local glaziers’ union had little presence in the region in the 1990s. The Commissioner had the discretion to weigh this evidence more heavily than that proffered by Lantry, which consisted largely of testimony and survey data indicating that glazier contractors considered preglazed window installation to be glaziers’ work.9 According the Commissioner deference based on her expertise in this area, we cannot say that her classification of the task at issue as ironwork was arbitrary or capricious or inconsistent with the nature of the work.
Nevertheless, Lantry urges that the Commissioner’s determination cannot be sustained because a prior determination of the Department directly conflicts with the result reached in this case. Lantry cites Matter of Cortland Glass Co. (PRC Nos. 94-8074A, 95-0790 [Jan. 10, 2000]), in which an Erie County contractor was charged with underpaying its employees at the glaziers’ rate rather than the ironworkers’ rate where they had installed preglazed windows on a public work project. After a hearing, the Department held that no underpayment occurred, concluding that the contractor properly used the glaziers’ rate. The Department relied heavily on evidence indicating that the actual practice undertaken by contractors in the region was to classify such work as glaziers’ work. Lantry submits that the Cortland Glass determination should inform the result in his case.
It is well settled that “[a] decision of an administrative agency which neither adheres to its own prior precedent nor indicates its reason for reaching a different result on essentially the same facts is arbitrary and capricious” (Matter of Charles A. Field Delivery Serv. [Roberts], 66 NY2d 516, 517 [1985]). Here, however, the Commissioner acknowledged that the Cortland Glass decision was inconsistent with the current determination *59and set forth her reasons for declining to follow its rationale and conclusion. The Commissioner explained that actual practice evidence is generally not relevant to the process of classifying work, overruling Cortland Glass, to the extent it held otherwise, as flawed and a departure from longstanding departmental precedent.10 We conclude that the Commissioner satisfied her obligation under Field Delivery.
Lantry’s remaining contentions are without merit.
Accordingly, the judgment of the Appellate Division should be affirmed, with costs.
. The Bureau also determined that other work compensated at the glaziers’ rate should have been paid at the rate for masons, laborers or carpenters. Lantry did not challenge these findings.
. The record indicates that, before preglazed windows were developed, an ironworker would first attach a metal frame to the building’s masonry walls with metal clips and screws, after which a glazier would affix the glass to the frame. Preglazed windows have the window glass installed at the factory thereby eliminating glass installation at the job site. As Lantry testified at the hearing, installation of a preglazed window now involves the attachment of the metal frame containing the window glass to the building’s masonry.
. The nonunion contractors testified that they paid the glaziers’ rate when working on public projects. They typically paid lower wages when performing similar work in the private sector.
. The hearing officer also concluded that Lantry’s failure to pay the prevailing wage and supplements was willful based on his failure to pay carpenters’ rates for ceiling tile work and laborers’ rates for washing and cleaning tasks. As a result of this finding of willful conduct, the hearing officer recommended that the Commissioner impose a civil penalty in the amount of 5% of the underpayment. The total amount withheld was $15,599.68, representing wages, supplements, interest and penalties.
. The Comptroller of the City of New York has the same responsibility within the city (see Labor Law § 220 [3], [5] [e]).
. The Commissioner does suggest in her brief, however, that empirical evidence may be relevant where “[historical data that correlate with changes in the work process [establish] that the nature of the work (and therefore the trade classification) has evolved over time.” The Commissioner also indicates that numerical data could conceivably be relevant to show that a particular trade does not perform any of the work in question.
. The dissent indicates that the issue in this case is whether the Commissioner must rely on prevailing practice to classify work, not whether the Department has to conduct surveys. But if prevailing practice is the standard, survey evidence seems unavoidable. In any given case, a contractor could selectively bring in a cadre of witnesses who would testify favorably for the contractor. In fact, Lantry testified at the hearing in Matter of Cortland Glass Co. (PRC Nos. 94-8074A, 95-0790 [Jan. 10, 2000]), a case involving a similar issue, and the contractor charged in Cortland Glass testified at Lantry’s hearing in this case. Accurate determination of prevailing practice—particularly where two unions claim jurisdiction over the same task—would therefore necessitate extensive surveying, a burden the Commissioner reasonably seeks to avoid.
. Lantry’s approach would also be unworkable because the Commissioner must classify the work in order to determine the pertinent locality. This is because the Labor Law now defines “[l]ocality” as “such areas of the state described and defined for a trade or occupation in the current collective bargaining agreements between bona fide labor organizations and employers of the private sector, performing public and private work” (Labor Law § 220 [5] [d]). Because a trade classification determines the relevant collective bargaining agreement governing a wage rate, the locality cannot be ascertained until after the Commissioner classifies the work. To do as Lantry suggests would require the Department to survey a geographical area before it knows the locality—the territorial scope of the relevant collective bargaining agreement.
. The dissent states that Lantry presented survey data demonstrating that glaziers performed 94% of preglazed window installation in eastern New York. But that survey collected information exclusively from glazing companies and indicated that 94% of glazier contractors considered such work to be glaziers’ work. The survey did not contemplate contractors who considered themselves to be ironwork firms or who were signatories to the ironworkers’ collective bargaining agreement. If, hypothetically, a survey of such companies similarly indicated that a vast majority considered the identical task to be ironwork, under a prevailing practice test, the inquiry would devolve into an analysis of which contractors actually installed more windows.
. During the pendency of the appeal in Cortland Glass, the Department in its brief to the Appellate Division repudiated the posture the agency had taken in the case and sought a remand for reconsideration (see Matter of International Assn. of Bridge, Structural & Ornamental Iron Workers, Local Union No. 6, AFL-CIO v State of New York, 280 AD2d 713 [3d Dept 2001] [dismissing for lack of standing a petition brought by ironworker unions that sought review of Cortland Glass]).