OPINION AND ORDER

To support their claims in this class action lawsuit, plaintiffs intend to rely on the testimony of Jeffrey Fagan, a criminologist with expertise in statistical analysis.1 To rebut Fagan’s testimony, defendants seek to offer the opinions of Dennis Smith, a political scientist with an expertise in police organizations. Plaintiffs now move to preclude parts of Smith’s opinions. Their motion is granted in part and denied in part.

I. BACKGROUND

Details of this litigation have been extensively covered in my decisions on defendants’ motion for summary judgment,2 defendants’ Daubert motion challenging Fagan’s report,3 and plaintiffs’ motion for class certification.4 Familiarity with these opinions is assumed. I include only new background here.

Dennis Smith is an associate professor of public policy at the Robert F. Wagner School of Public Service at New York University.5 He is an expert on police organizations and police behavior and has been studying the New York City Police Department (“NYPD”) since the 1970s.6 His scholarship (some of which has been pub*510lished in books and peer-reviewed journals and much of which has not) has focused on “performance management” and on evaluating the efficacy of public services, with a particular emphasis on police performance.7 He is not a statistician and his only formal study of statistics took place over thirty-five years ago in graduate school, when he received his Ph.D. in political science.8 However, according to Smith, his policy research has involved extensive collaboration with experts from other fields, including statistical experts. “Statistical analyses ... have consistently been critical components of [his] studies and publications.”9 Smith explains that he has “extensive experience critiquing the theories and assumptions underlying statistical models” and that he is “able to understand and develop hypotheses regarding specific multiple regression models.”10

Since 2006, Smith has collaborated with Robert Purtell on much of his work regarding the NYPD. Purtell is an assistant professor of finance and the director of the Masters in Public Administration Program at the University at Albany, Nelson A. Rockefeller College of Public Affairs & Policy.11 According to Smith, “[i]n all of my prior collaborative work with Purtell [on policing in New York City], the models tested by his statistical analysis were developed primarily by me based on my knowledge of policing — an area in which Purtell himself lacks expertise and relies upon my guidance.”12 Purtell uses his expertise in statistics and modeling to construct and implement regressions that will test Smith’s theories.13 Although Purtell was initially identified as a testifying expert by defendants, he did not produce an expert report.

Plaintiffs’ motion seeks to preclude Smith from testifying as follows:

1. Smith may not critique Fagan’s multivariate regression analyses and [Fagan’s] critique of the RAND study [that was commissioned by defendants];

2. Smith may not offer his correlation coefficient calculations and “alternative” regression analysis;

3. Smith may not opine on the meaning of low stop-and-frisk weapons recovery hit rates;

4. Smith may not opine on crime reduction in New York City, or otherwise testify about the results of the studies attached as Appendices D and E to his Expert report; and

5. Smith may not opine that NYPD officers do not conduct stops-and-frisks on the basis of race.14

11. LEGAL STANDARD

The proponent of expert evidence bears the initial burden of establishing admissibility by a “preponderance of proof.” 15 Rule 702 of the Federal Rules of *511Evidence states the following requirements for the admission of expert testimony:

If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise, if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of rehable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case.

Under Rule 702 and Daubert, the district court must determine whether the proposed expert testimony “both rests on a rehable foundation and is relevant to the task at hand.”16 The district court must act as “ ‘a gatekeeper to exclude invalid and unreliable expert testimony.’ ”17 However, “the Federal Rules of Evidence favor the admissibility of expert testimony, and [the court’s] role as gatekeeper is not intended to serve as a replacement for the adversary system.”18 In serving its gatekeeping function, the court’s focus must be on the principles and methodologies underlying the expert’s conclusions, rather than on the conclusions themselves.19 In assessing an expert’s methodology, courts may consider (1) “whether [the method or theory] can be (and has been) tested,” (2) “whether [it] has been subjected to peer review and publication,” (3) “the known or potential rate of error [associated with the technique] and the existence and maintenance of standards controlling the technique’s operation,” and (4) whether the method has achieved “general acceptance” with the relevant community.20

The courts’ gatekeeping function applies not only to “scientific” evidence, but also to proffers of “technical, or other specialized knowledge” under Rule 702.21 The objective of this function is to “make certain that an expert, whether basing testimony upon professional studies or personal experience, employs in the courtroom the same level of intellectual rigor that characterizes the practice of an expert in the relevant field.”22 However, recognizing that “there are many different kinds of experts, and many different kinds of expertise,” the Supreme Court has emphasized that the reliability inquiry “is a flexible one.”23 Accordingly, the factors “identified in Daubert may or may not be pertinent in assessing reliability, depending on the nature of the issue, the expert’s particular expertise, and the subject of his testimony.”24 Ultimately, the inquiry “depends upon the particular circumstances of the particular case at issue.”25 In sum, the trial court has “the same kind of latitude in deciding how to test an expert’s *512reliability ... as it enjoys when it decides whether or not that expert’s relevant testimony is reliable.”26

In addition, Rule 403 of the Federal Rules of Evidence states that relevant evidence “may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury.” “Expert evidence can be both powerful and quite misleading because of the difficulty in evaluating it. Because of this risk, the judge in weighing possible prejudice against probative force under Rule 403 ... exercises more control over experts than over lay witnesses.”27 Generally, “the rejection of expert testimony is the exception rather than the rule.”28 “The admission of expert testimony is committed to the broad discretion of the District Court and will not be disturbed on review unless found to be ‘manifestly erroneous.’ ”29

III. DISCUSSION

Plaintiffs seek to exclude five aspects of Smith’s report. I assess each one in the following discussion.

A. Smith’s Critique of Fagan’s Multivariate Regression Analyses

Fagan conducted various regression analyses in order to evaluate plaintiffs’ allegation that defendants are conducting stops and frisks on the basis of race, in violation of the Fourteenth Amendment.30 Among many other findings, Fagan’s regression analyses show that “NYPD stops- and-frisks are significantly more frequent for Black and Hispanic residents than they are for White residents, even after adjusting for local crime rates, racial composition of the local population, police patrol strength, and other social and economic factors predictive of police enforcement activity.”31 Notably, however, Fagan’s regression analyses do not account for the racial distribution of the perpetrators of crime in New York City.

In his report, Smith argues that this is a major flaw in Fagan’s analysis.32 Smith believes that by comparing the racial distribution of stops to the racial distribution of the population, instead of to the racial distribution of criminal suspects, Fagan is using the wrong “benchmark.”33 As an analogy, Smith notes that the NYPD stops women, children, and senior citizens at rates far below their share of the population.34 These groups of people are stopped at low rates, he argues, because they commit a disproportionately small number of crimes. Like the disparity in stops on the basis of gender and age, Smith argues, the disparity in stops on the basis of race and ethnicity is driven by the disparity in the commission of crime, not by unconstitutional discrimination. This dispute over “benchmarking” constitutes the parties’ central disagreement regarding Fagan’s analysis of discrimination.35

*513Plaintiffs argue that Smith is not qualified to make this benchmarking critique because he “has never conducted a statistical study to assess the racial bias of stop- and-frisk or any other law enforcement program,, practice, strategy or tactic, or claims of racial discrimination in any other governmental or private institution” and that, prior to his work on this case, Smith had done very little research into the question of racially-biased police stops.36

Plaintiffs’ point is better suited for cross-examination and closing argument than a Daubert motion. As an expert in police practices, Smith has the necessary expertise to evaluate Fagan’s hypotheses and the assumptions that Fagan used to build his model and he has the expertise to make his central “benchmark” criticism. Fagan explains that he chose not to use the crime suspect data in his benchmark because the data is incomplete; “race” is only known for approximately sixty percent of crimes and extrapolating that information to the remaining forty percent of crimes might not be appropriate.37 Whether such an extrapolation would be sound is as'much a question of social science as technical statistics and the answer turns largely on the investigator’s assumptions about the characteristics of the relevant populations and about the accuracy of alternative metrics (such as the regression analysis used by Fagan). Smith is therefore qualified to critique Fagan’s decision on this issue.38

B. Smith’s Alternative Regression Analysis and Correlation Coefficient Calculations

In order to highlight what he believed was the central flaw with Fagan’s benchmark decision, Smith devised (and Purtell constructed) an alternative regression analysis based on arrest and crime complaint reports.39 Plaintiffs correctly point out that because he is not an expert in statistics, Smith is unable to “establish that both the choice of variables and the actual running of the model were methodologically sound.”40 Without these verifications, Smith’s testimony is inadmissible.

However, given the importance of presenting otherwise admissible information to the jury in this case and given the lack of prejudice .to plaintiffs, the appropriate solution is to permit Purtell to testify about the. technical aspects of the alternative regression analysis and the correlation coefficient calculations. If he agrees with it, Purtell may therefore “adopt” Smith’s expert report, be deposed by plaintiffs, and testify at trial regarding technical aspects of Smith’s report. This late alteration to the expert schedule is appropriate because discovery and motion practice has *514taken many years, one additional deposition will not delay the trial date, and in a case of such significance involving complicated social science data it is important that the jury be permitted to fully consider both parties’ admissible arguments about what the data shows.

Plaintiffs argue that Purtell, like Smith, is not qualified to offer critiques of Fagan’s regressions because he has never conducted racial disparity studies or researched appropriate benchmarking techniques. But this argument goes to weight, not admissibility; even if Purtell has not previously conducted regression analyses on these particular topics, he has conducted many regressions on other topics and he is qualified to point out what he perceives as errors in Fagan’s methodology and explain the technical aspects of Smith’s alternative regression model.

After Purtell establishes the methodological soundness of the alternative regression analysis, Smith may testify regarding the qualitative theories and assumptions that went into its construction and to -the conclusions that he believes it supports. Although the regression was disclosed belatedly, fairness requires its admission.41

C. Smith’s Crime Reduction Studies and Opinions

Smith devotes a significant portion of his report to discussing the central role that he believes the NYPD’s stop and frisk program (and the related Operation Impact) has had in “the historic crime decline achieved by New York City.”42 One of Smith’s central critiques is that “[t]he Fagan analysis does not explicitly confront the historic shift at NYPD away from a primary mission of responding to crime to a mission of preventing crime through proactive and crime targeted police vigilance.” 43 Included as appendices to Smith’s report are two studies that he and Purtell wrote about reduced crime rates resulting from Operation Impact and the NYPD’s stop and frisk program.

Plaintiffs argue that “Smith’s crime reduction opinions” should be excluded because they “are irrelevant to the questions posed by Plaintiffs’ Fourth and Fourteenth Amendment claims: (1) Do NYPD officers conduct stops-and-frisks without reasonable suspicion?; (2) Do they stop civilians on the basis of their race?”44 Defendants respond by arguing that Smith’s opinion on the crime deterrent effects of these programs “are indeed relevant, as they represent alternative, race-neutral explanations for the racial patterns in [stops and frisks] which Fagan failed to consider in his analysis of the data” and that “[excluding Smiths opinions would be highly prejudicial by forcing the jury to accept Fagan’s word unchallenged ... when such strong evidence of methodological problems exists.”45

Defendants are conflating two different aspects of Smith’s report: his benchmarking critique and his separate conclusion that the NYPD’s programs are a proven strategy to combat crime and *515increase safety, particularly in minority neighborhoods. As I explained above, Smith’s benchmarking critique challenges Fagan’s finding that Blacks and Hispanics are stopped at disproportionately higher rates; it is a descriptive claim about the nature of racial disparities that is probative of the truth or falsity of plaintiffs’ Fourteenth Amendment claim, and it is therefore admissible.

However, Smith’s opinions about the deterrence and crime reduction impacts of the NYPD’s programs are inadmissible. Defendants argue that “Smith’s opinion that increased [stop and frisk] activity reduces neighborhood crime provides further evidence for his alternative hypothesis that [stops and frisks] are driven by where the ■ crime occurs rather than by racial discrimination.”46 But there are only two ways in which the effectiveness of the stop and frisk program can be probative of its legality. First, defendants could argue that they have adopted a program that discriminates on the basis of race but nevertheless survives strict scrutiny because it is narrowly tailored to achieve a compelling governmental objective.47 But defendants explicitly disclaim any such defense.48 The only other way in which “Smith’s opinion that increased [stop and frisk] activity reduces neighborhood crime” could be probative of the activity’s legality is by bolstering the defense that NYPD policymakers and officers have the intent only of reducing crime and are not deciding whom to stop on the basis of race.49 But Smith is not qualified to provide this testimony. Both in their court filings and in public comments, defendants regularly state that the intent behind the program is to reduce crime, not to discriminate. Defendants and their representatives may testify about why the City has adopted and pursued the stop and frisk program; on the basis of that testimony and the testimony of plaintiffs’ witnesses, the jury will determine whether defendants are' discriminating' on the basis of race, either when making individual stops or as a matter of policy or- custom. Defendáiíts clearly believe that the program has been effective in reducing crime. Whether or not Smith agrees with defendants’ assessment is irrelevant. Moreover, admitting Smith’s testimony — and permitting the parties to delve into the question of whether the-stop and frisk program actually reduces crime — would risk turning the trial into a policy debate over the wisdom of the program rather than a judicial proceeding that assesses plaintiffs’ constitutional claims.

D. Smith’s Hypotheses on Hit Rates

According to Fagan, the NYPD’s data shows that approximately five percent of stops result in an arrest, six percent of stops result in a summons, guns are seized in 0.15 percent of stops, and contraband of any kind is seized in 1.75 percent of stops.50 I found these statistics powerful *516evidence of a widespread pattern of unlawful stops: In “eighty-eight percent of cases [where no arrest was made or summons given], although they were required by law to have objective reasonable suspicion that crime was afoot when they made the stop, police officers ultimately concluded that there was no probable cause to believe that crime' was afoot. That is to say, according to their own records and judgment, officers’ “suspicion” was wrong nearly nine times out of ten.”51

In rebuttal, defendants seek to admit Smith’s opinion that “the test of success in a proactive, prevention-focused program is not the same as in an assessment of a reactive program” and that if “the goal of NYPD is to pursue practices that convince would be gun carriers to leave their guns at home, why would the fact that over time fewer guns are found in suspicion-based stops be a sign of failure?” 52 Defendants argue that Smith should be permitted to opine that the low rate of gun recoveries “may be explained in part by the deterrent effect of increased [stops and frisks],” because he is simply offering a “plausible alternative explanation” that will “challenge the reliability of Fagan’s conclusion” about the significance of this statistical pattern.53

Defendants are wrong. Fagan is proposing hypotheses about the objective circumstances that immediately precede stops on the' basis of those stops’ outcomes: if police officers only arrest or give summonses to twelve percent of the people they stop, then perhaps their initial “suspicion” that crime was afoot was in fact not reasonable; if guns are rarely recovered, then perhaps the “suspicious bulges” that police identify are not really suspicious. Smith, on the other hand, offers a hypothesis about the class members’ fears and motives: perhaps New York City residents are not carrying guns, he says, because they know that they will be stopped and frisked. Smith cites no evidence for this hypothesis. This theory about the reasons that so few stopped people are carrying weapons is too speculative to be admitted at trial, particularly through the testimony of an expert.

I also note that Smith’s hypothesis is not even couched as a rebuttal of Fagan’s hypothesis that the low hit rates suggest a lack of reasonable suspicion; instead, Smith appears simply to attempt to justify stops on the basis of their deterrent impact, regardless of their legality.54 Thus, the opinion is also irrelevant, and it is surely not based on any scientific methodology.

E. Smith’s Opinion that NYPD Stops-and-Frisks Are Not Racially Motivated

Plaintiffs seek to exclude Smith’s opinion that “there is no compelling evidence *517that NYPD officers are making stops based on race or ethnicity, but instead are pursuing a strategy and using tactics that prevent crime and benefit the City as a whole, and communities of color in particular.” 55 As stated above, Smith will. be permitted to testify that, because of the benchmarking dispute and other more minor critiques, he does not believe Fagan’s report presents compelling evidence that NYPD officers are making stops based on race or ethnicity.

But, as plaintiffs argue, “Smith did not conduct a study through which he determined that crime deterrence, rather than race, was a statistically significant and robust predictor of stop and frisks, nor did he conduct a statistical analysis that identified crime deterrence as the motivating factor in NYPD stops and frisks.”56 Smith will not be permitted to testify, regarding his opinion that the NYPD’s policies “prevent crime and benefit the City as a whole, and communities of color in particular.” As explained above, although his crime reduction opinions may be relevant as a policy matter, they will not help the jury answer the constitutional questions posed by this lawsuit. Neither expert will be permitted to speculate about the intent of NYPD policymakers or the efficacy of those policies, although defendants and their representatives may testify as to their intent.

IY. CONCLUSION

Plaintiffs’ motion is granted in part and denied in part. The Clerk of the Court is directed, to close the motion [Docket No. 215], A status conference is scheduled for August 27 at 3:30 p.m. Defendants are directed to make Robert Purtell available for deposition.as soon as practicable.

SO ORDERED.