dissents in a memorandum as follows: The issue raised on this appeal is whether a suspicionless stop of an occupied cab by officers in an unmarked police vehicle under the circumstances of this case was constitutionally permissible and valid.
*177Police Officer Ray Winslow, a member of the Street Crime Unit of the New York City Police Department, testified that he was working the 5:30 p.m. to 2:05 a.m. tour on February 27, 1995, and was assigned to the Taxi Livery Task Force in the 44th/46th Precinct. He explained that the Task Force is sent out to precincts with a high number of cab robberies, where its duties are to set up taxi check points, stop cabs, and hand out safety tips and other literature to cab drivers. A supervisor is supposed to be present at checkpoints and “usually” every third cab is stopped. The number, which “is different sometimes,” was possibly selected by the supervisor, but Winslow was “not exactly sure,” and he was unaware if there were criteria for the ratio selected.
On that date, Winslow was teamed with Sergeant Thomas Galati, who was the supervisor. Galati was not called by the People, so that the record does not contain further information about departmental policy, if any, concerning the selection criteria for cab stops. Both officers were in plainclothes; both sat in an unmarked police car secreted in the vicinity of 725 Exterior Avenue in the Bronx. At approximately 9:00 p.m., they pulled over the first cab of the evening, which was the third cab in sequence observed by the officers. This cab contained defendant as a passenger, leading to his arrest, as is related in the majority’s decision. On cross-examination, Winslow testified that he had not been given a specific location to locate the unmarked car, a choice that was basically up to the Sergeant and himself. The direction to stop every third cab was only verbally relayed; there was no written policy. Winslow conceded defense counsel’s observation that this was not a classic checkpoint, in that there was no marked police vehicle in a stationary position, at which location officers direct cars to pull over for a registration or DWI check. Here, the unmarked police car was parked on the side of the road at night, presumably out of sight of moving vehicles, awaiting passing cabs. Although Winslow indicated that they were stationary, they nevertheless operated by letting the targeted cab pass them first, after which the officers would follow, with turret light lit, and pull the cab over.
The stop in this case was pursuant to the New York City Police Department’s Taxi Livery Task Force program that has been criticized on Fourth Amendment grounds by other courts (see, e.g., United States v Santiago, 950 F Supp 590). In Matter of Muhammad F. (255 AD2d l68 [decided herewith]), we analyzed this program in the context of Federal and State constitutional law allowing stops, unsupported by reasonable suspicion, of *178cabs and other vehicles at fixed locations and “roving roadblocks”. The facts in Muhammad F. bore a striking similarity to those in Santiago, to the extent that apparently the same officer was involved, and the stop and arrest in Muhammad F. occurred three days after that in Santiago. Since the policy of mobile patrols stopping cabs in Muhammad F. did not derive from formal guidelines, but appeared to have been formulated at the street level, documentation of stops and results was not furnished, and the officers, rather than being identified, were in an unmarked car, we found insufficient indicia of a nonarbitrary and uniform policy to satisfy constitutional standards. Although the majority relies on the supposedly fixed situs of the police car in this case to authorize the stop as a checkpoint stop, under these facts, I see that as a distinction without a real difference. Notwithstanding my sympathy with the Department’s goals, I do not see how this stop can pass constitutional muster as previously articulated by this Court and the Court of Appeals.
Case law has developed an analytical model for judging the propriety of police stops of moving cars unsupported by reasonable suspicion and the policies or programs on which the stops are predicated. It is clear that a roadblock or checkpoint stop is as much of a seizure within the meaning of the Fourth Amendment as is a non-checkpoint stop of a moving car (United States v Hensley, 469 US 221, 226; Delaware v Prouse, 440 US 648; People v John BB., 56 NY2d 482, cert denied 459 US 1010). Although individualized suspicion is not a prerequisite to a constitutional seizure of a vehicle that is carried out according to a policy or program incorporating explicit, neutral limitations on the conduct of the individual officers undertaking the seizure (Delaware v Prouse, supra, at 663), nevertheless, the seizure must be justified in constitutional terms. This requires that a court evaluating the legality of the stop and seizure balance the consequential intrusion into the individual’s Fourth Amendment interests against the promotion of legitimate governmental interests (People v John BB., supra, at 487; People v Scott, 63 NY2d 518, 525; Delaware v Prouse, supra, at 657). The intrusion is characterized as both objective and subjective. The subjective intrusion would be measured in terms of the anxiety or inconvenience experienced by the individual being stopped; the objective intrusion is the stop itself (United States v Santiago, 950 F Supp 590, 595, supra). The analysis balancing these interests requires a preliminary evaluation of the nature and extent of the governmental interest involved, the effect of the program or policy in relation to it, and the degree of the intrusion on the individual, as measured *179by the subjective effect on that individual and the degree of discretion accorded the officers under the policy (People v Scott, supra, at 525).
There are two aspects of this police program that require analysis: the extent of, and justification for, the intrusion; and the degree of discretion vested in the officers responsible for carrying out the program.
The purported goal of the program was to hand out safety pamphlets to cab drivers. It is uncontroverted that the officers were not investigating a recent or ongoing crime or preventing the imminent commission of a robbery, despite some testimonial suggestion that the officers’ efforts were motivated by recent cab robberies in the vicinity. As such, the government interest in generally discouraging or investigating criminal activity, “does not implicate the same important social objectives that are at issue when police are investigating recent or ongoing suspected criminal activity” (People v Spencer, 84 NY2d 749, 754, cert denied 516 US 905). In “the absence of a crime ‘afoot’ ” (supra, at 756, citing United States v Ward, 488 F2d 162), the People bear the burden of demonstrating the “genuine need for so immediate and intrusive an action as pulling over [a] freely moving vehicle” (People v Spencer, supra, at 757), especially when less intrusive alternatives are available (supra, at 758). Parenthetically, to the extent that the police motive was informational, the same information might have been conveyed by other means, possibly by distributing safety pamphlets to dispatchers or the like for further distribution to cab drivers, but the record is silent on whether such methods were tried or even posited.
Turning to the nature of the stop, the “ ‘objective intrusion * * * checkpoint stops [are viewed] in a different light [from roving patrols] because the subjective intrusion — the generating of concern or even fright on the part of lawful travelers — is appreciably less in the case of a checkpoint stop’ ” (United States v Santiago, supra, at 595). The sine qua non of a valid checkpoint, typically for license and registration or sobriety checks, at which vehicles are stopped without regard to particularized suspicion, is in the very obviousness of the police presence and purposes and its stationary nature. For instance, in People v Scott (supra) the Court of Appeals found valid a roadblock, flagged by warning signs facing traffic in both directions some 300 feet from the checkpoint, manned by two marked police vehicles with flashing turret lights, and highlighted by flares placed in the center of the road, at which all vehicles were stopped. The Court of Appeals found that *180precautions as to safety, lighting and fair warning could alleviate drivers’ subjective reactions caused by sudden and apparently purposeless intrusions by patrolling police. The fact that the checkpoint was moved on a regular basis to other, preselected, locations did not diminish the stationary nature of the roadblocks (supra). We found a roadblock to be valid where several officers, responding to a spate of neighborhood car thefts, were assigned to conduct systematic mandatory car stops on a specified corner and directed to pull over every vehicle coming down the block to check ownership papers (People v Serrano, 233 AD2d 170, lv denied 89 NY2d 929). Although generally invalidating “roving patrols” which stop vehicles without reasonable suspicion, the Court of Appeals has allowed a “roving roadblock” under narrow circumstances not present here— when police uniformly stopped all vehicles in a sparsely populated rural area in connection with recent robberies (People v John BB., 56 NY2d 482, 489, supra).
In the present case, the People characterize the “checkpoint” as being stationary. While the record does not indicate that this was exactly the type of roving patrol that we invalidated in Muhammad F. (supra), and for which suppression was required in United States v Santiago (supra), it was nevertheless more akin to a mobile patrol than to a stationary checkpoint. It bears repeating that the police vehicle was unmarked and secreted in the night and, rather than stopping vehicles at the actual checkpoint, the officers, in plainclothes, were positioned so as to require a pursuit, even if a quick one, as the cab blithely passed beyond the supposedly stationary checkpoint.
Nor does the nature of the “policy” in this case instill confidence in its uniformity, in the indiscriminate nature of the selection criteria, and in its restriction of subjective motivations by the very officers carrying out the “policy.” In short, there was no policy, in the sense that identified criteria were not promulgated by the central command. Rather, at best, this record suggests that the very officers carrying out the policy were the ones who were formulating it. Although deciding on different grounds, we have noted elsewhere that the fact that the Police Department allows officers to exercise discretion to stop cabs ostensibly for the safety of the drivers in the absence of reasonable suspicion of actual criminal activity does not mean that the stop thereby is justified (People v Concepcion, 216 AD2d 141, lv denied 86 NY2d 792). The present record suggests just such street-level discretion, or, at best, as characterized by the Second Department, only “an unelabo*181rated precinct directive” (People v Vails, 170 AD2d 550, 551) as the predicate for the stop. By contrast, in People v Scott (supra), the County Sheriffs office had established a written policy, incorporating objective criteria (every car was stopped) and selecting the actual checkpoints as well as the time periods of operation and of rotation among the checkpoint sites. The officers in Scott conducting the stops had virtually no discretion as to whom to stop. This is not to suggest that all discretion ought to be removed from the officer at the scene, or that a policy, to be valid, must be in writing (a requirement that we declined to adopt in Serrano) but those circumstances have been identified by the Court of Appeals as factors enhancing the nonarbitrariness required by constitutional standards. It bears repeating that the supervisory officer who was on the scene might have clarified the parameters and formality of the policy, but he was not called as a witness by the People.
The People contend that the selection of an objective number (stopping one in three cabs) somehow makes the selection sufficiently neutral to insulate each stop from the officer’s unfettered discretion. The Southern District, examining the very argument advanced in the present case, characterized this logic as “plainly wrong” (United States v Santiago, supra, at 596). In that case, as in this case, there were no written rules, either the individual officers or unidentified persons in the precinct decided where and when to conduct stops, there was no reporting regarding the stops, there was no oversight as to whether stops were truly conducted in a neutral manner, and the patrols seemed to have been conducted disproportionately at night. All of this undermined the ostensible informational motive of the program. As previously noted, the informational ends in Santiago, as here, likely could have been accomplished with equal or even greater ease and efficiency by distributing pamphlets and making police presentations at central locations, such as taxi dispatch locations, taxi inspection stations, or through the Taxi and Limousine Commission. This is not to suggest that any of these various points should be necessary requirements in promulgating this type of program, but in the aggregate, the silence of the present record undermines the very clarity and neutrality that case law has required for vehicle stops unsupported by reasonable suspicion.
As I noted at the outset, since I cannot see how these facts can be rationally distinguished from those in Muhammad F. (supra), in which this Court requires suppression, or how a reversal in this case is in line with Court of Appeals authority, I conclude that the motion court properly construed governing *182case law and properly suppressed. For that reason, I respectfully dissent.