dissenting: I respectfully dissent. This is the first case in which this court has relied expressly and exclusively on part I, article 19 of the State Constitution in considering the constitutionality of a police roadblock. Prior roadblock cases either did not distinguish between State and national constitutional standards, see State v. Severance, 108 N.H. 404, 237 A.2d 683 (1968), or rested on the fourth amendment, see State v. Baldwin, 124 N.H. 770, 475 A.2d 522 (1984). Today’s decision applies a standard that is more restrictive than the limitations on police activity recognized in either of those cases. I dissent from the adoption of today’s standard and from the result to which it leads the majority.
There is no question that stopping a car even for a brief period is of constitutional significance, for it is a limited “seizure” of the car and its occupants. State v. Landry, 116 N.H. 288, 358 A.2d 661 (1976); Delaware v. Prouse, 440 U.S. 648 (1979). The ultimate touchstone for determining the constitutionality of such a stop is the language of article 19, with its prohibition of “unreasonable” seizures. It is settled law that such a stop is reasonable and, hence, constitutional if the police have an articulable basis to suspect that a driver is operating under the influence of intoxicants (DWI). Although the police do not have probable cause to arrest, they may constitutionally stop the car to question and examine the driver. State v. Brodeur, 126 N.H. 411, 493 A.2d 1134 (1985); State v. Arsenault, 115 N.H. 109, 336 A.2d 244 (1975); State v. Landry supra; Berkemer v. McCarty, 104 S. Ct. 3138, 3150 (1984); Terry v. Ohio, 392 U.S. 1 (1968).
The articulable basis that justifies such a stop as reasonable for constitutional purposes normally consists of erratic or dangerous driving. In this case, however, we consider whether there are any circumstances in which the public interest in preventing and prosecuting DWI renders it reasonable for the police to stop a car before its driver has operated in such an exceptionable manner. The question posed is a narrow one: does article 19 allow the police to employ a systematic roadblock to stop and detain drivers for a brief time, in order to observe their condition for signs of the influence of intoxicants?
The court has never held that a roadblock to detect violations of motor vehicle laws is per se unreasonable, see State v. Severance *296supra and State v. Baldwin supra, and I do not understand that the majority intends to hold that now. Rather, the question of the reasonableness of such a stop has turned on a traditional judicial balancing of the value of the roadblock to the State and broad public against its burden upon the individual drivers. For example, this court observed just last year that “[ejmploying a fourth amendment balancing test, courts have upheld the validity of road checks conducted for a number of different purposes, in each instance concluding that the State’s interest in conducting the road check outweighed the momentary inconvenience and minimal intrusion upon the privacy of each driver forced to stop.” State v. Baldwin, supra at 776, 475 A.2d at 525 (citations omitted). See State v. Severance supra; Brown v. Texas, 443 U.S. 47 (1979); Delaware v. Prouse supra; see also State v. Landry supra.
While the court today recognizes that such a comparative evaluation is necessary, the majority apparently do not employ the simple balancing test described in the preceding quotation. Rather, they state that the roadblock is improper unless it “significantly” advances the public interest in outweighing the individual disadvantage to given drivers who are stopped. Even assuming that such a weighted balancing test is appropriate, I could not join in the conclusion that the majority reach, for I believe that the evidence indicates that the value of the roadblocks in this case did significantly outweigh the minimal disadvantage to the drivers whose cars were stopped.
Looking first to the individual interests affected by a roadblock, we have already noted that an individual has a protected interest in privacy and security when driving a car. Absent an articulable basis for suspecting illegality, unbridled police discretion to stop a car is antithetical to the driver’s protected interests. A limitation on such discretion in the form of criteria to be systematically applied should therefore be a required feature of any roadblock procedure that could survive constitutional scrutiny. See Delaware v. Prouse, supra at 663, and examples of State cases cited in the majority opinion.
I would hold that the criteria employed by the police in this case were constitutionally sufficient. They included standards for selecting the location of the stops, for identifying the cars to be stopped, for limiting the number of cars stopped to the number that the police could expeditiously handle, and for governing the conduct of the police during the stops. There is no evidence or suggestion that the police departed from these standards or acted arbitrarily with any discriminatory result.
*297Since the roadblocks did not carry the risk of unbridled police discretion, we must next evaluate the actual intrusions that they involved. There was evidence that the areas of the stops were well-lit and that the police explained the objects of the stop to each driver, thus precluding any undue apprehension about police intentions. While the majority emphasize the number of stops in relation to the number of violations later charged (i.e., in six months, 1680 vehicles stopped, resulting in 18 DWI arrests), the individual stops themselves were extremely brief. There was evidence that the police normally detained a car for two minutes or less. It is therefore difficult to imagine a more minimal actual intrusion.
Turning to the public interest served by the stops, there is no need of argument to establish the legitimacy of public concern to protect life, health and property on the highways. Nor is there need to argue that this interest would be served by detecting impairment caused by intoxicants before the impairment manifests itself in the sort of behavior that would justify an individual stop. See State v. Landry supra. It appears from the evidence that the public interest was actually so served by the roadblocks in issue here. While we do not know the number of DWI convictions that resulted from arrests at roadblocks, the record does reveal that about ten percent of DWI arrests by the Concord Police Department during a period of six months resulted from roadblock stops. Considering the high risks involved in DWI, this percentage is significant. We need not assign a weight, then, to the probable deterrent effect of publicity about the roadblocks, in order to conclude that roadblocks have real value.
Contrasting this significant public benefit with the minimal private intrusion, I conclude that the roadblocks in question did not result in unreasonable seizures. Just as we held license and registration checks constitutional in State v. Severance, 108 N.H. 404, 237 A.2d 683 (1968), I would hold these roadblocks constitutional under article 19, reaffirming the rule that “‘[t]he State can practice preventative [sic] therapy by reasonable road checks to ascertain whether man and machine meet the legislative determination of fitness. That this requires a momentary stopping of the traveling citizen is not fatal.’” Id., at 406, 237 A.2d at 685 (quoting Myricks v. United States, 370 F.2d 901, 904 (5th Cir. 1967)).
Before concluding, however, I believe that I should speak to the final note of the majority opinion, which quotes the Supreme Court of Oklahoma in raising the spectre of a “police state” if DWI roadblocks should be allowed. See State v. Smith, 674 P.2d 562, 564-65 (Okla. 1984). The Oklahoma court suggests that if the State could stop all drivers to detect DWI, it could stop all pedestrians to detect *298shoplifting. The justices of that court thus see roadblocks as the thin-end of a constitutional wedge.
The suggestion is unjustifiable, however. Shopping, like most other normal public pursuits, is not so fraught with risk to others as to be a regulated activity. Driving a car, conversely, is a public pursuit that raises just such a risk and is pervasively regulated in order to mitigate its potential danger. Measures that would be reasonable in policing activities of great risk would not be reasonable as intrusions into the characteristically safe and innocent pursuits of social life. The suggestion of the Oklahoma judges ignores these distinctions. I believe that the constitution is well-served by recognizing them. I therefore dissent, despite my respect for the concern that prompts my Brothers to rule as they do.