delivered the opinion of the Court.'

This case is here on appeal from the Supreme Court of Arizona under § 237 of the Judicial Code as amended, 28 U. S. C. § 344 (now 28 U. S. C. § 1257). It involves the constitutional validity of the following amendment to the Arizona Constitution, adopted at the 1946 general election:

“No person shall be denied the opportunity to obtain or retain employment because of non-membership in a labor organization, nor shall the state or any subdivision thereof, or any corporation, individual or association of any kind enter into any agreement, written or oral, which excludes any person from employment or continuation of employment because of non-membership in a labor organization.”

The Supreme Court of Arizona sustained the amendment as constitutional against the contentions that it “deprived union appellants of rights guaranteed under the First Amendment and protected against invasion by the State under the Fourteenth Amendment to the *540United States Constitution”; that it impaired the obligations of existing contracts in violation of Art. I, § 10, of the United States Constitution; and that it deprived appellants of due process of law, and denied them equal protection of the laws contrary to the Fourteenth Amendment. All of these questions, properly reserved in the state court, were decided against the appellants by the State Supreme Court.1 The same questions raised in the state court are presented here.

For reasons given in two other cases decided today we reject the appellants' contentions that the Arizona amendment denies them freedom of speech, assembly or petition, impairs the obligation of their contracts, or deprives them of due process of law. Lincoln Federal Labor Union v. Northwestern Iron & Metal Co. and Whitaker v. North Carolina, ante, p. 525. A difference between the Arizona amendment and the amendment and statute considered in the Nebraska and North Carolina cases has made it necessary for us to give separate consideration to the contention in this case that the Arizona amendment denies appellants equal protection of the laws.

The language of the Arizona amendment prohibits employment discrimination against non-union workers, but it does not prohibit discrimination against union workers. It is argued that a failure to provide the same protection for union workers as that provided for non-union workers places the union workers at a disadvantage, thus denying unions and their members the equal protection of Arizona's laws.

Although the Arizona amendment does not itself expressly prohibit discrimination against union workers, that state has not left unions and union members without protection from discrimination on account of union mem*541bership. Prior to passage of this constitutional amendment, Arizona made it a misdemeanor for any person to coerce a worker to make a contract “not to join, become or remain, a member of any labor organization” as a condition of getting or holding a job in Arizona. A section of the Arizona Code made every such contract (generally known as a “yellow dog contract”) void and unenforceable.2 Similarly, the Arizona constitutional amendment makes void and unenforceable contracts under which an employer agrees to discriminate against non-union workers. Statutes implementing the amendment have provided as sanctions for its enforcement relief by injunction and suits for damages for discrimination practiced in violation of the amendment.3 Whether the same kind of sanctions would be afforded a union worker against whom an employer discriminated is not made clear by the opinion of the State Supreme Court in this case. But assuming that Arizona courts would not afford a remedy by injunction or suit for damages, we are unable to find any indication that Arizona’s amendment and statutes are weighted on the side of non-union as against union workers. We are satisfied that Arizona has attempted both in the anti-yellow-dog-contract law and in the anti-discrimination constitutional amendment to strike at what were considered evils, to strike where those evils were most felt, and to strike in a manner that would effectively suppress the evils.

In Labor Board v. Jones & Laughlin Corp., 301 U. S. 1, this Court considered a challenge to the National Labor Relations Act on the ground that it applied restraints against employers but did not apply similar restraints against wrongful conduct by employees. We there pointed out, at p! 46, the general rule that “legislative *542authority, exerted within its proper field, need not embrace all the evils within its reach.” And concerning state laws we have said that the existence of evils against which the law should afford protection and the relative need of different groups for that protection “is a matter for the legislative judgment.” West Coast Hotel Co. v. Parrish, 300 U. S. 379, 400. We cannot say that the Arizona amendment has denied appellants equal protection of the laws.

Affirmed.

concurring.*

Arizona, Nebraska, and North Carolina have passed laws forbidding agreements to employ only union members. The United States Constitution is invoked against these laws. Since the cases bring into question the judicial process in its application to the Due Process Clause, explicit avowal of individual attitudes towards that process may elucidate and thereby strengthen adjudication. Accordingly, I set forth the steps by which I have reached concurrence with my brethren on what I deem the only substantial issue here, on all other issues joining the Court’s opinion.

The coming of the machine age tended to despoil human personality. It turned men and women into “hands.” The industrial history of the early Nineteenth Century demonstrated the helplessness of the individual employee to achieve human dignity in a society so largely affected by technological advances. Hence the trade union made itself increasingly felt, not only as an indispensable weapon of self-defense on the part of work*543ers but as an aid to the well-being of a society in which work is an expression of life and not merely the means of earning subsistence. But unionization encountered the shibboleths of a pre-machine age and these were reflected in juridical assumptions that survived the facts on which they were based. Adam Smith was treated as though his generalizations had been imparted to him on Sinai and not as a thinker who addressed himself to the elimination of restrictions which had become fetters upon initiative and enterprise in his day. Basic human rights expressed by the constitutional conception of “liberty” were equated with theories of laissez faire.1 The result was that economic views of confined validity were treated by lawyers and judges as though the Framers had enshrined them in the Constitution. This misapplication of the notions of the classic economists and resulting disregard of the perduring reach of the Constitution led to Mr. Justice Holmes’ famous protest in the Lochner case against measuring the Fourteenth Amendment by Mr. Herbert Spencer’s Social Statics. 198 U. S. 45, 75. Had not Mr. Justice Holmes’ awareness of the impermanence of legislation as against the permanence of the Constitution gradually prevailed, there might indeed have been “hardly any limit but the sky” to the embodiment of “our economic or moral beliefs” in that Amendment’s “prohibitions.” Baldwin v. Missouri, 281 U. S. 586, 595.

The attitude which regarded any legislative encroachment upon the existing economic order as infected with unconstitutionality led to disrespect for legislative attempts to strengthen the wage-earner’s bargaining power. *544With that attitude as a premise, Adair v. United States, 208 U. S. 161, and Coppage v. Kansas, 236 U. S. 1, followed logically enough; not even Truax v. Corrigan, 257 U. S. 312, could be considered unexpected. But when the tide turned, it was not merely because circumstances had changed and there had arisen a new order with new claims to divine origin. The opinion of Mr. Justice Brandéis in Senn v. Tile Layers Union, 301 U. S. 468, shows the current running strongly in the new direction— the direction not of social dogma but of increased deference to the legislative judgment. “Whether it was wise,” he said, now speaking for the Court and not in dissent, “for the State to permit the unions to [picket] is a question of its public policy — not our concern.” Id. at 481. Long before that, in Duplex Printing Press Co. v. Deering, 254 U. S. 443, 488, he had warned:

“All rights are derived from the purposes of the society in which they exist; above all rights rises duty to the community. The conditions developed in industry may be such that those engaged in it cannot continue their struggle without danger to the community. But it is not for judges to determine whether such conditions exist, nor is it their function to set the limits of permissible contest and to declare the duties which the new situation demands. This is the function of the legislature which, while limiting individual and group rights of aggression and defense, may substitute processes of justice for the more primitive method of trial by combat.”

Unions are powers within the State. Like the power of industrial and financial aggregations, the power of organized labor springs from a group which is only a fraction of the whole that Mr. Justice Holmes referred to as “the one club to which we all belong.” The power of the former is subject to control, though, of course, the *545particular incidence of control may be brought to test at the bar of this Court. E. g., Northern Securities Co. v. United States, 193 U. S. 197; North American Co. v. S. E. C., 327 U. S. 686. Neither can the latter claim constitutional exemption. Even the Government — the organ of the whole people — -is restricted by the system of checks and balances established by our Constitution. The designers of that system distributed authority among the three branches “not to promote efficiency but to preclude the exercise of arbitrary power.” Mr. Justice Brandéis, dissenting in Myers v. United States, 272 U. S. 52, 293. Their concern for individual members of society, for whose well-being government is instituted, gave urgency to the fear that concentrated power would become arbitrary. It is a fear that the history of such power, even when professedly employed for democratic purposes, has hardly rendered unfounded.

If concern for the individual justifies incorporating in the Constitution itself devices to curb public authority, a legislative judgment that his protection requires the regulation of the private power of unions cannot be dismissed as insupportable. A union is no more than a medium through which individuals are able to act together; union power was begotten of individual helplessness. But that power can come into being only when, and continue to exist only so long as, individual aims are seen to be shared in common with the other members of the group. There is a natural emphasis, however, on what is shared and a resulting tendency to subordinate the inconsistent interests and impulses of individuals. From this, it is an easy transition to thinking of the union as an entity having rights and purposes of its own. An ardent supporter of trade unions who is also no less a disinterested student of society has pointed out that “As soon as we personify the idea, whether it is a country or a church, a trade union or an employers’ association, we obscure *546individual responsibility by transferring emotional loyalties to a fictitious creation which then acts upon us psychologically as an obstruction, especially in times of crisis, to the critical exercise of a reasoned judgment.” Laski, Morris Cohen’s Approach to Legal Philosophy, 15 U. of Chi. L. Rev. 575, 581 (1948).

The right of association, like any other right carried to its extreme, encounters limiting principles. See Hudson County Water Co. v. McCarter, 209 U. S. 349, 355. At the point where the mutual advantage of association demands too much individual disadvantage, a compromise must be struck. See Dicey, Law and Public Opinion in England 465-66 (1905). When that point has been reached — where the intersection should fall — is plainly a question within the special province of the legislature. This Court has given effect to such a compromise in sustaining a legislative purpose to protect individual employees against the exclusionary practices of unions. Steele v. Louisville & N. R. Co., 323 U. S. 192; Wallace Corp. v. Labor Board, 323 U. S. 248; Railway Mail Assn. v. Corsi, 326 U. S. 88; cf. Elgin, J. & E. R. Co. v. Burley, 325 U. S. 711, 733-34. The rationale of the Arizona, Nebraska, and North Carolina legislation prohibiting union-security agreements is founded on a similar resolution of conflicting interests.2 Unless we are to treat *547as unconstitutional what goes against the grain because it offends what we may strongly believe to be socially desirable, that resolution must be given respect.

It is urged that the compromise which this legislation embodies is no compromise at all because fatal to the survival of organized labor. But can it be said that the legislators and the people of Arizona, Nebraska, and North Carolina could not in reason be sceptical of organized labor’s insistence upon the necessity to its strength of power to compel rather than to persuade the allegiance of its reluctant members? In the past fifty years the total number of employed, counting salaried workers and the self-employed but not farmers or farm laborers, has not quite trebled, while total union membership has increased more than thirty-three times; at the time of the open-shop drive following the First World War, the ratio of organized to unorganized non-agricultural workers was about one to nine, and now it is almost one to three.3 However necessitous may have been the circumstances of unionism in 1898 or even in 1923, its status in 1948 precludes constitutional condemnation of a legislative judgment, whatever we may think of it, that the need of this type of regulation outweighs its detriments. It would be arbitrary for this Court to deny the States the right to experiment with such laws, especially in view of the fact that the Railroad Brotherhoods have held their own de*548spite congressional prohibition of union security4 and in the light of the experience of countries advanced in industrial democracy, such as Great Britain and Sweden, where deeply rooted acceptance of the principles of collective *549bargaining is not reflected in uncompromising demands for contractually guaranteed security.5 Whether it is preferable in the public interest that trade unions should be subjected to State intervention or left to the free play of social forces, whether experience has disclosed “union unfair labor practices” and, if so, whether legislative correction is more appropriate than self-discipline and the *550pressure of public opinion — these are questions on which it is not for us to express views. The very limited function of this Court is discharged when we recognize that these issues are not so unrelated to the experience and feelings of the community as to render legislation addressing itself to them wilfully destructive of cherished rights. For these are not matters, like censorship of the press or separation of Church and State, on which history, through the Constitution, speaks so decisively as to forbid legislative experimentation.

But the policy which finds expression in the prohibition of union-security agreements need not rest solely on a legislative conception of the public interest which includes but transcends the special claims of trade unions. The States are entitled to give weight to views combining opposition to the “closed shop” with long-range concern for the welfare of trade unions. Mr. Justice Brandéis, for example, before he came to this Court, had been a staunch promoter of unionism. In testifying before the Commission on Industrial Relations, he said:

“I should say to those employers who stand for the open shop, that they ought to recognize that it is for their interests as well as that of the community that unions should be powerful and responsible; that it is to their interests to build up the union; to aid as far as they can in making them stronger; and to create conditions under which the unions shall be led by the ablest and most experienced men.” 6

*551Yet at the same time he believed that “The objections, legal, economic, and social, against the closed shop are so strong, and the ideas of the closed shop so antagonistic to the American spirit, that the insistence upon it has been a serious obstacle to union progress.” Letter of Sept. 6, 1910, to Lawrence F. Abbott of the Outlook.7 On another occasion he wrote, “But the American people should not, and will not, accept unionism if it involves the closed shop. They will not consent to the exchange of the tyranny of the employer for the tyranny of the employee.” Letter of Feb. 26, 1912, to Lincoln Steffens.8 In summing up his views on unionism, he said:

“It is not true that the 'success of a labor union’ necessarily means a ‘perfect monopoly’. The union, in order to attain or preserve for its members industrial liberty, must be strong and stable. It need not include every member of the trade. Indeed, it is desirable for both the employer and the union that it should not. Absolute power leads to excesses and to weakness: Neither our character nor our intelligence can long bear the strain of unrestricted power. The union attains success when it reaches *552the ideal condition, and the ideal condition for a union is to be strong and stable, and yet to have in the trade outside its own ranks an appreciable number of men who are non-unionists. In any free community the diversity of character, of beliefs, of taste — indeed mere selfishness — -will insure such a supply, if the enjoyment of this privilege of individualism is protected by law. Such a nucleus of unorganized labor will check oppression by the union as the union checks oppression by the employer.” Quoted from Louis D. Brandéis’ contribution to a discussion entitled Peace with Liberty and Justice in 2 Nat. Civic Federation Rev., No. 2, pp. 1, 16 (May 15, 1905).

Mr. Brandéis on the long view deemed the preferential shop a more reliable form of security both for unions and for society than the closed shop; that he did so only serves to ¡Drove that these are pragmatic issues not appropriate for dogmatic solution.

Whatever one may think of Mr. Brandéis’ views, they have been reinforced by the adoption of laws insuring against that undercutting of union standards which was one of the most serious effects of a dissident minority in a union shop. Under interpretations of the National Labor Relations Act undisturbed by the Taft-Hartley Act,9 and of the Railway Labor Act, the bargaining representative designated by a majority of employees has exclusive power to deal with the employer on matters of wages and working conditions. Individual contracts, whether on more or less favorable terms than those obtained by the union, are barred. J. I. Case Co. v. Labor Board, 321 U. S. 332; Order of R. R. Telegraphers v. Railway Express Agency, 321 U. S. 342; Medo Photo Supply Corp. v. Labor Board, 321 U. S. 678; see Elgin, *553J. & E. R. Co. v. Burley, 325 U. S. 711, 737, n. 35. Under these laws, a non-union bidder for a job in a union shop cannot, if he would, undercut the union standards.

Even where the social undesirability of a law may be convincingly urged, invalidation of the law by a court debilitates popular democratic government. Most laws dealing with economic and social problems are matters of trial and error.10 That which before trial appears to be demonstrably bad may belie prophecy in actual operation. It may not prove good, but it may prove innocuous. But even if a law is found wanting on trial, it is better that its defects should be demonstrated and removed than that the law should be aborted by judicial fiat. Such an assertion of judicial power deflects responsibility from those on whom in a democratic society it ultimately rests — the people. If the proponents of union-security agreements have confidence in the arguments addressed to the Court in their “economic brief,” they should address those arguments to the electorate. Its endorsement would be a vindication that the mandate of this Court could never give. That such vindication *554is not a vain hope has been recently demonstrated by the voters of Maine, Massachusetts, and New Mexico.11 And although several States in addition to those at bar now have such laws,12 the legislatures of as many other States have, sometimes repeatedly, rejected them.13 What one State can refuse to do, another can undo.

*555But there is reason for judicial restraint in matters of policy deeper than the value of experiment: it is founded on a recognition of the gulf of difference between sustaining and nullifying legislation. This difference is theoretical in that the function of legislating is for legislatures who have also taken oaths to support the Constitution, while the function of courts, when legislation is challenged, is merely to make sure that the legislature has exercised an allowable judgment, and not to exercise their own judgment, whether a policy is within or without “the vague contours” of due process. Theory is reinforced by the notorious fact that lawyers predominate in American legislatures.14 In practice also the difference is wide. In the day-to-day working of our democracy it is vital that the power of the non-democratic organ of our Government be exercised with rigorous self-restraint. Because the powers exercised by this Court are inherently oligarchic, Jefferson all of his life thought of the Court as “an irresponsible body”15 and “independent of the nation itself.” 16 The Court is not saved from being oli*556garchic because it professes to act in the service of humane ends. As history amply proves, the judiciary is prone to misconceive the public good by confounding private notions with constitutional requirements, and such misconceptions are not subject to legitimate displacement by the will of the people except at too slow a pace.17 Judges appointed for life whose decisions run counter to prevailing opinion cannot be voted out of office and supplanted by men of views more consonant with it. They are even farther removed from democratic pressures by the fact that their deliberations are in secret and remain beyond disclosure either by periodic reports or by such a modern device for securing responsibility to the electorate as the “press conference.” But a democracy need not rely on the courts to save it from its own unwisdom. If it is alert — and without alertness by the people there can be no enduring democracy — unwise or unfair legislation can readily be removed from the statute books. It is by such vigilance over its representatives that democracy proves itself.

Our right to pass on the validity of legislation is now too much part of our constitutional system to be brought *557into question. But the implications of that right and the conditions for its exercise must constantly be kept in mind and vigorously observed. Because the Court is without power to shape measures for dealing with the problems of society but has merely the power of negation over measures shaped by others, the indispensable judicial requisite is intellectual humility, and such humility presupposes complete disinterestedness. And so, in the end, it is right that the Court should be indifferent to public temper and' popular wishes. Mr. Dooley’s “th’ Supreme Coort follows th’ iliction returns” expressed the wit of cynicism, not the demand of principle. A court which yields to the popular will thereby licenses itself to practice despotism, for there can be no assurance that it will not on another occasion indulge its own will. Courts can fulfill their responsibility in a democratic society only to the extent that they succeed in shaping their judgments by rational standards, and rational standards are both impersonal and communicable. Matters of policy, however, are by definition' matters which demand the resolution of conflicts of value, and the elements of conflicting values are largely imponderable. Assessment of their competing worth involves differences of feeling; it is also an exercise in prophecy. Obviously the proper forum for mediating a clash of feelings and rendering a prophetic judgment is the body chosen for those purposes by the people. Its functions can be assumed by this Court only in disregard of the historic limits of the Constitution.

concurring.*

I concur in the Court’s judgment in No. 34, Whitaker v. North Carolina. The appellants were convicted under *558a warrant which charged only, in effect, that they had violated the statute “by executing a written agreement or contract” for a closed or union shop.1 There was neither charge nor evidence that the employer, after the statute became effective, had refused employment to any person because he was not a member of a union. The charge, therefore, and the conviction were limited to the making of the contract. No other provision of the statute is now involved, as the state’s attorney general conceded, indeed as he strongly urged, in the argument here. As against the constitutional objections raised to this application of the statute, I agree that the legislature has power to proscribe the making of such contracts, and accordingly join in the judgment affirming the convictions.

In No. 27, American Federation of Labor v. American Sash & Door Company, and in No. 47, Lincoln Federal Labor Union v. Northwestern Iron & Metal Company, as against the constitutional questions now raised, *559I am also in agreement with the Court’s decision, but subject to the following reservation. Because no strike has been involved in any of the states of fact, no question has been presented in any of these cases immediately involving the right to strike or concerning the effect of the Thirteenth Amendment. Yet the issues so closely approach touching that right as it exists or may exist under that Amendment that the possible effect of the decisions upon it hardly can be ignored.2 Strikes have been called throughout union history in defense of the right of union members not to work with nonunion men. If today’s decision should be construed to permit a state to foreclose that right by making illegal the concerted refusal of union members to work with nonunion workers, and more especially if the decision should be taken as going so far as to permit a state to enjoin such a strike,3 I should want a complete and thorough reargument of these cases before deciding so momentous a question.

But the right to prohibit contracts for union security is one thing. The right to force union members to work with nonunion workers is entirely another. Because of this difference, I expressly reserve judgment upon the latter question until it is squarely and inescapably presented. Although this reservation is not made expressly by the Court, I do not understand its opinion to foreclose this question.