Considerations on Representative Government

by John Stuart Mill

Chapter XIV—Of the Executive in a Representative Government

It would be out of place in this treatise to discuss the question into what departments or branches the executive business of government may most conveniently be divided. In this respect the exigencies of different governments are different; and there is little probability that any great mistake will be made in the classification of the duties when men are willing to begin at the beginning, and do not hold themselves bound by the series of accidents which, in an old government like ours, has produced the existing division of the public business. It may be sufficient to say that the classification of functionaries should correspond to that of subjects, and that there should not be several departments independent of one another, to superintend different parts of the same natural whole, as in our own military administration down to a recent period, and in a less degree even at present. Where the object to be attained is single (such as that of having an efficient army), the authority commissioned to attend to it should be single likewise. The entire aggregate of means provided for one end should be under one and the same control and responsibility. If they are divided among independent authorities, the means with each of those authorities become ends, and it is the business of nobody except the head of the government, who has probably no departmental experience, to take care of the real end. The different classes of means are not combined and adapted to one another under the guidance of any leading idea; and while every department pushes forward its own requirements, regardless of those of the rest, the purpose of the work is perpetually sacrificed to the work itself.

As a general rule, every executive function, whether superior or subordinate, should be the appointed duty of some given individual. It should be apparent to all the world who did every thing, and through whose default any thing was left undone. Responsibility is null when nobody knows who is responsible; nor, even when real, can it be divided without being weakened. To maintain it at its highest, there must be one person who receives the whole praise of what is well done, the whole blame of what is ill. There are, however, two modes of sharing responsibility; by one it is only enfeebled, by the other absolutely destroyed. It is enfeebled when the concurrence of more than one functionary is required to the same act. Each one among them has still a real responsibility; if a wrong has been done, none of them can say he did not do it; he is as much a participant as an accomplice is in an offense: if there has been legal criminality, they may all be punished legally, and their punishment needs not be less severe than if there had been only one person concerned. But it is not so with the penalties any more than with the rewards of opinion; these are always diminished by being shared. Where there has been no definite legal offense, no corruption or malversation, only an error or an imprudence, or what may pass for such, every participator has an excuse to himself and to the world in the fact that other persons are jointly involved with him. There is hardly any thing, even to pecuniary dishonesty, for which men will not feel themselves almost absolved, if those whose duty it was to resist and remonstrate have failed to do it, still more if they have given a formal assent.

In this case, however, though responsibility is weakened, there still is responsibility: every one of those implicated has in his individual capacity assented to, and joined in the act. Things are much worse when the act itself is only that of a majority—a board deliberating with closed doors, nobody knowing, or, except in some extreme case, being ever likely to know, whether an individual member voted for the act or against it. Responsibility in this case is a mere name. "Boards," it is happily said by Bentham, "are screens." What "the Board" does is the act of nobody, and nobody can be made to answer for it. The Board suffers, even in reputation, only in its collective character; and no individual member feels this further than his disposition leads him to identify his own estimation with that of the body—a feeling often very strong when the body is a permanent one, and he is wedded to it for better for worse; but the fluctuations of a modern official career give no time for the formation of such an esprit de corps, which, if it exists at all, exists only in the obscure ranks of the permanent subordinates. Boards, therefore, are not a fit instrument for executive business, and are only admissible in it when, for other reasons, to give full discretionary power to a single minister would be worse.

On the other hand, it is also a maxim of experience that in the multitude of councillors there is wisdom, and that a man seldom judges right, even in his own concerns, still less in those of the public, when he makes habitual use of no knowledge but his own, or that of some single adviser. There is no necessary incompatibility between this principle and the other. It is easy to give the effective power and the full responsibility to one, providing him when necessary with advisers, each of whom is responsible only for the opinion he gives.

In general, the head of a department of the executive government is a mere politician. He may be a good politician, and a man of merit; and, unless this is usually the case, the government is bad. But his general capacity, and the knowledge he ought to possess of the general interests of the country, will not, unless by occasional accident, be accompanied by adequate, and what may be called professional knowledge of the department over which he is called to preside. Professional advisers must therefore be provided for him. Wherever mere experience and attainments are sufficient—wherever the qualities required in a professional adviser may possibly be united in a single well-selected individual (as in the case, for example, of a law officer), one such person for general purposes, and a staff of clerks to supply knowledge of details, meet the demands of the case. But, more frequently, it is not sufficient that the minister should consult some one competent person, and, when himself not conversant with the subject, act implicitly on that person's advice. It is often necessary that he should, not only occasionally, but habitually, listen to a variety of opinions, and inform his judgment by the discussions among a body of advisers. This, for example, is emphatically necessary in military and naval affairs. The military and naval ministers, therefore, and probably several others, should be provided with a Council, composed, at least in those two departments, of able and experienced professional men. As a means of obtaining the best men for the purpose under every change of administration, they ought to be permanent; by which I mean that they ought not, like the Lords of the Admiralty, to be expected to resign with the ministry by whom they were appointed; but it is a good rule that all who hold high appointments to which they have risen by selection, and not by the ordinary course of promotion, should retain their office only for a fixed term, unless reappointed, as is now the rule with staff appointments in the British army. This rule renders appointments somewhat less likely to be jobbed, not being a provision for life, and the same time affords a means, without affront to any one, of getting rid of those who are least worth keeping, and bringing in highly qualified persons of younger standing, for whom there might never be room if death vacancies, or voluntary resignations were waited for.

The councils should be consultative merely, in this sense, that the ultimate decision should rest undividedly with the minister himself; but neither ought they to be looked upon, or to look upon themselves as ciphers, or as capable of being reduced to such at his pleasure. The advisers attached to a powerful and perhaps self-willed man ought to be placed under conditions which make it impossible for them, without discredit, not to express an opinion, and impossible for him not to listen to and consider their recommendations, whether he adopts them or not. The relation which ought to exist between a chief and this description of advisers is very accurately hit by the constitution of the Council of the Governor General and those of the different Presidencies in India. These councils are composed of persons who have professional knowledge of Indian affairs, which the governor general and governors usually lack, and which it would not be desirable to require of them. As a rule, every member of council is expected to give an opinion, which is of course very often a simple acquiescence; but if there is a difference of sentiment, it is at the option of every member, and is the invariable practice, to record the reasons of his opinion, the governor general, or governor, doing the same. In ordinary cases the decision is according to the sense of the majority; the council, therefore, has a substantial part in the government; but if the governor general, or governor, thinks fit, he may set aside even their unanimous opinion, recording his reasons. The result is, that the chief is individually and effectively responsible for every act of the government. The members of council have only the responsibility of advisers; but it is always known, from documents capable of being produced, and which, if called for by Parliament or public opinion always are produced, what each has advised, and what reasons he gave for his advice; while, from their dignified position, and ostensible participation in all acts of government, they have nearly as strong motives to apply themselves to the public business, and to form and express a well-considered opinion on every part of it, as if the whole responsibility rested with themselves.

This mode of conducting the highest class of administrative business is one of the most successful instances of the adaptation of means to ends which political history, not hitherto very prolific in works of skill and contrivance, has yet to show. It is one of the acquisitions with which the art of politics has been enriched by the experience of the East India Company's rule; and, like most of the other wise contrivances by which India has been preserved to this country, and an amount of good government produced which is truly wonderful considering the circumstances and the materials, it is probably destined to perish in the general holocaust which the traditions of Indian government seem fated to undergo since they have been placed at the mercy of public ignorance and the presumptuous vanity of political men. Already an outcry is raised for abolishing the councils as a superfluous and expensive clog on the wheels of government; while the clamor has long been urgent, and is daily obtaining more countenance in the highest quarters, for the abrogation of the professional civil service, which breeds the men that compose the councils, and the existence of which is the sole guaranty for their being of any value.

A most important principle of good government in a popular constitution is that no executive functionaries should be appointed by popular election, neither by the votes of the people themselves, nor by those of their representatives. The entire business of government is skilled employment; the qualifications for the discharge of it are of that special and professional kind which can not be properly judged of except by persons who have themselves some share of those qualifications, or some practical experience of them. The business of finding the fittest persons to fill public employments—not merely selecting the best who offer, but looking out for the absolutely best, and taking note of all fit persons who are met with, that they may be found when wanted—is very laborious, and requires a delicate as well as highly conscientious discernment; and as there is no public duty which is in general so badly performed, so there is none for which it is of greater importance to enforce the utmost practicable amount of personal responsibility, by imposing it as a special obligation on high functionaries in the several departments. All subordinate public officers who are not appointed by some mode of public competition should be selected on the direct responsibility of the minister under whom they serve. The ministers, all but the chief, will naturally be selected by the chief; and the chief himself, though really designated by Parliament, should be, in a regal government, officially appointed by the crown. The functionary who appoints should be the sole person empowered to remove any subordinate officer who is liable to removal, which the far greater number ought not to be, except for personal misconduct, since it would be vain to expect that the body of persons by whom the whole detail of the public business is transacted, and whose qualifications are generally of much more importance to the public than those of the minister himself, will devote themselves to their profession, and acquire the knowledge and skill on which the minister must often place entire dependence, if they are liable at any moment to be turned adrift for no fault, that the minister may gratify himself, or promote his political interest, by appointing somebody else.

To the principle which condemns the appointment of executive officers by popular suffrage, ought the chief of the executive, in a republican government, to be an exception? Is it a good rule which, in the American Constitution, provides for the election of the President once in every four years by the entire people? The question is not free from difficulty. There is unquestionably some advantage, in a country like America, where no apprehension needs be entertained of a coup d'état, in making the chief minister constitutionally independent of the legislative body, and rendering the two great branches of the government, while equally popular both in their origin and in their responsibility, an effective check on one another. The plan is in accordance with that sedulous avoidance of the concentration of great masses of power in the same hands, which is a marked characteristic of the American federal Constitution. But the advantage, in this instance, is purchased at a price above all reasonable estimates of its value. It seems far better that the chief magistrate in a republic should be appointed avowedly, as the chief minister in a constitutional monarchy is virtually, by the representative body. In the first place, he is certain, when thus appointed, to be a more eminent man. The party which has the majority in Parliament would then, as a rule, appoint its own leader, who is always one of the foremost, and often the very foremost person in political life; while the President of the United States, since the last survivor of the founders of the republic disappeared from the scene, is almost always either an obscure man, or one who has gained any reputation he may possess in some other field than politics. And this, as I have before observed, is no accident, but the natural effect of the situation. The eminent men of a party, in an election extending to the whole country, are never its most available candidates. All eminent men have made personal enemies, or, have done something, or at the lowest, professed some opinion obnoxious to some local or other considerable division of the community, and likely to tell with fatal effect upon the number of votes; whereas a man without antecedents, of whom nothing is known but that he professes the creed of the party, is readily voted for by its entire strength. Another important consideration is the great mischief of unintermitted electioneering. When the highest dignity in the state is to be conferred by popular election once in every few years, the whole intervening time is spent in what is virtually a canvass. President, ministers, chiefs of parties, and their followers, are all electioneerers: the whole community is kept intent on the mere personalities of politics, and every public question is discussed and decided with less reference to its merits than to its expected bearing on the presidential election. If a system had been devised to make party spirit the ruling principle of action in all public affairs, and create an inducement not only to make every question a party question, but to raise questions for the purpose of founding parties upon them, it would have been difficult to contrive any means better adapted to the purpose.

I will not affirm that it would at all times and places be desirable that the head of the executive should be so completely dependent upon the votes of a representative assembly as the prime minister is in England, and is without inconvenience. If it were thought best to avoid this, he might, though appointed by Parliament, hold his office for a fixed period, independent of a Parliamentary vote, which would be the American system minus the popular election and its evils. There is another mode of giving the head of the administration as much independence of the Legislature as is at all compatible with the essentials of free government. He never could be unduly dependent on a vote of Parliament if he had, as the British prime minister practically has, the power to dissolve the House and appeal to the people; if, instead of being turned out of office by a hostile vote, he could only be reduced by it to the alternative of resignation or dissolution. The power of dissolving Parliament is one which I think it desirable he should possess, even under the system by which his own tenure of office is secured to him for a fixed period. There ought not to be any possibility of that deadlock in politics which would ensue on a quarrel breaking out between a president and an assembly, neither of whom, during an interval which might amount to years, would have any legal means of ridding itself of the other. To get through such a period without a coup d'état being attempted, on either side or on both, requires such a combination of the love of liberty and the habit of self-restraint as very few nations have yet shown themselves capable of; and though this extremity were avoided, to expect that the two authorities would not paralyze each other's operations is to suppose that the political life of the country will always be pervaded by a spirit of mutual forbearance and compromise, imperturbable by the passions and excitements of the keenest party struggles. Such a spirit may exist, but even where it does there is imprudence in trying it too far.

Other reasons make it desirable that some power in the state (which can only be the executive) should have the liberty of at any time, and at discretion, calling a new Parliament. When there is a real doubt which of two contending parties has the strongest following, it is important that there should exist a constitutional means of immediately testing the point and setting it at rest. No other political topic has a chance of being properly attended to while this is undecided; and such an interval is mostly an interregnum for purposes of legislative or administrative improvement, neither party having sufficient confidence in its strength to attempt things likely to provoke opposition in any quarter that has either direct or indirect influence in the pending struggle.

I have not taken account of the case in which the vast power centralized in the chief magistrate, and the insufficient attachment of the mass of the people to free institutions, give him a chance of success in an attempt to subvert the Constitution, and usurp sovereign power. Where such peril exists, no first magistrate is admissible whom the Parliament can not, by a single vote, reduce to a private station. In a state of things holding out any encouragement to that most audacious and profligate of all breaches of trust, even this entireness of constitutional dependence is but a weak protection.

Of all officers of government, those in whose appointment any participation of popular suffrage is the most objectionable are judicial officers. While there are no functionaries whose special and professional qualifications the popular judgment is less fitted to estimate, there are none in whose case absolute impartiality, and freedom from connection with politicians or sections of politicians, are of any thing like equal importance. Some thinkers, among others Mr. Bentham, have been of opinion that, although it is better that judges should not be appointed by popular election, the people of their district ought to have the power, after sufficient experience, of removing them from their trust. It can not be denied that the irremovability of any public officer to whom great interests are intrusted is in itself an evil. It is far from desirable that there should be no means of getting rid of a bad or incompetent judge, unless for such misconduct as he can be made to answer for in a criminal court, and that a functionary on whom so much depends should have the feeling of being free from responsibility except to opinion and his own conscience. The question however is, whether, in the peculiar position of a judge, and supposing that all practicable securities have been taken for an honest appointment, irresponsibility, except to his own and the public conscience, has not, on the whole, less tendency to pervert his conduct than responsibility to the government or to a popular vote. Experience has long decided this point in the affirmative as regards responsibility to the executive, and the case is quite equally strong when the responsibility sought to be enforced is to the suffrages of electors. Among the good qualities of a popular constituency, those peculiarly incumbent upon a judge, calmness and impartiality, are not numbered. Happily, in that intervention of popular suffrage which is essential to freedom they are not the qualities required. Even the quality of justice, though necessary to all human beings, and therefore to all electors, is not the inducement which decides any popular election. Justice and impartiality are as little wanted for electing a member of Parliament as they can be in any transaction of men. The electors have not to award something which either candidate has a right to, nor to pass judgment on the general merits of the competitors, but to declare which of them has most of their personal confidence, or best represents their political convictions. A judge is bound to treat his political friend, or the person best known to him, exactly as he treats other people; but it would be a breach of duty, as well as an absurdity, if an elector did so. No argument can be grounded on the beneficial effect produced on judges, as on all other functionaries, by the moral jurisdiction of opinion; for even in this respect, that which really exercises a useful control over the proceedings of a judge, when fit for the judicial office, is not (except sometimes in political cases) the opinion of the community generally, but that of the only public by whom his conduct or qualifications can be duly estimated, the bar of his own court. I must not be understood to say that the participation of the general public in the administration of justice is of no importance; it is of the greatest; but in what manner? By the actual discharge of a part of the judicial office in the capacity of jurymen. This is one of the few cases in politics in which it is better that the people should act directly and personally than through their representatives, being almost the only case in which the errors that a person exercising authority may commit can be better borne than the consequences of making him responsible for them. If a judge could be removed from office by a popular vote, whoever was desirous of supplanting him would make capital for that purpose out of all his judicial decisions; would carry all of them, as far as he found practicable, by irregular appeal before a public opinion wholly incompetent, for want of having heard the case, or from having heard it without either the precautions or the impartiality belonging to a judicial hearing; would play upon popular passion and prejudice where they existed, and take pains to arouse them where they did not. And in this, if the case were interesting, and he took sufficient trouble, he would infallibly be successful, unless the judge or his friends descended into the arena, and made equally powerful appeals on the other side. Judges would end by feeling that they risked their office upon every decision they gave in a case susceptible of general interest, and that it was less essential for them to consider what decision was just, than what would be most applauded by the public, or would least admit of insidious misrepresentation. The practice introduced by some of the new or revised State Constitutions in America, of submitting judicial officers to periodical popular re-election, will be found, I apprehend, to be one of the most dangerous errors ever yet committed by democracy; and, were it not that the practical good sense which never totally deserts the people of the United States is said to be producing a reaction, likely in no long time to lead to the retraction of the error, it might with reason be regarded as the first great downward step in the degeneration of modern democratic government.

With regard to that large and important body which constitutes the permanent strength of the public service, those who do not change with changes of politics, but remain to aid every minister by their experience and traditions, inform him by their knowledge of business, and conduct official details under his general control—those, in short, who form the class of professional public servants, entering their profession as others do while young, in the hope of rising progressively to its higher grades as they advance in life—it is evidently inadmissible that these should be liable to be turned out, and deprived of the whole benefit of their previous service, except for positive, proved, and serious misconduct. Not, of course, such delinquency only as makes them amenable to the law, but voluntary neglect of duty, or conduct implying untrustworthiness for the purposes for which their trust is given them. Since, therefore, unless in case of personal culpability, there is no way of getting rid of them except by quartering them on the public as pensioners, it is of the greatest importance that the appointments should be well made in the first instance; and it remains to be considered by what mode of appointment this purpose can best be attained.

In making first appointments, little danger is to be apprehended from want of special skill and knowledge in the choosers, but much from partiality, and private or political interest. Being all appointed at the commencement of manhood, not as having learned, but in order that they may learn, their profession, the only thing by which the best candidates can be discriminated is proficiency in the ordinary branches of liberal education; and this can be ascertained without difficulty, provided there be the requisite pains and the requisite impartiality in those who are appointed to inquire into it. Neither the one nor the other can reasonably be expected from a minister, who must rely wholly on recommendations, and, however disinterested as to his personal wishes, never will be proof against the solicitations of persons who have the power of influencing his own election, or whose political adherence is important to the ministry to which he belongs. These considerations have introduced the practice of submitting all candidates for first appointments to a public examination, conducted by persons not engaged in politics, and of the same class and quality with the examiners for honors at the Universities. This would probably be the best plan under any system; and under our Parliamentary government it is the only one which affords a chance, I do not say of honest appointment, but even of abstinence from such as are manifestly and flagrantly profligate.

It is also absolutely necessary that the examinations should be competitive, and the appointments given to those who are most successful. A mere pass examination never, in the long run, does more than exclude absolute dunces. When the question, in the mind of an examiner, lies between blighting the prospects of an individual and performing a duty to the public which, in the particular instance, seldom appears of first rate importance, and when he is sure to be bitterly reproached for doing the first, while in general no one will either know or care whether he has done the latter, the balance, unless he is a man of very unusual stamp, inclines to the side of good-nature. A relaxation in one instance establishes a claim to it in others, which every repetition of indulgence makes it more difficult to resist; each of these, in succession, becomes a precedent for more, until the standard of proficiency sinks gradually to something almost contemptible. Examinations for degrees at the two great Universities have generally been as slender in their requirements as those for honors are trying and serious. Where there is no inducement to exceed a certain minimum, the minimum comes to be the maximum: it becomes the general practice not to aim at more; and as in every thing there are some who do not attain all they aim at, however low the standard may be pitched, there are always several who fall short of it. When, on the contrary, the appointments are given to those, among a great number of candidates, who most distinguish themselves, and where the successful competitors are classed in order of merit, not only each is stimulated to do his very utmost, but the influence is felt in every place of liberal education throughout the country. It becomes with every schoolmaster an object of ambition and an avenue to success to have furnished pupils who have gained a high place in these competitions, and there is hardly any other mode in which the state can do so much to raise the quality of educational institutions throughout the country. Though the principle of competitive examinations for public employment is of such recent introduction in this country, and is still so imperfectly carried out, the Indian service being as yet nearly the only case in which it exists in its completeness, a sensible effect has already begun to be produced on the places of middle-class education, notwithstanding the difficulties which the principle has encountered from the disgracefully low existing state of education in the country, which these very examinations have brought into strong light. So contemptible has the standard of acquirement been found to be, among the youths who obtain the nomination from the minister, which entitles them to offer themselves as candidates, that the competition of such candidates produces almost a poorer result than would be obtained from a mere pass examination; for no one would think of fixing the conditions of a pass examination so low as is actually found sufficient to enable a young man to surpass his fellow-candidates. Accordingly, it is said that successive years show on the whole a decline of attainments, less effort being made, because the results of former examinations have proved that the exertions then used were greater than would have been sufficient to attain the object. Partly from this decrease of effort, and partly because, even at the examinations which do not require a previous nomination, conscious ignorance reduces the number of competitors to a mere handful, it has so happened that though there have always been a few instances of great proficiency, the lower part of the list of successful candidates represents but a very moderate amount of acquirement; and we have it on the word of the commissioners that nearly all who have been unsuccessful have owed their failure to ignorance, not of the higher branches of instruction, but of its very humblest elements—spelling and arithmetic.

The outcries which continue to be made against these examinations by some of the organs of opinion are often, I regret to say, as little creditable to the good faith as to the good sense of the assailants. They proceed partly by misrepresentation of the kind of ignorance which, as a matter of fact, actually leads to failure in the examinations. They quote with emphasis the most recondite questions [8] which can be shown to have been ever asked, and make it appear as if unexceptionable answers to all these were made the sine quâ non of success. Yet it has been repeated to satiety that such questions are not put because it is expected of every one that he should answer them, but in order that whoever is able to do so may have the means of proving and availing himself of that portion of his knowledge. It is not as a ground of rejection, but as an additional means of success, that this opportunity is given. We are then asked whether the kind of knowledge supposed in this, that, or the other question, is calculated to be of any use to the candidate after he has attained his object. People differ greatly in opinion as to what knowledge is useful. There are persons in existence, and a late Foreign Secretary of State is one of them, who think English spelling a useless accomplishment in a diplomatic attaché or a clerk in a government office. About one thing the objectors seem to be unanimous, that general mental cultivation is not useful in these employments, whatever else may be so. If, however (as I presume to think), it is useful, or if any education at all is useful, it must be tested by the tests most likely to show whether the candidate possesses it or not. To ascertain whether he has been well educated, he must be interrogated in the things which he is likely to know if he has been well educated, even though not directly pertinent to the work to which he is to be appointed. Will those who object to his being questioned in classics and mathematics, tell us what they would have him questioned in? There seems, however, to be equal objection to examining him in these, and to examining him in any thing but these. If the Commissioners—anxious to open a door of admission to those who have not gone through the routine of a grammar-school, or who make up for the smallness of their knowledge of what is there taught by greater knowledge of something else—allow marks to be gained by proficiency in any other subject of real utility, they are reproached for that too. Nothing will satisfy the objectors but free admission of total ignorance.

We are triumphantly told that neither Clive nor Wellington could have passed the test which is prescribed for an aspirant to an engineer cadetship; as if, because Clive and Wellington did not do what was not required of them, they could not have done it if it had been required. If it be only meant to inform us that it is possible to be a great general without these things, so it is without many other things which are very useful to great generals. Alexander the Great had never heard of Vauban's rules, nor could Julius Cæsar speak French. We are next informed that book-worms, a term which seems to be held applicable to whoever has the smallest tincture of book-knowledge, may not be good at bodily exercises, or have the habits of gentlemen. This is a very common line of remark with dunces of condition; but, whatever the dunces may think, they have no monopoly of either gentlemanly habits or bodily activity. Wherever these are needed, let them be inquired into and separately provided for, not to the exclusion of mental qualifications, but in addition. Meanwhile, I am credibly informed that in the Military Academy at Woolwich the competition cadets are as superior to those admitted on the old system of nomination in these respects as in all others; that they learn even their drill more quickly, as indeed might be expected, for an intelligent person learns all things sooner than a stupid one; and that in general demeanor they contrast so favorably with their predecessors, that the authorities of the institutions are impatient for the day to arrive when the last remains of the old leaven shall have disappeared from the place. If this be so, and it is easy to ascertain whether it is so, it is to be hoped we shall soon have heard for the last time that ignorance is a better qualification than knowledge for the military, and, à fortiori, for every other profession, or that any one good quality, however little apparently connected with liberal education, is at all likely to be promoted by going without it.

Though the first admission to government employment be decided by competitive examination, it would in most cases be impossible that subsequent promotion should be so decided; and it seems proper that this should take place, as it usually does at present, on a mixed system of seniority and selection. Those whose duties are of a routine character should rise by seniority to the highest point to which duties merely of that description can carry them, while those to whom functions of particular trust, and requiring special capacity, are confided, should be selected from the body on the discretion of the chief of the office. And this selection will generally be made honestly by him if the original appointments take place by open competition, for under that system his establishment will generally consist of individuals to whom, but for the official connection, he would have been a stranger. If among them there be any in whom he, or his political friends and supporters, take an interest, it will be but occasionally, and only when to this advantage of connection is added, as far as the initiatory examination could test it, at least equality of real merit; and, except when there is a very strong motive to job these appointments, there is always a strong one to appoint the fittest person, being the one who gives to his chief the most useful assistance, saves him most trouble, and helps most to build up that reputation for good management of public business which necessarily and properly redound to the credit of the minister, however much the qualities to which it is immediately owing may be those of his subordinates.