AustLII Home | Databases | WorldLII | Search | Feedback

University of Technology Sydney Law Research Series

You are here:  AustLII >> Databases >> University of Technology Sydney Law Research Series >> 2016 >> [2016] UTSLRS 2

Database Search | Name Search | Recent Articles | Noteup | LawCite | Author Info | Download | Help

Opeskin, Brian --- "Judicial Exits: The Tenure of Judges in Three Apex Courts" [2016] UTSLRS 2; Judicial Independence in Australia (The Federation Press, 2016) Ch 7

Last Updated: 16 February 2017




Judicial tenure is an important facet of judicial independence and a key principle underpinning the rule of law. Robust provisions for tenure allow judges the freedom to decide cases according to law, without fearing reprisal through demotion or dismissal, or anticipating favour through promotion or re-appointment, by executive government. It has not always been so. In 17th Century England, judges were commonly transferred or dismissed after deciding cases in a manner that displeased the monarch who appointed them. Since then, systems of government have evolved to give much greater protection to judicial officers by enshrining principles of tenure in law and practice, although the evolutionary process has been uneven and some countries have a chequered history of judicial independence.

Tenure is not a binary concept that either does or does not exist. Provisions relating to tenure can be crafted in diverse ways that support a greater or lesser degree of independence. This paper examines three modes of judicial exit, namely: death, attaining a mandatory retirement age, and reaching the end of a fixed term appointment. These correspond to three models of tenure that are here called life limits, age limits and term limits. This chapter examines constitutional models of judicial tenure by investigating three apex courts that exemplify disparate practices — the Supreme Court of the United States in the case of life limits (‘the Supreme Court’); the High Court of Australia in the case of age limits (‘the High Court’); and the Constitutional Court of South Africa in the case of term limits (‘the Constitutional Court’). A novel feature of this chapter is that it provides a comparative empirical assessment of the tenure practices of the selected courts by examining changes over time in the judges’ age at appointment, age at termination and length of service.

An examination of models of judicial tenure is timely because there appears to be dissatisfaction with historical models and growing divergence in contemporary constitutional practice. While the importance of judicial independence has not diminished, the social context in which independence is to be protected has altered, necessitating a reappraisal of available tenure options.

This chapter is organised as follows. Part II examines the changing social context in which judicial tenure must now be assessed. Then, commencing with the historical antecedents of judicial tenure in England under the Stuart kings, Part III discusses the constitutional framework for regulating judicial tenure in three courts that exemplify the use of life limits, age limits and term limits. Part IV turns to the empirical record and considers how different models of tenure are reflected in the experience of the courts with regard to age at appointment, age at termination and length of service. Part V examines the tensions that inhere in different models of tenure through the lens of five dyadic relationships, namely: judicial independence versus accountability; constancy versus change in the composition of the bench; cost versus effectiveness; rigidity versus flexibility in the level of regulation; and the contrast between courts at different levels of the hierarchy or invested with different subject matter jurisdiction. Part VI concludes that the choice between models, and between key parameters within each model, reflect complex policy preferences. Hybrid arrangements that incorporate age limits and term limits — illustrated by the experience of the Constitutional Court of South Africa — provide an appropriate compromise between competing policy objectives.


The plurality of constitutional practice with respect to judicial tenure reflects the growing complexity of the social environment in which legal policy is formulated. While the relevance of judicial independence has not diminished, the social facts against which that principle operates have changed, as considered below.

A Increasing Longevity

The developed world has experienced dramatic improvements in human longevity since the Industrial Revolution. In 1701, when life tenure for judges was first adopted in England as a matter of regular constitutional practice, life expectancy was only 37 years.[1] Today in Australia the life expectancy at birth is 80.1 years for males and 84.3 years for females, and by 2060 this is projected to rise to as much as 92.1 years and 93.6 years, respectively, adding about a decade to the average life span.[2] These gains are attributable to an epidemiological transition in which humans have progressed in stages from an ‘age of pestilence and famine’ to an ‘age of degenerative and man-made diseases’.[3]

For the present study, a more significant statistic is life expectancy at the average age of judicial appointment. In 2011–2013, an Australian judge appointed at age 54 (typical of the High Court) could expect a total life span of 82.6 years for males and 86.0 years for females.[4] These figures exceed life expectancy at birth because a person who reaches the age of judicial appointment has already survived the vicissitudes of life for many decades, including the first year of life, which is the most perilous.

There are many instances of judges serving well into old age, leading Judge Richard Posner to the colourful observation that the United States judiciary is ‘the nation’s premier geriatric occupation’.[5] For example, in the United States Federal District Court, Judge Wesley Brown died in office in 2012 at age 104, having served nearly 50 years on the bench, making him both the longest serving and oldest federal judge still hearing cases at that date. In Australia too there are examples of judges serving well into their eighties — records that were set in the era, now passed, in which appointment was for life.[6]

Declining mortality has the potential to impact all models of judicial tenure. Judges appointed for life may serve substantially longer terms, judges appointed to a fixed age are likely to have longer periods in retirement, and judges appointed for fixed terms are likely to have longer periods of their productive working lives in non-judicial roles.

B Mental Decrepitude

One of the sequelae of increased life expectancy is that the frailties of human ageing are exposed to public view in circumstances that were often masked when life was ‘brutish and short’.[7] Notably, an ageing bench may become susceptible to incapacitation through senescence. In a pioneering study of ‘mental decrepitude’ on the United States Supreme Court, David Garrow concluded that ‘historical evidence convincingly demonstrates that mental decrepitude among aging justices is a persistently recurring problem that merits serious attention’.[8] Noting that the 20th Century featured 11 Supreme Court justices whose declining mental capacity should have led to earlier departure, Garrow advocated a constitutional amendment to impose mandatory retirement for justices at age 75. Without that barrier, judges might continue to adjudicate into old age because, as Posner has remarked, ‘judging is light work [and] senility is virtually the only condition short of death that disables a judge from performing at a satisfactory although not necessarily distinguished level’.[9]

There are documented examples in Australia as well. One concerns Australia’s youngest appointee to the High Court, H V Evatt, who left that office in 1930 at the age of 46 to pursue a career in national politics. Twenty years later, when he was ‘far removed from the practice of law and already showing signs of mental deterioration’, he was made Chief Justice of New South Wales.[10] However, he was unable to function at the most basic level — he was unfocussed, had no grasp of the cases at hand, and relied on others to write his judgments. Ultimately, he was encouraged to resign and spent his remaining years in a ‘regressed state under the care of his wife and a nurse’.[11]

The manner in which age affects mental aptitude has a bearing on the merits of life limits and age limits as models of judicial tenure. The prospect of mental incapacity raises concerns about life tenure and provides a yardstick for determining an appropriate retirement age if mandatory retirement is adopted.

C Age Discrimination

The third social change is the shift in public attitudes towards ageing and age-based discrimination. Term limits are an age-neutral model of tenure because the duration of a judge’s appointment is unrelated to his or her chronological age. Life tenure is a pro-ageing model of tenure because it implicitly assumes judges have full capacity to discharge their functions until their death or earlier voluntary retirement, even if this entails service into extreme old age. By contrast, age limits embody an ageist conception of tenure because judges cease to hold office on reaching the mandatory retirement age regardless of their individual capacities. This presumption of ‘statutory senility’, as mandatory retirement is sometimes called, is the price paid for avoiding the risk of judicial over-stayers.[12] It is this model of tenure that most significantly challenges societal views of age discrimination.

The prohibition of age discrimination in many industrialised states has been driven in large measure by anxieties about their macro economies. Many countries now recognise that their populations are ageing inexorably due to declining fertility and increasing longevity. As the proportion of the older population continues to rise, many economies will face declining labour force participation rates and reduced growth unless they can retain the skills of older workers.[13] There is a broad consensus that ‘policies that remove barriers to employment and enhance the productivity of older men and women are an essential part of any effective response to population aging’.[14]

The upshot of these concerns is that mandatory retirement has been all but abolished for most public and private employees. Should judges be subject to mandatory retirement laws that do not apply to the public at large? It must be asked whether the state’s interest in ensuring the proper administration of justice through the retention of only the most able judges justifies a mandatory retirement rule whose application may be overly broad in particular instances.


Early English experience of judicial tenure was based on a model in which judges held office at the discretion of the executive (quam diu nobis placuerit — at the King’s pleasure) and could be dismissed at whim. This was the antithesis of judicial independence and bore the institutional risks of bias in decision-making, erosion of public confidence in the judiciary, and a tenuous grip on office. The worst excesses of monarchical power occurred under the Stuart kings, whose transfer and removal of troublesome judges made some ‘pretty black pages of history’.[15] In the last 11 years of his reign, Charles II (1660–1685) dismissed 11 of his judges, while his brother James II (1685–1688) dismissed 13 judges in four years, including four in one day.[16]

Pressure for reform mounted, and by the start of the 18th Century the discretionary model gave way. The Act of Settlement 1701 limited the royal prerogative and established the regular practice of appointing judges during good behaviour (quam diu se bene gesserit), which in practice meant appointment for life, unless the judge ceased to be of good behaviour. Life tenure provides a very high degree of independence because a judge is largely placed beyond the reach of the executive’s opprobrium from the moment of appointment.

A Til Death Us Do Part: Life Tenure in the United States

Life tenure became the standard practice for judicial appointments in England after the Act of Settlement, and a model for other legal systems that looked to English constitutional history for guidance. Although that model has not endured in the United Kingdom itself,[17] the influence of the Act of Settlement on the common law world cannot be gainsaid. Thus it was that art III § 1 of the United States Constitution provided that federal judges ‘shall hold their Offices during good Behaviour’, while art II § 4 made provision for judges to be removed from office only by impeachment for ‘Treason, Bribery, or other high Crimes and Misdemeanors’.

Writing in support of art III § 1 during the ratification period, Alexander Hamilton advanced several reasons for preferring life tenure over ‘temporary commissions’.[18] Permanency in office not only comported with the best practice of state constitutions of the day and the ‘illustrious’ experience of Great Britain, but also contributed so much to ‘that independent spirit in the judges which must be essential to the faithful performance of so arduous a duty’.[19] Hamilton was particularly ardent in his criticism of age limits for federal judges. Rejecting the approach in the New York Constitution of taking a particular age (60 years) as the ‘criterion of inability’, he argued that a judge’s ‘deliberating faculties’ preserved their strength well beyond that age, and that very few men outlived ‘the season of intellectual vigor’.[20] The life tenure model that was adopted for federal judges continues to this day, although it is no longer widely embraced at the state level.

Yet some of the difficulties inherent in the life tenure model were masked by the short life expectancy of appointees in those early times. A life tenure judge who is disinclined to retire may hold office into extreme old age and this too may adversely affect public confidence in the judiciary if the judge lacks capacity to decide cases according to law. In the 20th Century, polities began to experiment with other models of tenure, including mandatory retirement at a specified age.

B ‘Statutory Senility’: Age Limits in Australia

The second model of tenure is one in which judicial office expires when the judge attains a specified age. It is thought that New Zealand was the first jurisdiction in the British Empire to introduce an age limit for judges (72 years) in 1903, followed by New South Wales (70 years) in 1918.[21] Age limits now have a substantial pedigree and are used in many countries at all levels of the judicial hierarchy. This section examines their adoption in the Australian Constitution, which regulates the federal judiciary, and compares this with the practice in the constituent states.

The Australian Constitution was modelled closely on the United States Constitution. The key provision is s 72, which contemplated only one mode of termination. It provided that federal judges ‘shall not be removed except by the Governor-General in Council, on an address from both Houses of the Parliament in the same session, praying for such removal on the ground of proved misbehaviour or incapacity’. However, s 72 failed to specify the tenure of federal judges — it made no express mention of holding office during good behaviour but only provided for removal. This uncertainty was not resolved until 1918 when the High Court held in Alexander’s Case that a judicial appointment for a fixed term was invalid because every judge of a court created by the Parliament is required to be appointed for life, subject only to the power of removal on the grounds of ‘proved misbehaviour or incapacity’.[22]

The original text of s 72 remained the basis of judicial tenure for federal judges for the next 60 years. However, in 1977 a Senate Committee conducted an inquiry into the tenure of federal judges,[23] prompted by parliamentarians who were shocked at being sworn into office by the aged and feeble Acting Chief Justice, Sir Edward McTiernan.[24] The Committee’s recommendations became the basis of a constitutional referendum to introduce a mandatory retirement age for federal judges. For justices of the High Court this was to be 70 years of age. For judges of other federal courts, 70 years was to be the maximum age but Parliament could set a lower limit if it chose to do so. Faced with bipartisan support for the referendum proposal, voters gave it their overwhelming endorsement — the referendum passed in all states and was approved by over 80 per cent of the population.

The Government’s rationale for proposing the constitutional amendment can be seen in the second reading speech made when introducing the Bill.[25] In the opinion of the Attorney-General there was an almost universal practice that holders of public office retire on attaining a maximum retirement age.[26] He noted that a fixed retirement age had been adopted in all State Supreme Courts and that it was appropriate to make similar provision for the growing number of federal judges.[27] Perhaps more revealing are the comments made during the parliamentary debate, namely, that judges are not immune from the geriatric processes of mental decay, and that the proposed age limit would lead to a younger body of judges who are ‘closer to the people’ and have ‘current day sets of values’.[28]

The notion that there should be a mandatory retirement age for federal judges appears to have been quickly accepted but there was virtually no discussion of the appropriateness of selecting 70 years as the maximum age. Even the report of the Senate Committee said little on the subject, noting only that 70 was the ‘retiring age most commonly established for judges of State and territory Supreme Courts’.[29] The desire for uniformity between state and federal practice was conspicuous. New South Wales was the first Australian state to introduce a mandatory retirement age for judges of superior courts when it legislated for retirement at 70 years during the First World War.[30] The choice proved influential. Today, 70 years remains the mandatory retirement age for judicial officers in all States except in New South Wales and Tasmania, where it has since been increased to 72 years,[31] and in magistrates’ courts in Western Australia and the Australian Capital Territory, where it is 65 years.[32]

C Term Limits in South Africa

A third model of tenure is one in which a judge’s term has a fixed duration and thus comes to an end by effluxion of time. This section focuses on the evolution of term limits in the Constitutional Court in South Africa, a new court that was established to adjudicate sensitive constitutional questions in the post-apartheid era.[33]

South Africa’s 1993 interim Constitution[34] established a Constitutional Court comprising a President and ten other members, and invested it with jurisdiction over all matters relating to the interpretation, protection and enforcement of the Constitution. The members of the Court were to hold office for a non-renewable period of seven years, which would take their service beyond the life of the Interim Constitution and the first Parliament.[35] The limited terms were thought appropriate for judges of a court that was to exercise substantial political authority through its constitutional mandate.[36]

Important changes were made to tenure arrangements by the adoption of the final Constitution in 1996.[37] Under s 176(1), a Constitutional Court judge was to be appointed ‘for a non-renewable term of 12 years, but must retire at the age of 70’. The two new features — longer fixed terms coupled with mandatory retirement at age 70 — were modelled on Germany’s Federal Constitutional Court whose judges hold office for 12 years without the possibility of re-election, but subject to mandatory retirement at age 68.[38] Reflecting on the rationale for term limits, Kate O’Regan, a judge of the South African Constitutional Court from 1994 to 2009, remarked that ‘[i]t is unfortunate if a senior court turns over in membership too quickly, but it is doubly unfortunate if it turns over too slowly. Our (effective) 12-year term limit steers a course between these evils.’[39]

The tinkering with the tenure arrangements did not end with the passage of the 1996 Constitution. In 2001, s 176(1) of the Constitution was amended to give a new role to the legislature.[40] The subsection now reads: ‘A Constitutional Court judge holds office for a non-renewable term of 12 years, or until he or she attains the age of 70, whichever occurs first, except where an Act of Parliament extends the term of office of a Constitutional Court judge.’ The change effected by the final clause is the addition of a legislative power to extend judges’ terms by relaxing either the 12-year term limit or the 70-year age limit, or both.

Parliament wasted no time in exercising its new power. In 2001 it altered both the term limit and the age limit, albeit in complex ways.[41] The 2001 Act affirms the 12 year term limit and 70 year age limit as alternative criteria for departure from the Court, but there are some circumstances in which a judge may be required to serve on the Constitutional Court for 15 years or to age 75.[42]


The optimal design of constitutional provisions and cognate legislation regarding judicial tenure requires a clear understanding of how the contrasting models operate in practice. Unfortunately, little attention has been paid to empirical evidence outside the United States, leaving many key questions unanswered. This Part seeks to fill that gap by examining data on the three apex courts. Although the data required for the analysis are undemanding, they are surprisingly difficult to procure.

For the purpose of comparative analysis, this chapter utilises the methodology employed by Steven Calabresi and James Lindgren in their path-breaking study of tenure on the United States Supreme Court.[43] However, it updates the analysis to 30 June 2013, incorporates more detailed data on age at appointment, and adopts an improved graphical representation that allows for inter-jurisdictional comparisons.

Calabresi and Lindgren found that there have been significant changes in the practical meaning of life tenure for justices of the Supreme Court, as evidenced in their length of service, age at termination, and the interval between vacancies. Comparing the 30 year period 1941–1970 with 1971–2006, the authors found that the justices’ mean length of service had more than doubled to 26.1 years (compared with 12.2 years in the earlier period), the mean age at which justices left office had risen to 78.7 years (compared with 67.6 years), and the frequency of vacancies had declined from an average of one every 1.6 years to one every 3.1 years.

In their view, there were three vices in the emerging patterns of tenure. They reduced democratic control over the Supreme Court by making appointments infrequent and irregular, thus limiting ‘the democratic instillation of public values on the Court through the selection of new judges’.[44] They made the Senate confirmation process more political, to the point of dysfunction, because the irregular occurrence of vacancies made the stakes so high. And they resulted in an increasing prevalence of mental decrepitude on the Court as justices stayed for longer periods and to more advanced ages. The authors considered these problems sufficiently serious to warrant a constitutional amendment abandoning life tenure in favour of term limits.

What does the empirical record reveal about life limits, age limits and term limits in the three apex courts examined in this study? Table 1 presents summary data, while Figure 1 illustrates the data graphically, focussing on the period 1970-2013.

Table 1: Summary Statistics: United States, Australia and South Africa[45]

United States
South Africa

Supreme Court
High Court
Constitutional Court
Current Membership

32 yrs
36.6 yrs
36.8 yrs
65 yrs
61.8 yrs
64.0 yrs
Mean age
52.5 yrs
54.1 yrs
54.4 yrs

47 yrs
46.3 yrs
52.1 yrs
90 yrs
87.0 yrs
74.4 yrs
Mean Age
69.6 yrs
69.4 yrs
66.5 yrs
Length of Service

1 yr
0.1 yrs
4.0 yrs
37 yrs
45.7 yrs
15.3 yrs
Mean duration
17.0 yrs
15.9 yrs
11.5 yrs
Number (completed service)

A Mean Age at Appointment

The first point of comparison is that there is marked similarity in the mean age at appointment, and in the age range of appointments, to the three courts (Figure 1A). For appointments made in the period 1970–2013, the mean age is 53.1 (United States), 54.9 (Australia) and 54.4 years (South Africa). There has also been considerable stability in the age at appointment over time, except in Australia, where the mean age has risen since 1995 from 51.4 years to 58.3 years. This reflects the recent tendency in Australia to appoint individuals who have had substantial judicial experience on other courts, and who are necessarily older when appointed to the High Court.

B Mean Age at Termination

In contrast, there are significant disparities in the mean age at termination (Figure 1B). The Supreme Court and the High Court had similar experience in the period 1970–1985, when the mean age at termination hovered at around 70 years. For the High Court, the mean age at termination then declined in the period 1985–2010. This was not due to the introduction of mandatory retirement but to idiosyncratic factors affecting individual retirees, such as ill health and appointment to other offices. The mean age has since edged back towards 70 years because the six terminations prior to 30 June 2013 have all occurred at the mandatory retirement age. In the Supreme Court, however, there has been a relentless rise in the mean age at termination, which is now over 80 years — a full 11.5 years greater than the High Court. There is no ceiling to that rise other than the limits of human longevity itself. Interestingly, the rise in tenure has closely tracked improvements in longevity — between 1970 and 2010 the mean age at termination on the Supreme Court rose by 19 per cent, while life expectancy at age 50 (both sexes, all races) rose by 21.7 per cent.[46] In the Constitutional Court the mean age at termination was initially around 70 years (2004–2009) but has since fallen sharply to 61.0 years.

C Mean Length of Service

The patterns in mean length of service reflect the combined effect of changes in the mean age at appointment and the mean age at termination (Figure 1C). In the Supreme Court, the mean length of service has been 25 years or more every year since 1993. In the High Court, length of service has rarely exceeded 15 years over the same period, so that over the past 20 years (1994–2013) Supreme Court tenure has exceeded High Court tenure by an average of 11.3 years. In the High Court, the mean length of service has edged downwards in recent years because the mean age at appointment has risen while facing a fixed age limit of 70 years. In the Constitutional Court, the mean length of service has risen to just under 15 years, reflecting the rising term limits for some judges, discussed above.

Figure 1: Appointment, termination and length of service in three apex courts, 1970–2013[47]



There are complex policy choices at stake in crafting arrangements for judicial tenure in modern constitutional democracies. This Part examines some of the central tensions, across five dyadic relationships.

A Independence and Accountability

Life limits, age limits and terms limits are different solutions to the problem of executive interference in the exercise of judicial functions, such as the English experienced during the reign of the Stuart kings. Although external interference is not the only threat to be guarded against, this type of decisional independence has become a value of the highest order in assessing models of tenure. Life tenure provides the greatest bulwark against executive interference but the other models do not seriously compromise that goal so long as the age limit is not too young nor the term limit too short (requiring departing judges to seek post-judicial professional employment).

In the case of term limits, it is also important that judicial terms are non-renewable so that judges have nothing to gain or lose by deciding cases for or against the government of the day. Judges are very much alive to such threats to their independence, as can be seen in South Africa’s Justice Alliance Case, in which the Constitutional Court invalidated a legislative attempt to extend the term of the Chief Justice upon the request of the President of South Africa.[48]

Weighed against the demands of judicial independence is the need for judicial accountability, and the notion that ‘public officials who are entrusted with power should be held accountable to the nation, and must be fit persons to exercise these powers.’[49] The problem was posed during the United States ratification debates when Robert Yates, writing as Brutus in the Anti-Federalist Papers, proffered the vatic observation that life tenure would make judges ‘soon feel themselves independent of heaven itself’.[50] In the present context, the problem is at its starkest when a life tenure judge is incapable of discharging his or her duties by reason of mental decrepitude but nonetheless refuses to retire. The usual mechanisms of judicial accountability — conduct of proceedings in open court, delivery of reasons for judgment, and appellate review — would ordinarily expose such judges to adverse public scrutiny and might nudge them towards retirement, but it is no guarantee in the face of a judge’s intransigence.

B Constancy and Change: The Composition of the Bench

A second tension in the models of judicial tenure is the balance between constancy and change in the composition of the bench. Joseph Raz has argued that stability is an important principle underpinning the rule of law because it allows people to know the law for the purpose of short-term decision-making and long-term planning.[51] It is a short step to conclude that stability in the composition of a court promotes stability in its jurisprudence, especially in appellate courts that have a clear law-making function. This is not to say that long-serving judges do not change their opinions over time, but seismic shifts in legal principles are less likely when the personnel interpreting and developing them are not subject to frequent change.

Pitted against the need for stability on the bench is the need for change to ensure that evolving social values are broadly reflected in decisions of the courts. As Justice Michael Kirby once observed, the rules relating to tenure should provide for ‘regular and seemly exits’ so that the final court can be ‘dynamic and open-minded’.[52] The argument is not that judges should be a simulacrum of the people on whose behalf they interpret and apply the law but that frequent turnover ensures that judges are not too far out of kilter with the ethos of the community they serve. For courts at the highest level, the policy preferences of judges should be refreshed regularly for reasons of good governance and democratic legitimacy.

The goal of stability favours life limits, reasonably long term limits, or age limits that are not too young. In contrast, the goals of change and renewal are best served by shorter term limits and younger age limits, and are the antithesis of life tenure, which permits the development of a geriatric bench.

C Cost and Effectiveness

A third but often unarticulated tension exists between the economic cost of different models of tenure and their effectiveness in achieving the objectives of the judicial system. The issue of effectiveness has been canvassed above — the critical question is whether the tenure model, which affects the age of judges and their length of service, adversely impacts on the adjudicatory process. It could do so if aged or long-serving judges are less able to resolve disputes in a timely manner, or produce decisions of lower quality, or are more prone to error. Judicial error imposes burdens on the parties and the legal system through the instigation of avoidable appeals or the failure to resolve disputes according to law, with a consequent erosion of the rule of law.

It is the economic cost of different tenure models that has gone most unremarked. The appointment of judges imposes direct costs to public finance through salaries, pensions, personal staff and facilities, and these costs vary from model to model. Different tenure arrangements create complex trade-offs between different cost components. These can be difficult to quantify because legislatures tend to bias judicial compensation towards non-salary benefits, such as pensions, which are harder to value and less likely to provoke public discontent.[53] It is generally the case, however, that models of tenure that encourage greater turnover entail higher public cost due to the emoluments that adhere to each new appointee.

The tension between economic cost and effectiveness can be elucidated by considering the consequences of hypothetically increasing the age limit for federal judges in Australia. This change would permit judges to retire later than currently required. Every extra year of judicial service requires the state to pay more in salary than it saves in pension, but the marginal cost of that option is less than the cost of appointing a new judge to provide equivalent labour, since a new appointee generates no pension saving. On the other hand, each extra year of service increases the likelihood of falling productivity or rising decrepitude, which impacts on an ageing judge’s effectiveness. Although the trade-offs are unique to the retirement and remuneration arrangements in each jurisdiction, the general principle is well understood by executive government — longer serving judges save direct public expenditure but at the risk of diminishing effectiveness.

D Rigidity and Flexibility: The Level of Regulation

A fourth tension is the respective roles of the constitution and legislation in regulating judicial tenure. Comparative practice suggests that constitutions need not contain every provision that affects judicial tenure but nor should they remain silent. The quest is to find an appropriate balance between rigidity and flexibility. The tipping point will depend, among other things, on the ease with which the constitution can be amended in theory and practice. The Australian and South African experience show that constitutions falter when they over-specify and thus entrench policy choices that are impervious to changing social circumstances.

The rules most suitable for constitutional inclusion are those that go to foundational architecture (such as the choice between alternative models of tenure) and those that seek to protect judicial independence from legislative or executive encroachment. Within a socially acceptable range, finer details can be left to statute-makers. For example, if term limits are the chosen architecture, the constitution should specify that terms are non-renewable, and stipulate a minimum term or a span of years within which the legislature may make choices. If age limits are the chosen architecture, the constitution should specify a minimum age for appointment and a maximum age for retirement, subject to legislative extension.

E The High and the Low: Differentiating between Courts

A final tension is whether the same tenure model should apply to all courts within a country or whether justifiable distinctions can be drawn between courts. This chapter has focussed on judicial tenure in three pre-eminent appellate courts but tenure arrangements are not uniform within those countries.

One justification for differentiated approaches relates to subject matter. In Australia there was a period (1977–1991) in which judges of the Family Court faced a mandatory retirement age of 65 years although the age limit for other federal judges was 70.[54] The difference was explained by the ‘demanding and arduous nature’ of custody disputes and the need for judges in this field to ‘keep abreast of current social values and attitudes’.[55] Ultimately these considerations yielded to another concern, namely, the desire to improve the status of the Family Court in the eyes of the public by giving its judges the same tenure as other federal judges, and parity in retirement age was duly restored.

A second justification for differentiation relates to the political function of the court. The most delicate tensions between the branches of government often fall to be decided in an apex court or constitutional court, and the politically charged questions they adjudicate may call for unique measures. In the United States, these considerations led Calabresi and Lindgren to propose that the life tenure of Supreme Court justices — but not other federal judges — be replaced by a non-renewable 18-year term, staggered in such a way that each President would be able to nominate two justices during a four-year presidential term.[56]

A third justification relates to the level of the judicial officer within the court hierarchy. In several Australian states, magistrates face younger mandatory retirement than judges appointed to courts above them — typically 65 years compared to 70 years.[57] Pragmatic reasons for the difference include younger appointment to the magistracy and heavy daily caseloads that might be thought too onerous for ageing judicial officers.


If the competing values at stake in different models of judicial tenure were simple to reconcile it is reasonable to suppose that mature constitutional democracies might have arrived at a stable solution by now — even if that solution is not common to all legal systems because of the differing weights attached to competing considerations. Instead, the three countries examined here show a flux in legal thinking, revealing dynamic tension between alternative models of tenure and between key parameters within the models.

Examined over the long run, there has been dissatisfaction with judicial tenure arrangements in each country, prompting reform or proposals for reform. In the United States, this is reflected in legislative palliatives relating to judicial pensions and senior status for life tenure judges, as well as in proposals made in the 1930s, 1940s and 1950s to introduce mandatory retirement for federal judges[58] and more recently to introduce fixed terms for the Supreme Court. In Australia, dissatisfaction is evident in the rejection of life tenure for federal judges in favour of age limits in the 1970s, coupled with experiments in lowering the retirement age and more recent proposals to increase it. In South Africa, dissatisfaction is apparent in the steps taken to increase the duration of term limits of the Constitutional Court and in the adoption of a hybrid model that combines term limits with age limits.

It is unsurprising that life tenure — adopted with such confidence in the United States in 1789 as an antidote to the English experience of royal caprice — should now be viewed with greater scepticism. Although life tenure provides the sturdiest protection against executive interference, contemporary experience of increasing longevity and concomitant exposure to mental frailty in old age suggests that such an elevated degree of judicial independence comes at too high a cost.

On the other hand, it is hard to see how the independence of the judiciary is adversely affected by judges having to retire at a fixed age, or after a fixed period, if the parameters of those models are intelligently set. Whether the appropriate age limit is 70 or 75 years, or the term limit 10 or 15 years, is a matter of detailed policy about which polities may come to different conclusions, from each other and over time, depending on the social facts underpinning their political communities. As the Judicial Committee of the United States Senate observed many years ago, ‘the age at which retirement is compulsory is not nearly so important as the proposition that there be an age at which retirement is compulsory’.[59] What is elementary is that the constitutional design preserves core features of judicial independence while giving flexibility to modify other features as changing circumstances require.

In this context, South Africa’s hybrid model is an inventive solution that guards against some of the weaknesses of individual models of tenure and offers a potential template for other jurisdictions. The age limit controls for the risk of decrepitude among elderly judges, while the term limit ensures there are ‘regular and seemly exits’ that generate change and renewal on the court. Moreover, these objectives are achieved through a mixture of constitutional and legislative regulation, which provides rigidity in some dimensions and flexibility in others. However, the current arrangements have been in place for the Constitutional Court only since 2001 and it is too early to make a full empirical assessment of their impact. In time, it may be a model well worth deliberating in other apex courts.

[*] Professor of Legal Governance, Macquarie University, Sydney. This chapter is based on an article published in (2015) 35 Oxford Journal of Legal Studies 627, where the topic is addressed in greater depth.

[1] Edward Wrigley and Roger Schofield, The Population History of England, 1541-1871: A Reconstruction (Cambridge University Press, 1981) 230.

[2] Australian Bureau of Statistics, 'Life Tables, States, Territories and Australia, 2011-2013, Cat. No. 3302.0.55.001' (Australian Bureau of Statistics, 2014) Table 1.9; Australian Bureau of Statistics, 'Population Projections, Australia, 2012 (Base) to 2101 Cat. No. 3222.0' (Australian Bureau of Statistics, 2013) 9.

[3] Abdel Omran, 'The Epidemiological Transition: A Theory of the Epidemiology of Population Change' (1971) 49 Milbank Memorial Fund Quarterly 509.

[4] Australian Bureau of Statistics, above n 2.

[5] Richard Posner, Aging and Old Age (University of Chicago Press, 1995) 180.

[6] Keith Mason, Lawyers Then and Now: An Australian Legal Miscellany (Federation Press, 2012) 86–8.

[7] Thomas Hobbes, Leviathan (Penguin Classics, first published 1651, 1985 ed) ch XIII, 186.

[8] David Garrow, 'Mental Decrepitude on the US Supreme Court: The Historical Case for a 28th Amendment' (2000) 67 University of Chicago Law Review 995, 995.

[9] Posner, above n 6, 352.

[10] Mason, above n 7, 63.

[11] Andrew Campbell, 'Dr H V Evatt - Part One: A Question of Sanity' (2007) 73 National Observer 25, 39.

[12] See Frederick Schauer, Profiles, Probabilities, and Stereotypes (Belknap Press, 2003) 108–30, on the similar question of mandatory retirement for pilots at age 60.

[13] Australian Productivity Commission, Australian Government, Economic Implications of an Ageing Australia (2005).

[14] Andrew Mason and Ronald Lee, 'Population Aging and the Generational Economy: Key Findings' in Ronald Lee and Andrew Mason (eds), Population Aging and the Generational Economy: A Global Perspective (Edward Elgar, 2011) 3, 27.

[15] C H McIlwain, 'The Tenure of English Judges' (1913) 7(2) American Political Science Review 217, 222.

[16] Sir Henry Brooke, 'Judicial Independence: Its History in England and Wales' in Helen Cunningham (ed), Fragile Bastion: Judicial Independence in the Nineties and Beyond (Judicial Commission of New South Wales, 1997) 89, 97; McIlwain, above n 16, 223–4.

[17] Judicial Pensions Act 1959 (UK) s 2; Judicial Pensions and Retirement Act 1993 (UK) s 26.

[18] Alexander Hamilton, 'The Federalist No. 78' in Robert Scigliano (ed), The Federalist: A Commentary on the Constitution of the United States (Random House Publishing Group, 2010) 495.

[19] Ibid 500.

[20] Alexander Hamilton, 'The Federalist No. 79' in Robert Scigliano (ed), The Federalist: A Commentary on the Constitution of the United States (Random House Publishing Group, 2010) 504.

[21] Supreme Court Judges Act 1903 (NZ) s 2; Judges Retirement Act 1918 (NSW) s 3; Tony Cunneen, 'A Creature of a Momentary Panic' [2010] (Winter) Bar News 74, 74.

[22] Waterside Workers' Federation of Australia v J W Alexander Ltd [1918] HCA 56; (1918) 25 CLR 434.

[23] Commonwealth, Report on Retiring Age for Commonwealth Judges, Parl Paper No 414 (1976).

[24] Michael Kirby, 'Sir Edward McTiernan: A Centenary Reflection' (1991) 20 Federal Law Review 165, 181.

[25] Constitution Alteration (Retirement of Judges) Bill 1977 (Cth).

[26] Commonwealth, Parliamentary Debates, House of Representatives, 16 February 1977, 147 (Robert Ellicott, Attorney-General).

[27] Ibid.

[28] Commonwealth, Parliamentary Debates, House of Representatives, 17 February 1977, 203 (Ian Sinclair, Minister for Primary Industry), 206 (Lionel Bowen), 209 (Michael Hodgman), 220 (Maurice Neil).

[29] Commonwealth, above n 24, [50].

[30] Judges Retirement Act 1918 (NSW) s 3. See also Cunneen, above n 22.

[31] Judicial Officers Act 1986 (NSW) s 44(1); Supreme Court Act 1887 (Tas) s 6A.

[32] Magistrates Court Act 2004 (WA) sch 1, cl 11(1)(a); Magistrates Court Act 1930 (ACT) s 7D.

[33] Hugh Corder and Jason Brickhill, ‘The Constitutional Court' in Cora Hoexter and Morné Olivier (eds), The Judiciary in South Africa (Juta, 2014) 355, 357–62.

[34] Consitution of the Republic of South Africa 1993 (South Africa).

[35] Constitution of the Republic of South Africa 1993 (South Africa) ss 97(2), 99(1).

[36] Hugh Corder, 'Judicial Accountability' in Cora Hoexter and Morné Olivier (eds), The Judiciary in South Africa (Juta, 2014) 200, 210.

[37] Consitution of the Republic of South Africa 1996 (South Africa).

[38] Bundesverfassungsgericht [Federal Constitutional Court Act] (Germany) 12 March 1951, BGB1 I, 1951, 243, 4.

[39] Kate O'Regan, 'The Constitutional Court: A Judge's Perspective' in Cora Hoexter and Morné Olivier (eds), The Judiciary in South Africa (Juta, 2014) 403, 412.

[40] Constitution Sixth Amendment Act 2001 (South Africa).

[41] Judges' Remuneration and Conditions of Employment Act 2001 (South Africa) ss 3, 4.

[42] Judges' Remuneration and Conditions of Employment Act 2001 (South Africa) ss 4(1), (2), (4): A Constitutional Court judge whose 12-year term of office as a Constitutional Court judge expires before they have completed 15 years’ active service must continue to act as a Constitutional Court judge until they have completed 15 years’ active service or attained the age of 75 years, whichever occurs first. A non-Constitutional Court judge in the same predicament may choose to follow this same course of action.

[43] Steven Calabresi and James Lindgren, 'Term Limits for the Supreme Court: Life Tenure Reconsidered' (2006) 29(3) Harvard Journal of Law & Public Policy 769.

[44] ibid 811.

  1. [45] United States: Lee Epstein et al, The U.S. Supreme Court Justices Database (2 March 2013) University of Southern California <> . Australia: Michael Coper, Tony Blackshield and George Williams (eds), Oxford Companion to the High Court of Australia (Oxford University Press, 2001); ConnectWeb, Who’s Who in Australia (Crown Content, 2013); High Court of Australia, <>. South Africa: Constitutional Court of South Africa, <>; Rapule Tabane and Barbara Ludman (eds), The Mail & Guardian A-Z of South African Politics (Jacana Media, 2009).

[46] Human Mortality Database, United States of America Life Tables (1x1) (2013) <>

[47] Data sources are those listed at the foot of Table 1.

[48] Judicial Alliance of South Africa v President of the Republic of South Africa [2011] ZACC 23; [2011] 5 SA 388 (Constitutional Court).

[49] A A Paterson, 'The Infirm Judge' (1974) 1(1) British Journal of Law and Society 83, 87.

[50] Charles Cooper, 'Federalist Society Symposium: Term Limits for Judges?' (1997) 13 Journal of Law and Politics 669, 672–4.

[51] Joseph Raz, The Authority of Law: Essays on Law and Morality (Oxford University Press, 2nd ed, 2009), 214–5.

[52] Michael Kirby, 'Neville Wran - A Lawyer Politician: Reflections on Law Reform and the High Court of Australia' (Speech delivered at the inaugural Neville Wran Lecture, Sydney, Parliament House, 13 November 2008) 39–40.

[53] Posner, above n 6, 353.

[54] Family Law Amendment Act 1977 (Cth) s 4; reversed by Family Law Amendment Act (No 2) 1991 (Cth) s 3.

[55] Commonwealth, Parliamentary Debates, Senate, 18 August 1977, 147 (Peter Durack, Minister for Veterans Affairs).

[56] Calabresi and Lindgren, above n 44, 824–54.

[57] Kathy Mack and Sharyn Roach Anleu, 'The Security of Tenure of Australian Magistrates' [2006] MelbULawRw 13; (2006) 30 Melbourne University Law Review 370, 386.

[58] See Garrow, above n 9, 1018–26, 1028–43, discussing the reform campaigns in 1937 and 1946–1955.

[59] U.S. Senate, Judicial Committee. Composition and Jurisdiction of the Supreme Court (S. Rpt. 83-1091) Washington: Government Printing Office, 1954, cited in Garrow, above n 9, 1037.

AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback