Advocates’ immunity: intimate connection and finality

Yesterday, on 4 May 2016, the High Court delivered a judgment that should be of great interest to every lawyer working in litigation: Attwells v Jackson Lalic Lawyers Pty Ltd [2016] HCA 16.


Lawyers acting in litigation enjoy a number of privileges in their work. One of them is the right to make allegations in court without fear of being sued for defamation. Another is not to be accountable to clients when their work is carried out negligently – although, as we will see, not all aspects of their work enjoy immunity from suit.

For obvious reasons, many people in the community think this is bad law. Comparison is often made with the work of surgeons. A surgeon working under pressure and in high-risk areas of the body may make a mistake, and of course is liable to be sued. Whether in all the circumstances the mistake is negligent is not to the point; the point is that the aggrieved patient (or someone else on their behalf) may sue. Sure, there’s insurance, but there are also reputational consequences, not to mention the chilling effect on the willingness to operate on certain patients.

At another level, the fact that there is no statute that provides for advocates’ immunity, and that it is purely a matter of judge-made law, gives the impression in the minds of some that the immunity is the product of lawyers looking after their own at the expense of their clients. Of course, it is open to Parliament to change judge-made law. High Court authority supporting the immunity has been with us for nearly 30 years, and there is no sign of any parliament doing anything to change or abolish the immunity.

Attwells’ case

In Attwells, the High Court rejected an argument that the immunity should be abolished, clarified the general scope of the immunity and, on the narrow issue squarely raised by that case, held that the immunity does not extend to “negligent advice which leads to the settlement of a case by agreement between the parties” ([4]), or put differently, “acts or advice of the advocate which do not move litigation towards a determination by a court” ([38]), or “negligent advice which leads to the settlement of a claim in civil proceedings” ([45]).

Mr Attwells was a guarantor for a company’s liability to the ANZ Bank. The ANZ’s counsel said in opening that Mr Attwells’ liability pursuant to the guarantee was limited to a little over $1.8 million, in contrast to the company’s liability of almost $3.4 million. Before the trial proceeded to evidence, the parties settled. The terms of settlement resulted in consent orders being entered to give effect to them. The consent orders were that there be judgment against Mr Attwells for $3.4 million, but it was noted (not ordered) that the ANZ had agreed not to enforce that order if Mr Attwells paid the ANZ $1.75 million by 19 November 2010. His solicitor (negligently) advised (and it seems that this was an agreed fact in the proceeding against the solicitor) that it did not make any difference that Mr Attwells was consenting to judgment against him for that larger sum, which was far in excess of what his maximum liability without the settlement would have been. Of course, Mr Attwells did not pay the lesser sum in time and the ANZ sought to enforce the judgment (which is probably why the second appellant in the High Court was the assignee of Mr Attwells’ cause of action courtesy of Mr Attwells’ trustee in bankruptcy).

After failing to set aside the judgment as an unenforceable penalty, Mr Attwells sued his solicitor. The parties sought to have determined as a preliminary question whether the solicitor could rely on the immunity as a complete defence. The primary judge declined to answer the question. The NSW Court of Appeal decided that it should be answered in favour of the solicitor. The High Court decided that it should be answered in favour of Mr Attwells.

Mr Attwells also asked the High Court to abolish the immunity. That required reconsidering the High Court’s decisions in Giannerelli v Wraith (1988) 165 CLR 543 and D’Orta-Ekenaike v Victoria Legal Aid (2005) 223 CLR 1. The High Court unanimously refused to do so, with the majority expressly stating, among other things, it was up to Parliament to effect such a change in the law ([28]).


The High Court was at pains to restate, in the clearest of terms, the policy underpinning for the immunity. Here’s a highlights reel of that restatement by the majority (French CJ, Kiefel, Bell, Gageler and Keane JJ):

  1. “[O]nce a controversy has been finally resolved by the exercise of the judicial power of the State, the controversy should not be reopened by a collateral attack which seeks to demonstrate that that judicial determination was wrong.” ([34])

  2. “The advocate’s immunity is … justified as an aspect of the protection of the public interest in the finality and certainty of judicial decisions by precluding a contention that the decisions were not reached lawfully.” ([35])

  3. “[I]t is the participation of the advocate as an officer of the court in the quelling of controversies by the exercise of judicial power which attracts the immunity.” ([38])

  4. “[T]he basis of the immunity is the protection of the finality and certainty of judicial determinations …” ([46])

  5. “[T]he public policy which justifies the immunity is not concerned with the desirability or otherwise of settlements, but with the finality and certainty of judicial decisions.” ([52])

  6. “The advocate’s immunity is grounded in the necessity of ensuring that the certainty and finality of judicial decisions, values at the heart of the rule of law, are not undermined by subsequent collateral attack.” ([52])

You get the picture.

Whilst the justification for a test and the test are not the same thing, as will be seen below, in the case of the immunity, the justification sets a limit on the reach of the test.

The test

What conduct is immune from suit? The starting point is that an advocate’s conduct of a case in court is caught by the immunity.

The next step is that “work done out of court which leads to a decision affecting the conduct of the case in court” is also caught by the immunity. Mason CJ said Gianarelli (165 CLR 543, 559) that “it would be artificial in the extreme to draw the line at the courtroom door” (quoted by the majority in Attwells, [2]). (The concept of “a decision” – in particular, whose decision: the client’s or the advocate’s? – was the subject of submissions, but the majority in Attwells does not enter into that debate. The answer seems to be that it does not matter. Mr Attwells’ case was that he was given negligent advice to consent to the proposed orders being made. The relevant conduct was the advice; the impugned “decision” in the conduct of the case in court was the consenting to the proposed orders. In other words, both the advocate and the client participated, albeit in different capacities, in the conduct (allegedly) resulting in loss to the client.)

The out-of-court work requires a certain quality about it. The words typically used are ([2] quoting McCarthy P in Rees v Sinclair [1974] 1 NZLR 180, 187):

where the particular work is so intimately connected with the conduct of the cause in Court that it can fairly be said to be a preliminary decision affecting the way that cause is to be conducted when it comes to a hearing.

Here’s how the majority in Attwells explained the test ([5]-[6], emphasis added):

[T]he intimate connection required to attract the immunity is a functional connection between the advocate’s work and the judge’s decision. As Mason CJ said in Giannarelli, the required connection is between the work in question and the manner in which the case is conducted in court. Both D’Orta and Giannarelli were concerned with claims which impugned a judicial determination to which the allegedly negligent work of the advocate contributed. … [T]he public policy, protective of finality, which justifies the immunity at the same time limits its scope so that its protection can only be invoked where the advocate’s work has contributed to the judicial determination of the litigation.

In short, in order to attract the immunity, advice given out of court must affect the conduct of the case in court and the resolution of the case by that court. …

As will be seen, the court’s function of actively resolving or determining the dispute must be engaged. It is not enough simply that there is a court order.


Applying the test to advice in respect of settlements of civil proceedings, the majority said ([38]) “the immunity does not extend to acts or advice of the advocate which do not move litigation towards a determination by a court”. The reason immediately relied on for that statement was the justification of the immunity: “the participation of the advocate … in the quelling of controversies by the exercise of judicial power”.

It is not explained in the majority’s judgment whether or how advice to enter into a settlement is “intimately connected” with conduct of the case in court. As seen above, the justification acts as a limit on the scope of that test. So, presumably, the majority had in mind that an advocate’s advice to a client about a settlement is indeed “intimately connected” with the advocate’s in-court work, because the settlement (if assented to by all relevant parties) brings the work to an end. (That is how Gordon J, dissenting, seems to have regarded it: [106] and [108].)

The majority later said ([46]):

Once it is appreciated that the basis of the immunity is the protection of the finality and certainty of judicial determinations, it can be more clearly understood that the “intimate connection” between the advocate’s work and “the conduct of the case in court” must be such that the work affects the way the case is to be conducted so as to affect its outcome by judicial decision. The notion of an “intimate connection” between the work the subject of the claim by the disappointed client and the conduct of the case does not encompass any plausible historical connection between the advocate’s work and the client’s loss; rather, it is concerned only with work by the advocate that bears upon the judge’s determination of the case.

That last clause is worth highlighting: “[the intimate connection] is concerned only with work by the advocate that bears upon the judge’s determination of the case”. There are any number of steps in litigation that, if handled negligently, can lead to a disappointing result for a client. Unless that disappointing result is the subject of a “judge’s determination”, it appears that the negligent conduct is beyond the scope of the immunity. Query whether, for example, the failure to secure the attendance of a key witness by serving a subpoena on the witness is an example of such conduct. (This is obviously contrasted with a negligent decision not to call such a witness.)

Further illustrating the bounds of the immunity ([50]):

The insufficiency of a mere historical connection between an advocate’s work and a litigious event may be illustrated by reference to negligent advice to commence proceedings which are doomed to fail. No one suggests that the immunity is available in such a case. Likewise, advice to cease litigating or to continue litigating does not itself affect the judicial determination of a case.

Importantly, the majority excluded from their consideration whether the immunity extends to “cases where, although the parties have agreed upon the terms of the order which a court is asked to make, the making of the order itself requires the resolution of issues by the exercise of judicial power” ([61]). A number of examples are given. Added to that list might be settlement of financial disputes under the Family Law Act 1975: consider for example Bell J’s decision in Goddard Elliott v Fritsch [2012] VSC 87 (which goes for a whopping 1,146 paragraphs (plus an appendix) over 321 pages), where his Honour said ([815]):

Under the current state of the law, it is not appropriate to go into any of these matters. For the purposes of the immunity, the final character of the court’s determination results from the exercise of the judicial responsibility of the court to approve or otherwise consider the merits of the proposed orders, not from the quality or extent of the approval or consideration which was given in fact. As Pill LJ held in Kelley [v Corston [1998] QB 686, 707]: ‘Immunity from suit (if not otherwise present) should not … depend upon the degree of conscientiousness of the judge who makes the consent order’.

Consent orders

The majority succinctly summarised the position with respect to consent orders ([62], emphasis added):

In the present case, the consent order and associated notation by the Court reflected an agreement of the parties for the payment of money in circumstances where no exercise of judicial power determined the terms of the agreement or gave it effect as resolving the dispute. The consent order may have facilitated the enforcement of the compromise, but it was the agreement of the parties that settled its terms.

The dissent

The two Victorian Justices of the High Court dissented. Gordon J wrote the dissenting judgment, and Nettle J agreed with her Honour, but added a few comments of his own, which reflected a different characterisation of consent orders from that given by the majority.

Gordon J expressed a similar view to that of the majority about the justification of the immunity. Her Honour said ([101]), “Of course, the law does not assume the processes or outcomes of judicial determination are perfect. That is not the reason the immunity attempts to protect judicial determinations. Rather, it does so because there must be finality in the resolution of disputes.” And, in reliance on some passages from D’Orta, her Honour said further ([102]), “[T]he re-litigation of disputes already judicially quelled is a threat to finality wherever a client alleges that, but for their lawyer’s negligence, a judicial determination would have been different.”

In applying the test, Gordon J came to a different conclusion from the majority. Her Honour said ([104]), “The issue is resolved by understanding that there was a final quelling of the controversy between the parties by the [consent orders]”. Her Honour then relied on two main points supporting that conclusion:

  1. The consent to the entry of judgment against Mr Attwells was analogous to the plea of guilty in D’Orta. The antecedent rights and liabilities of both disappointed clients merged in the judgments of the court – the making of the orders being the relevant exercise of judicial power.

  2. The consent orders stand unimpeached, but Mr Attwells was seeking another judicial determination “by challenging the bases on which the judgment was entered”.

Her Honour then said ([111]):

[The judgment] stands unimpeached. Yet central to the claim against Jackson Lalic Lawyers is the contention that Mr Attwells was not indebted to the ANZ Bank in the amount recorded in that judgment. That is, the claim necessarily disputes the judgment that has been entered in respect of the dispute commenced by the ANZ Bank – the Recovery Proceedings. That is a direct challenge to finality. It is impermissible.


There is obviously some power in what Gordon J says. It seems to me that the difference between the majority’s view of consent orders as a relevant exercise of judicial power and Gordon J’s view is whether, quite simply, the judge had to think about what order to make and had to give reasons for doing so (eg following a trial) – so query what that means for consent orders or settlements the subject of express judicial approval.

%d bloggers like this: