Not Your Grandparents’ Civil Law: Decisions Are Getting Longer. Why and What Does It Mean in France and Québec?

(I’m very pleased to welcome Antoine Dusséaux from Doctrine as a guest contributor on this post. You can read more about and from Antoine below.)

Given my job (CEO at CanLII – saved you a click ) and law degree from a civil law program, I often get to talk about the differences between legal information in Québec compared to the rest of Canada.

I was an intellectual property (IP) lawyer before joining CanLII and although I wasn’t a litigator per se, I was routinely involved in IP cases before the Federal Court. Before that, I did a bit of commercial litigation, mostly at the Québec Superior Court.

I often explain to whoever is unlucky enough to be stuck with me at a cocktail that, in my experience, there was no difference in how we used case law in “pure” civil law matters (if there is such a thing) versus how I used case law when practising IP before the Federal Court. No need to point out that IP is statutory law and that the Civil Code is actually statute, I realize that… but it’s still logical to expect that the use of case law will be different under civil law than in anything that is covered by Common Law.

It turns out that lawyers in Québec, just like in the rest of the country, love case law, and it’s not uncommon to see fairly large books of authorities being filed for relatively routine interlocutory motions in civil law matters. That often surprises my common-law-trained interlocutors who think civil law in Québec as being much more exotic than it is.

It’s not uncommon during these conversations for me to point out that the few times I had to do legal research in French law I was completely lost. That is true even if I was trained as a civilist in a province where the Civil Code is said to have the Napoleonic Code as one of its ancestors.

Court decisions from France threw me off by their short length, relative lack of factual details, and extremely brief reasons. The highest court in France, the Cour de cassation, generally issues decisions a handful of paragraphs long. This is, of course, a significant difference with our highest courts’ novel-length work products.

I recently had the chance to spend some time on Skype with Antoine Dusséaux from the French legal research service Judging by how much we ran over the time we had allotted to our call in our respective calendars, we had a good time comparing the Canadian and French situations with respect to legal information and access.

One comment by Antoine struck me: He explained that lower court decisions in France are fairly long and detailed (I didn’t remember that), and that since they are increasingly made available publicly, this influences the “legal information tastes” of French professionals. He mentioned that decisions in France keep getting longer. He also mentioned that the Cour de cassation itself recently announced that it would follow a new model for its decisions, and would now start drafting judgments that are much more similar to how court decisions are drafted in Canada.

Antoine thus hypothesized that the increased availability of legal information changes legal information itself. That got me to dust off a little old project I had of studying the evolution of decision length in Québec since the launch of CanLII and of the other free (as in “unimpeded”) access resource for caselaw in Québec ( Before these resources came into existence, Québec law was much less easily available. Observations about the results of my humble “research project” follows.


A quick word on my approach:

At first, I took a sample of 400 cases from each year where CanLII’s coverage is complete for the Cour du Québec (2003 to 2017 inclusively), out of a population of about 19,000 decisions issued per year. I didn’t include 2018 and 2019 since there seems to be a lag in receiving 2018 decisions from the CQ (we think it’s because of a backlog in redactions to be made in the Chambre de la jeunesse decisions in case you are asking) and since we’re only half into 2019.

Since the results were “promising” (in the sense that I felt I would have something to say about it), I asked Patrick Bourdon, one of Lexum’s awesome programmers, to help me retrieve data about all decisions issued by the three judicial courts in Québec, namely the Cour du Québec (CQ), the Cour supérieure (CS) and the Cour d’appel (CA) over the same period.

I calculated two metrics:

  • The average length (in terms of number of characters) of the decisions issued by these courts per year.
  • The “citation density” (Antoine’s term, and a most elegant one I should say), i.e. the average number of decisions cited per decision.

The trend is pretty clear:

  • From 2003 to 2017, the length of lower court decisions (CQ and CS) has increased by about 40% and the length of the CA decisions has increased by 20%.
  • The citation density of CQ and CS decisions has doubled. The CA decisions density increased by 36%, a smaller increase but still a material one considering that the CA decisions were already “denser” than the CQ and CS decisions in 2003 (which is normal for an appeal court, in Canada at least), leaving less room for spectacular increases (expressed in percentage).

Quick note: Something happened around 2008 at the CA since both citation density and length dropped significantly, before starting to shoot up from 2009 to 2017. For now I have no idea what this “something” was.

(More charts and data here.)

The correlation coefficient between the evolution of the length of the decisions from the CS from 2003 to 2017, and the yearly increase in the approximate size of the Québec court decisions collection on CanLII over the same period is a nearly perfect 0.986.

Correlation is not causation, so I tried to think of other factors that could explain this and came up with the following hypotheses:

  • Complexity of the cases: I don’t think cases have increased all that much in “legal” or even “factual” complexity (I discuss below that cases may have increased in “evidentiary complexity”), but maybe I’m missing something, in which case feel free to comment.
  • The impact of the increasing number of self-represented litigants (SRL): I don’t know much about the use of authorities by SRL compared to professionals, but I can imagine ways this factor can influence both up and down the length and “density” of decisions. Again, feel free to comment.
  • More evidence due to the “data boom” aka an increase in “evidentiary complexity”: It’s possible that as we are collectively generating more data in our daily (work and personal) lives, what used to be relatively simple matters with a handful of key pieces of evidence (a few pieces of paper correspondence and a contract, for instance) are now made more complex by the production of more evidence from all possible digital sources (emails, instant messages, SMS, pictures, social media posts, etc). One can speculate that judges have to spend more paragraphs discussing this evidence and that this results in longer decisions. I think this factor could indeed influence decision length to some degree. That said, I’m not sure this can be a reasonable explanation for the even bigger increase in density that I have observed. I also doubt it can explain why the trend up in decision length continued in recent years when the increases in smartphone sales and social media adoption (two important sources of data) in Canada have been flattening (see this and this, for what they are worth).

All that to say that for now, the data looks like it could support the hypothesis that greater availability of caselaw changes the nature of caselaw itself. The fact that there’s a similar trend in France, as per Antoine’s observations below, strengthens this conclusion.

Three parting questions:

  • Why was there an uptick starting in 2011 in both length and (for CA and CS at least) density? My hypothesis for now is that in the early years of CanLII, the coverage was just too thin to generate more than small-but-steady growth in impact, but that by 2011, the collection covered a big enough range of years to be relied upon by more people, and in turn to have more impact on the form of legal decisions. This could be supported by previous research that showed that the average lifespan of CA decisions is around 10 years, and by my own quick perusal of historical usage stats at CanLII that indicate that significant growth in traffic year-to-year was still happening at the time.
  • Is this good or bad news? I think that it’s a good thing to have more detailed explanations of the reasons that support a legal finding. I would go as far as to say that this could mean that the quality and predictability of decisions in Québec increased over the same period, thanks to free access to law. While I suspect that some might say that longer decisions are not necessarily better when it comes to informing the parties and the public, I would insist that judges are not just chattier: they also cite more law which is, to me at least, an indication that the increase is also qualitative.
  • What does it say about civil law? This might be a truism to anybody who practices law in Québec, but these results may indicate that it’s not just the form (length, density) of the decisions that changes: Civil law itself is changing. As judges cite more cases, the reasoning may be becoming less purely civilist (deductive and rational), and more inductive and empirical through the combined influence of the Common Law and free access to law. I’ll stop here since Antoine covers this in greater detail.

And now on to Antoine.




Thanks Xavier for inviting me to contribute on this post!

I’m the cofounder and Chief Product Officer of the legal search engine Doctrine and I was never predestined to one day write about court decision length… Indeed, I didn’t study law, but Bioengineering at École Polytechnique (Paris). However, I discovered during my last year of master’s the challenge of access to law on the internet and together with two friends we decided to create to advance open justice in France.

As the cofounder of the largest collection of French cases I’m often asked by non-French lawyers and law librarians: “But… isn’t case-law useless in France?”

The recent call I had with Xavier made me discover the impressive work behind CanLII to disseminate Canadian court decisions. But Xavier also told me he had to answer similar questions to his fellow Canadian lawyers regarding his practice in Québec. As we talked, we noticed more and more converging trends in the judicial systems on both sides of the Atlantic.

And yet, common law and civil law systems are often described as totally different and incompatible, no matter the external forces of globalisation. For instance, Pr. Pierre Legrand wrote in his seminal paper European Legal Systems are not Converging published in 1996: “A civilian can never understand the English legal experience like an English lawyer. […] A lawyer brought up within a system of judge-made law has a legal outlook utterly different from one who has grown up within a codified system. […] Common law lawyers and civilians continue to take a different view of what it is to have knowledge of the law, of what counts as legal knowledge. The positivities in terms of which the real is known, the images of the real that are formed have not coalesced: there continue to exist two fundamentally discrete understandings and representations of truth.”

To prove that common lawyers and civilians had fundamentally different legal reasonings, Pr. Legrand considered, among many other things, the length of court decisions and the description of facts in judgments: “The contrast can be introduced thus: where, as in Continental legal systems, a propositional logic prevails, the question put by the courts (and by the interpretive community at large) is: “Quid juris!”; however, where, as in English law, a situational logic governs, the relevant question to ask is: “Quid facti?” In this way the sources of credibility are seen to differ in the two legal traditions. While the Cour de cassation devotes but a few lines to facts and while the Tatbestand of a German decision also tends to be rather short, an English court will, through its different judges, who may each enter their own statement of facts, continue for pages. The aim is a description of facts that is as thorough as possible. The importance of facts in the decision-making process is confirmed by the insistence on certain facts which may differ from one judge to another.”

He then concluded that common law and civil law legal systems “have not been converging, are not converging and will not be converging”.

His assertion may have been true in 1996, but is it still true in 2019?

Together with Xavier, we decided to look at the evolution of decision length in French and Quebecois courts.

Before looking at the data, here’s a comprehensive diagram of the French hierarchy of courts:

For this study, I didn’t consider administrative courts (right side of the diagram), that belong to a separate system in France, and focused on the following courts (on the left side of the diagram) and years (full years with enough data available to be representative):

  • Court of Cassation (Cour de cassation), “published” decisions (both civil and criminal cases) (CASS), 1960-2018: about 10k decisions given per year. The Court of Cassation is the supreme appellate court for civil and criminal matters.
  • Courts of Appeal (cours d’appel), social, commercial and civil divisions (CA), 2006-2018: about 150k decisions given per year.
  • Commercial Courts (tribunaux de commerce), all cases (inc. small claims) (TCOM), 2006-2018: about 250k decisions given per year.

Regarding the Court of Cassation, only 50% of its decisions are published online. “Unpublished” decisions (“décisions non publiées”) are mostly appeals that were dismissed without providing the detailed rationale for the decision (they’re called in French “rejets non spécialement motivés”). The text of such decisions is usually sparse, containing only a minimal recitation of the facts and a limited description of the law. The term “unpublished opinions” is somewhat of a misnomer, as they are still public records and they should actually all be published online but that’s another story…

It is important to note that all French court decisions are written, there’s no ex tempore judgments in France. This point, together with the fact that France is 8 times more populous than Québec should explain why there are way more decisions given each year in France.

The methodology was quite simple: I considered the average HTML content length (in terms of number of characters) of the decisions, in their anonymised format, issued by these courts per year. The number of characters in HTML is a bit longer (+5-10%) than in normal text because of tags (paragraphs, bold, italic, etc.) but on the other hand anonymisation (“Monsieur Antoine DUSSÉAUX” replaced by “M. X”) reduces the total number of characters so at the end it is fairly similar to the total number of characters of the original court decision in a text format.

I didn’t look at citation density as most French courts don’t cite precedents but only cite legislation, even though this “tradition” is gradually changing too, but it could be the subject of another article…

(Data per year here.)

So the trends are clear:

  • Commercial Courts (TCOM): From 2006 to 2018, the length of first instance court decisions from the Commercial Courts has slowly increased (+7% in 10 years).
  • Courts of Appeal (CA): From 2006 to 2018, the length of court decisions has increased by about +4% per year, which leads to a +58% increase over this period.
  • Court of Cassation (CASS): From 1960 to 2008, the length of court decisions from the Cour de cassation has also slowly, but steadily, increased so that it almost doubled in 50 years. In 2009, there was a huge increase in decision length (2x). Since 2009, the length has continued to increase, at an even faster rate than before (+50% from 2009 to 2018).

Also, if we compare France to Québec, even though French court decisions are still shorter than their Canadian equivalents, they’re catching up at a fast pace.

What are the reasons for this continuous increase of French court decision length?

For the Court of Cassation, in 2008 the Court started to annex at the end of its decisions the original grounds of law of the parties copied from their written submissions (“moyens annexés” in French). Hence the surge that year. Those “moyens annexés” written by lawyers are often more developed than the decision itself and they can in particular focus more on facts. The continuous increase that followed after 2008 is probably the consequence of a gradual change of the writing style, with more weight given to the reasoning and sometimes to facts. The Court of Cassation’s writing style was indeed more and more criticized by legal practitioners and legal scholars for being too concise. In 2015, under the leadership of visionary Chief Justice Bertrand Louvel the Court launched a working group to reform its writing style. The Court in particular cited globalisation and the ongoing competition between legal systems as the main drivers for reform. As a result, the Court announced two months ago that all its decisions will now follow a new writing style: more direct, with more emphasis on facts, and if necessary with citations of precedents (quite a revolution in a civil law system!). In its press release the Court provided with examples before and after the new writing style: the new style appears to be almost twice as long as the previous one, which means that length should continue to grow.

There is no single writing style for the 36 French courts of appeal but appeal judges may have followed the trends of the Court of Cassation, given its influence on lower courts.

So, coming back to Pr. Legrand’s assertion, his analysis proved to be wrong. First of all, decisions from the Court of Cassation were +40% longer in 1996 than they were in 1960 and he didn’t mention this point, as he was probably not aware of it. But the increase continued after the publication of his paper so that decisions from the Court of Cassation are now four times longer than they were at the time he wrote his article! Moreover, the Court of Cassation now devolves longer paragraphs to the description of facts.

But, not only is there a “common-lawisation” of French courts, with longer court decisions and a growing insistence on facts. There is also an “international influence on the common law” to quote the late Lord Toulson. For instance, Pr. Legrand wrote in 1996 that “an English court will, through its different judges, who may each enter their own statement of facts, continue for pages”. And yet, single judgments that are the standard in France are now common in England as well as explained Lord Reed, Deputy President of the UK Supreme Court: In practice, it has also become increasingly common in recent years for Justices to produce joint judgments.”

Another example of such an influence is the shift toward a purposive approach to statutory interpretation by English courts. As Lord Steyn in R (on the application of Quintavalle) v Secretary of State for Health [2003] UKHL 13 put it: “The pendulum has swung towards purposive methods of construction. This change was not initiated by the teleological approach of European Community jurisprudence, and the influence of European legal culture generally, but it has been accelerated by European ideas.”

There are many other examples and Pr. Ugo Mattei described some of them in Civil Law and Common Law: Toward Convergence? in 2008. The reasons usually given for this convergence are globalisation and the growing importance of transnational courts such as the European Court of Justice and the European Court of Human Rights.

Is this good or bad news? According to Lord Bingham the first sub-rule of the rule of law is that “the law must be accessible and so far as possible intelligible, clear and predictable. This seems obvious: if everyone is bound by the law they must be able without undue difficulty to find out what it is, even if that means taking advice (as it usually will), and the answer when given should be sufficiently clear that a course of action can be based on it.” So, I, like Xavier, think that it’s a good thing for the rule of law to have longer decisions, with a more detailed legal reasoning, as long as we avoid “the length, complexity and sometimes prolixity of modern common law judgments” criticized by Lord Bingham.


  1. This is a nice post, but there is no need to refer to the courts only by their French names. The courts in Québec have English names as well because s. 133 of the Constitution Act, 1867 provides that English and French are the languages of justice and legislation in Québec. Thus, the Courts of Justice Act, CQLR c T-16, s. 1 provides:
    The Courts of Québec, in civil, criminal and mixed matters, are:
    The Court of Appeal;
    The Superior Court;
    The Court of Québec;
    The Municipal Courts.