Lower Court Influence on High Courts: Evidence from the Supreme Court of the United Kingdom

Do lower court judges influence the content of Supreme Court opinions in the United Kingdom? Leveraging original data, we analyze opinion language adoption practices of the UK Supreme Court. We advance a theory where the justices ’ choices to adopt language from lower court opinions are influenced by Supreme Court-level attributes and Court of Appeal case characteristics. We uncover compelling evidence that UK Supreme Court justices incorporate language extensively from the written opinions of the Court of Appeal of England and Wales. Our findings have significant implications for opinion formulation, doctrinal development, and higher and lower court interactions within comparative courts

In December 2011, the Supreme Court of the United Kingdom issued a landmark decision on criminal liability in R v. Gnango [2011 UKSC 59].The case facts involved the murder of a nursing home worker caught in the crossfire of two teenagers engaged in a gunfight in an adjacent parking lot in south London.The Supreme Court's decision to reinstate the murder conviction of Armel Gnango under the principle of joint enterpriseeven though Gnango did not fire the fatal shotcaptured headlines in the UK.Less obvious, though was that the Supreme Court's lead opinion directly adopted nearly 12% of the language from the lower court opinion.The phenomena of opinion language adoption raises important questions.Do lower court opinions influence the content of Supreme Court opinions in the United Kingdom, and if so, what factors motivate language borrowing tendencies among justices in the UK Supreme Court?Prior scholarship demonstrates that adherence to higher court precedent is especially central in the UK given the country's lack of a written constitution as well as an apolitical appointment process predicated on a judge's experience and reputation (Hanretty 2020).While the lower courts' tendencies to overwhelmingly adhere to the principle of stare decisis demonstrates the Supreme Court's capacity to shape lower court behavior (Masood and Bowie 2023), existing studies have not yet explored whether and the extent to which lower court judges are able to influence decision makers in the UK Supreme Court.
Language borrowing speaks directly to influence because, at a minimum, it indicates that the higher court is relying on the lower court's opinion to justify an outcome and is ultimately allowing lower court judges to shape Supreme Court doctrine (Corley, Collins, and Calvin 2011). 1 Nearly all cases that reach the UK Supreme Court have been heard by a lower court, meaning that lower court opinions can be a crucial starting point for the justices to reference when they are tasked with writing their own opinions (Corley, Collins, and Calvin 2011; Hanretty 2020; Bowie  and Savchak 2022). 2 Scholars have made significant strides in improving our understanding of how lower court judges affect higher court opinion content in the American context, yet little is known about whether language borrowing is a phenomenon that extends to other judiciaries, such as the United Kingdom.
We assess whether the Supreme Court of the United Kingdom incorporates language from decisions issued by the court directly below, the Court of Appeal for England and Wales. 3We offer a theoretical framework in which language adoption tendencies are influenced by both Supreme Court-level attributes and Court of Appeal case characteristics.We argue that the limited pervasiveness of ideology during the decision-making process as well as institutional norms and idiosyncrasies that are unique to the UK will influence the extent to which Supreme Court justices adopt language from lower court opinions.More specifically, we theorize that the career-based nature of the UK judiciary in which an individual's movement up the hierarchy is predicated on judicial experience and a judge's capacity to adhere to precedent create an environment where judicial actors are socialized to borrow language frequently.To our knowledge, this is the first study to systematically analyze the extent to which a court of last resort borrows language from lower court opinions outside of the American context.As such, our work sheds new light on the mechanisms motivating language borrowing tendencies in a comparative judicial environment.
To evaluate opinion language borrowing patterns in the United Kingdom, we analyze over 500 Supreme Court-Court of Appeal opinion dyads between 2009 and 2019.We assemble three distinct datasets, which include (1) the universe of decisions issued by the UK Supreme Court, (2) the universe of decisions issued by the Court of Appeal of England and Wales, and (3) data on language borrowing between the top two tiers of the UK judiciary.We find that justices on the UK Supreme Court adopt language extensively from the opinions of the Court of Appeals of England and Wales and that they do so because of both Supreme Court attributes and lower court case characteristics.Notably, our results demonstrate that although lower court judges are meaningful drivers of the development of Supreme Court doctrine through the content of their own opinions, language borrowing practices are not influenced by the ideological proximity between the judicial actors within these courts.These results provide new and important insights into the critical role that lower court judges play in advancing the development of law, as bottom-up influences seemingly help shape both the behavior of Supreme Court justices and the broader contours of legal doctrine within the United Kingdom.More broadly, our analysis offers implications toward a more nuanced understanding of opinion development and hierarchical interactions within comparative courts.

Doctrinal development and the importance of judicial opinions
Opinion writing is central to judicial decision-making.Hazelton and Hinkle note in their recent work that, "[i]t is the majority opinions, not outcomes, that lower courts and government officials interpret and implement" (2022, 157).Perhaps most significantly, written opinions clarify the law for the parties involved in any particular case and, by extension, serve as a source of transparency for other judicial actors, governmental agencies, and the public at large (Corley, Collins, and Calvin 2011).Written opinions are critical in establishing and refining relevant precedents that inform how the law is adjudicated in subsequent cases throughout each tier of the judicial hierarchy (Callander and Clark 2017).
Studies examining judicial opinion content often focus on the practice of language borrowing, where high court judges and justices directly implement language from lower court opinions into their own decisions, thus codifying the language from the lower courts as national precedent (Corley, Collins, and Calvin 2011).There is considerable evidence that higher courts routinely borrow language from lower court opinions because these opinions provide information on the state of the law as it relates to the case at hand, offer persuasive legal arguments, and cover complex issues (Corley, Collins, and Calvin 2011; Savchak and Bowie 2016; Bowie and Savchak  2022).Other work on bottom-up influences reveals that Supreme Court justices ascertain critical information on the applicability and policy consequences of legal rules through implementation patterns within the lower courts (Hansford, Spriggs,  and Stenger 2013).In addition, research by Clark and Carrubba (2012) and Carruba and Clark (2012) suggests that Supreme Court justices utilize the lower courts as a means to mitigate the substantial costs associated with developing new doctrine (see also Beim 2017).Beyond bottom-up accounts, studies demonstrate that justices are not averse to turning to external sources to help bolster the content of their opinions (Corley 2008; Collins, Corley, and Hamner 2015; Black, Owens, and Brookhart 2016).Given the evidence in support of opinion content being shaped by dynamic factors, studying language borrowing tendencies between courts helps provide key insight into how justices craft their opinions and the ways in which lower court judges help shape the development of legal doctrine from the bottom up.In the next section, we explore the factors that influence language borrowing tendencies among Supreme Court justices and lower court opinions in the United Kingdom.

The institutional dynamics of the UK judiciary
The Supreme Court of the United Kingdom, which began operations in 2009, was created by the Constitutional Reform Act of 2005.The new UK Supreme Court replaced the Appellate Committee of the House of Lords to provide greater judicial independence from Parliament and to further distinguish the functions of the Court as inherently nonpolitical (Hanretty 2020).The UK Supreme Court hears appeals from the three distinct judicial systems -England and Wales, Scotland, and Northern Ireland4 that make up the UK legal structure (see Hanretty 2020, 6-7).Most cases begin in the High Court (i.e., trial court of first instance) and are heard by a single judge.Decisions made by the High Court may be appealed to the Court of Appeal, which is the principal intermediate appellate court in the UK.The Court of Appeal of England and Wales is made up of the Criminal and Civil divisions, where cases are decided by three-judge panels (Blanes i Vidal and Leaver 2013; Hanretty 2020).Unless the UK Supreme Court decides to review a case on appeal, the Court of Appeal of England and Wales is effectively the final arbiter of most judicial appeals.
The mechanisms that help explain language borrowing in the United Kingdom can be attributed to the judiciary's unique institutional structure and professional norms.Unlike other common law judiciaries in which the appointment process is political, the UK system is a career-based judiciary, meaning that judicial advancement is largely predicated on prior record of service and a judge's reputation among his or her judicial peers (Liu and Zhang 2001).To be appointed to the Court of Appeal of England and Wales, candidates not only have to be recommended to the monarch5 by the Prime Minister on behalf of the Lord Chancellor, but they are also required to have a combined twenty-five years of experience serving as barristers or advocates of the court (Hanretty 2015; Masood and Lineberger 2020).Such substantial tenure requirements for judicial advancement distinguish the UK courts from most other legal systems where judicial appointments are, at least in part, political in nature (Blom-Cooper, Dickson, and Drewry 2009).
Advancement to the Supreme Court is also based on experience and expertise.Candidates for elevation to one of the twelve positions on the Court must have either served on a High Court 6 or the Court of Appeals for at least two years or have practiced law for a minimum of fifteen years.After the Judicial Appointments Commission 7 reviews each candidate's application, the committee's nominee must be approved by the Law Lord Chancellor, the Prime Minister's Office, and, ultimately, the monarch.Importantly, nominees cannot be an active member of a political party and justices must retire at the age of 70. 8 Given the apolitical and merit-based nature of the appointment process, the structure of the UK judiciary incentivizes lower court judges to craft opinions grounded in legal precedent rather than motivated by ideological considerations (Atkins 1990;  Salzberger and Fenn 1999; Malleson and Moules 2010; Darbyshire 2011).Decisions that are based largely on precedential and legal factors enhance the legitimacy of judicial opinions and, in turn, increase the likelihood of further promotion up the judicial ladder (Drewry, Blom-Cooper, and Blake 2007; Masood and Lineberger 2020; Masood  and Bowie 2023).The size of the UK judiciary also limits the pervasiveness of ideology.The twelve Supreme Court justices monitor only 38 appellate court judges, and they review approximately 30 to 40 percent of petitions in a given term.The small ratio of higher court justices to lower court judges coupled with considerable auditing capacity severely limits the lower court's ability to hand down ideological decisions without facing scrutiny or reversal from their Supreme Court superiors (Caminker 1994; Klein  and Hume 2003; Posner 2005; Bowie and Songer 2009).
Because the United Kingdom's career-based judiciary and robust oversight capabilities significantly minimize the extent to which lower court judges write opinions with the sole intention of pursuing their political preferences (Hanretty 2020), precedentnot ideologyis central to the decision-making process.Supreme Court justices should be less inclined to question the logic and motivations behind Court of Appeal decisions because those decisions are consistently grounded in the precedents that the justices themselves have created and reaffirmed.Therefore, the institutional norm for the appellate court to rely on Supreme Court stare decisis should result in high levels of opinion language borrowing, as the justices can be confident that Court of Appeal decisions align with what the Supreme Court has already established as legitimate law.

Theoretical expectations
Given that there is little empirical evidence showing that attitudinal preferences are central to the UK's opinion writing and decision-making processes (Malleson and  Moules 2010; Hanretty 2020), we premise our theoretical expectations on the notion that language borrowing tendencies within the United Kingdom are driven by unique legal and institutional factors rather than distinctly political preferences.Prior studies on hierarchical interactions suggest that in the absence of ideological motivations, institutional (Masood and Lineberger 2020) and casespecific factors (Corley, Collins, and Calvin 2011) can impact judicial decision-making and opinion writing behavior.As a result, we expect language borrowing to be influenced by a combination of Supreme Courtlevel attributes and Court of Appeal case characteristics.
We begin with a discussion of the Supreme Court level factors that should impact the propensity for justices to borrow language from lower court opinions.
Our initial expectation is that the ideological congruence between the Supreme Court and lower court panels should have little to no effect on whether Supreme Court justices borrow lower court opinion language.The UK Supreme Court typically employs pseudo-random panels of five justices when deciding a case, meaning there is uncertainty regarding which set of justices will serve on a given panel.Therefore, there is no reason to expect, a priori, that a particular Supreme Court panel configuration will borrow more or less based on ideological differences in the aggregate.Panel assignment is conditional on each justice's area of specialization, as well as the justices' schedules, professional recusals, and workload considerations (Blom-Cooper, Dickson, and Drewry 2009; Hanretty 2017;  Hanretty 2020).Similarly, the mandatory retirement age of 70 creates low membership stability and ultimately engenders additional ambiguity about the composition of justices who will review appellate court decisions (Hanretty 2020).Random panels along with significant membership turnover make it difficult for lower court judges to consistently predict the ideological preferences of the reviewing justices, meaning lower court judges will write opinions that largely adhere to precedent in order to avoid Supreme Court reversal (Masood and  Lineberger 2020).As such, we expect Supreme Court justices to borrow language from lower court opinions regardless of the ideological similarities between appellate and Supreme Court panels.
Hypothesis 1 Ideological congruence between the Supreme Court and Court of Appeal panels should not impact language borrowing tendencies.
Although we do not expect ideology to play a meaningful role in the justices' decisions to incorporate lower court language within their own opinions, they should intuitively be more inclined to borrow language when they are affirming a lower court decision.Bowie and Savchak offer that, "when affirming a judgment below, a justice may look to lower court opinions for shortcuts in useful language that may be helpful for explaining their own legal rationale" (2022, 11).We might expect UK Supreme Court justices to frequently consult and borrow language from lower court opinions when they affirm because upholding a lower court decision reflects a mutually agreeable outcome that likely involves overlapping or highly similar legal justifications between the two levels of the judiciary.However, if the Supreme Court reverses a lower court decision, the justices have to articulate original arguments that explain why the lower decision is problematic and why an alternative rationale is needed to justify the legal outcome.In such an instance, it is highly unlikely for the justices to extensively borrow language from the lower court opinion.Therefore, we expect justices to have a greater propensity to borrow language from opinions when they agree with the lower court outcome.
Hypothesis 2 Justices on the UK Supreme Court will borrow more language from the Court of Appeal opinion when they affirm the lower court decision.
Finally, the size of the reviewing Supreme Court panel should play a considerable role in the justices' choices to borrow lower court opinion language.We address two theoretical explanations that explain this expectation.First, larger panels reflect comparatively more important decisions (Hanretty 2020).Although the justices typically make decisions in pseudo-random panels of five, the Court employs larger panels of seven, nine, or even eleven justices when a case is particularly consequential. 9The UK Supreme Court notes within its rules that cases assigned more than five justices indicate, among other things, "a case of great public importance."We expect that justices will devote more time to crafting original opinions in salient cases given their gravity and heightened likelihood of public attention (Corley 2008; Corley, Collins, and Calvin 2011; Hanretty 2020).Second, compared to the standard five-justice panel, larger panels provide the justices with comparatively more opportunities to deliberate amongst each other, which should reduce the need to borrow from lower court opinions.As such, we expect borrowing rates to be higher when the Supreme Court panel consists of the standard, fivejustice configuration as opposed to panels greater than five.
Hypothesis 3 Justices on the UK Supreme Court will borrow more language from the Court of Appeal opinion when their panel composition consists of five justices.
In addition to the Supreme Court attributes discussed above, casespecific characteristics at the appellate leveloutside of the control of the justicesmay also impact language borrowing tendencies within the UK.First, Supreme Court justices may be more inclined to borrow lower court language when a lower court panel decides a case unanimously.Research demonstrates that judicial decision-makers sitting on courts of last resort (Corely, Steigerwalt, and Ward 2013) and lower courts (Hansford and Spriggs 2006; Masood, Kassow, and Songer 2019; Masood and  Kassow 2023) are more likely to adopt legal rationales in the absence of a counter signal by a dissenting judge (Beim, Hirsch, and Kastellec 2014).Dissenting opinions also provide judicial decision-makers with additional avenues from which to borrow language that might be more persuasive than the content of the lower court majority opinion.The absence of dissenting opinions removes such opportunities that may dissuade justices from borrowing language from the majority opinion.
Previous research also finds that the type of opinion influences whether a higher court borrows from the resulting opinion (Corley, Collins, and Calvin 2011).For instance, split decisions may signal to the justices that there is ideological divisiveness or disagreement over the legal justifications on which the decision is based (Masood  and Lineberger 2020).Such disagreements should decrease the likelihood for the Supreme Court to borrow lower court language, particularly due to the importance that the justices attach to precedent rather than ideological considerations.Because unanimous decisions in the lower court provide justices with a single avenue from which to adopt language and signal that the decision is grounded in legal rationale based on consensus, we expect to see higher rates of borrowing in instances where lower court decisions are unanimous.
Hypothesis 4 Justices on the UK Supreme Court will borrow more language from unanimous Court of Appeal opinions.

9
The official criteria for when the UK Supreme Court would consider the use of larger panels are stated in the rules and procedures of the Court.The rules state that more than five justices should sit on a panel if (1) the Court was being asked to depart or may decide to depart from a previous decision, (2) a case was of high constitutional importance, (3) a case was of great public importance, (4) there was conflict between decisions of the UK Supreme Court, the Judicial Committee of the Privy Council, or (5) a case concerned an important aspect of the European Convention on Human Rights (see also Hanretty 2020).This information is available at https://www.supremecourt.uk/procedures/panel-numbers-criteria.html.Finally, an additional Court of Appeal case characteristic that should impact Supreme Court language borrowing is the length of the lower court opinion.Lengthier opinions, by definition, provide more content for the justices to potentially incorporate within their own opinions.Opinion length might also be indicative of a more thoughtful and deliberative process of crafting an opinion (Leonard and Ross  2016; Hinkle 2017).As such, the justices might be more willing to adopt language from lengthier opinions because longer opinions may, in the aggregate, better justify the logic of a lower court decision compared to shorter, less precise lower court justifications.This could be due to the fact that lengthier opinions may be better embedded within existing doctrine and are therefore more persuasive or informational.Lengthier opinions are also arguably applicable to a wider set of factual situations if we assume that longer opinions discuss the applicability of the legal reasoning more thoroughly.For instance, Masood and Lineberger (2020) demonstrate that lengthier opinions supply more information to judicial actors in the UK compared to shorter opinions.Beyond this finding, the justices might view longer opinions as more legitimate because length could imply greater attentiveness and care in articulating the decision.For these reasons, we expect Supreme Court justices to borrow language more extensively from lengthier appellate decisions compared to opinions that are comparatively shorter.
Hypothesis 5 Justices on the UK Supreme Court will borrow more language from Court of Appeal opinions as the length of those opinions increase.

Data and methods
In order to empirically assess how the United Kingdom's career-based judiciary and institutional norms affect the degree to which justices borrow language from Court of Appeal opinions, we collect data from multiple sources to construct several new and original datasets.First, we collect data from the British and Irish Legal Institute (BAILII) website10 to identify the universe of cases decided by the UK Supreme Court between the years 2009 and 2019. 11The final dataset includes 548 judicial decisions, where our unit of analysis is the Supreme Court majority opinion -Court of Appeal opinion dyad. 12o create our dependent variable, we utilize BAILII's case history function 13 to access higher and lower court opinion texts.After locating each Supreme Court opinion 14 and the accompanying appellate court decision under review, we convert each opinion into a text format and use WCopyfind 4.1.5 15oftware which identifies language similaritiesto assess the degree to which the lower court opinion words and phrasing were replicated, or "adopted," in the UK Supreme Court opinion (Corley 2008; Corley, Collins, and Calvin 2011; Bloomfield 2016; Savchak and Bowie  2016; Bowie and Savchak 2019; Bowie and Savchak 2022). 16Our dependent variable is the percentage of a Supreme Court opinion's language that is borrowed directly from the lower court decision.We set the shortest phrase match to ten words, which tells the software to ignore language matches of nine words or less.This provides more conservative estimates than the frequently used six-word match. 17The dependent variable ranges from 0 to 42 percent.Table 1 provides an example of the type of borrowing reported by the WCopyfind program.Returning to the case of R v. Gnango, the Supreme Court's lead opinion highlights that the account of the case facts comes "almost verbatim [from the] judgment of the Court of Appeal." 18This direct incorporation of the lower court's language into the Supreme Court opinion does not, however, end with a summary of the facts.Table 1 illustrates one of the many instances of word-for-word language adoption interspersed throughout the Supreme Court's majority opinion.As the Gnango example demonstrates, the language used in opinions is essential to the judicial process for identifying the issue, clarifying legal doctrine, and establishing guidelines for judges to follow when adjudicating similar cases (Corley, Collins, and  Calvin 2011).lead opinion by issuing a short "speech" that lacks in any substantive contribution to the lead opinion, à la a "me too" opinion (2020, 136).For instance, the full text of Lord Brown's opinion in HH v. Deputy Prosecutor of the Italian Republic ([2012] UKSC 25) highlights the "me too" opinion.However, substantive opinions are always longer than two pages.If a case possesses more than one substantive opinion, then a quick rule of thumb, according to Hanretty (2020), is the lead substantive opinion will either be listed first or be the longest (2020, 137).Additionally, the lead opinion will present the facts of the case, where other substantive opinions will not (Paterson 2013,13).Hanretty shows that by following these two "rules" one can quickly identify the substantive lead opinion 93 percent of the time, and approximately half of the UK Supreme Court opinions feature just one substantive opinion (Hanretty 2020, 138-39). 15The WCopyfind program can be assessed here: https://plagiarism.bloomfieldmedia.com/software/wcopyfind/ (last accessed April 2, 2022). 16More recently, Hazelton and Hinkle (2022) use an alternative measurethe cosine similarity scoreto evaluate "shared language" between litigant briefs and US Supreme Court opinions.The cosine similarity measure calculates the likeness "between two documents … based on individual words used without paying attention to the word order" (Hazelton and Hinkle 2022, 113).We use CopyFind because it is a simpler yet more conservative measure of language borrowing that accounts for word order while allowing for minor differences. 17We follow the standard CopyFind programming as explained in Corley (2008) and Corley, Collins, and  Calvin (2011), which identifies "matches" between the opinion dyads but also excludes citation duplicates.For example, we program the CopyFind software to disregard indiscriminate language, like numbers, outer punctuation, non-words, and word capitalizations.These programming measures ensure that case citations do not count as borrowed language (Corley 2008; Corley, Collins, and Calvin 2011).We set the program to permit minor editing.The minor editing features allow for two minor imperfections within the matching text.Additionally, the minimum percentage of matches that a phrase can contain was set to eighty, which permits us to incorporate matches between opinions even with minor editing of the text (Corley, Collins, and Calvin  2011).Finally, we set the software to skip non-words.We also report the results of an alternate six-word match dependent variable in the appendix.
18 R. v.We include explanatory variables to account for the Supreme Court level attributes and Court of Appeal case characteristics detailed in the theory.Three independent variables capture effects of Supreme Court level attributes on language borrowing.Although we believe that the judiciary in the United Kingdom is predisposed to "getting the law right," nevertheless, it is necessary to assess whether ideological preferences at the Supreme Court and Court of Appeal level exert any influence on decision-making in the UK Supreme Court.Ideological Congruence measures the ideological similarity between the Supreme Court's opinion author and the opinion author of the Court of Appeal.Each judge's ideology is determined by the party of the appointing prime minister. 21A value of 1 indicates ideological congruence between the Supreme Court's lead opinion author and the lead author of the Court of Appeal opinion, and a value of 0 indicates otherwise. 22o assess whether UK Supreme Court justices have a preference toward borrowing lower court opinions when they are affirming a decision, we include the Affirm variable that is coded as 1 when the lower court decision is affirmed, and 0 otherwise.We expect a positive estimate for this variable, demonstrating that the justices are more likely to borrow when they are in agreement with the lower court about the application of the law.In addition, we include a variable to capture whether the size of the Supreme Court panel influences whether the justices adopt language from the Court of Appeal opinion.To be consistent with existing measures, Panel Size equals 1 if the case is decided by the standard panel of five justices and 0 if the panel includes more than five justices (Hanretty 2020; Masood and Lineberger 2020). 23The panel size variable should be signed in the positive direction, as justices should be more apt

Court of Appeal majority opinion Supreme Court majority opinion
There is at the heart of this issue a question of policy.Does the justice and effectiveness of the criminal justice system require the imposition of liability in cases of genuinely agreed duels by acceptance that there can be a joint enterprise of the first type between opposing persons if they agree not only to hit but to be hit? 19ere is at the heart of this issue a question of policy.It is important to note that relying on the party of the Prime Minister is, at best, a proxy for determining the ideology of the justices.Although there are ideal point estimates available for some justices, work by Hanretty (2020) demonstrates that ideal points of the Law Lords predict an inclination to dissent rather than revealing their ideological preferences.Therefore, we revert to an imperfect but established proxy in the literature. 22We estimated models where we consider an alternate measure for ideological congruence based on the proximity between the Supreme Court's opinion author and the ideological direction of the Court of Appeal decision.The results across the operationalization of the ideological congruence variables are highly robust.These results are reported in the appendix. 23We also estimated a model with the full range of the panel size variable as an interval variable and found the results to be highly consistent.We provide these results in the appendix.
to borrow language in cases that are considered more routine and less salient to the public.
We also include two variables to capture appellate level case characteristics.Unanimous equals 1 if the decision from the Court of Appeal is unanimous, and 0 otherwise.We expect this variable to be positively signed, suggesting that justices tend to borrow more language from the lower court opinion when they are certain there is decisional and jurisprudential agreement among the lower court judges.Finally, we include an Opinion Length variable in order to account for the length of each lower court decision.We first determine the number of words within each appellate opinion and then transform the variable by calculating the natural logarithm of the total number of words.This approach allows us to account for any nonlinearity or variation within the values of the variable (see Randazzo, Waterman,  and Fine 2006; Bowie and Savchak 2022).We expect the estimate of this variable to be signed in a positive direction, as it would indicate that justices borrow more language from lengthier opinions compared to those that are less verbose.
Beyond the key variables of interest, we account for several confounders.At the Supreme Court level, we include a control for whether a Supreme Court justice is new to the Court.Freshman Justice equals 1 if the UK Supreme Court opinion writer is within the first two years of serving on the Supreme Court, and 0 otherwise. 24We expect a positive estimate, which would account for an initial period of acclimation to the work of the higher court (Hagle 1993; Bowen 1995).Next, we include a control for the Supreme Court's caseload because a disproportionately higher number of decisions may reduce the amount of time a justice can dedicate to crafting an opinion and increase his or her likelihood to borrow language from the lower court.The Caseload variable is coded based on the number of cases adjudicated by the Supreme Court in each term.Given nonlinearity in the values of this variable, we take the natural logarithm of the total number of cases by term.
Finally, we control for two key factors at the Court of Appeal level that may shape how much (or little) a justice borrows from the Court of Appeal decision.Prior scholarship demonstrates that writing clarity is an important factor affecting the judicial process, as higher levels of opinion clarity increase the transparency and perceived legitimacy of the ruling at hand (Goelzhauser and Cann 2014; Black et al.  2016; Nelson and Hinkle 2018; Bowie and Savchak 2019; Bowie and Savchak 2022).As such, lower court opinions that are clearly written may be more strongly perceived as rationally grounded decisions, meaning we should see higher rates of borrowing from those opinions.To account for these effects, we gauge the clarity of an opinion by taking into account its reading ease as well as the extent to which the opinion uses the passive voice.It stands to reason that it should be easier to decipher and understand the logic of an opinion that is written absent difficult language.Similarly, the absence of passive voice language in legal writing maximizes opinion clarity and decreases the ambiguity of judicial decisions (Wydick 1998; Bowie and Savchak  2022).We gather data for two variables meant to capture lower court opinion clarity.The first is Reading Ease and the second is Passive Voice.We collect both pieces of information by copying the lower court opinion text into a Microsoft Word document and running the text through the software's automated readability tests.Word calculates reading ease using the Flesch-Kincaid readability test, which assigns the document a score between 0 and 100.Higher scores for Reading Ease indicate easier material to read, whereas lower scores indicate more difficulty.The Passive Voice variable reports the percentage of sentences in each lower court opinion that use the passive voice.

Empirical results
Do UK Supreme Court justices frequently borrow language from the opinions of lower court judges?The short answer is yes.We find that justices on the UK Supreme Court directly incorporate some language from the opinions of Court of Appeal judges in over 90% of cases.This suggests an overwhelming embrace of language adoption by the Supreme Court. 25We also discover that UK Supreme Court justices adopt, on average, approximately 9% of the language directly from lower court opinions.To better understand how frequently the justices borrow language from lower court opinions, Figure 1 presents the percentage of Supreme Court majority opinions that include borrowed language by justice.The data show that 11 out of 26 justices in our data (42%) borrow language in every opinion for which they are the lead opinion writer.The remaining 15 justices incorporate borrowed language in at least 80% of the majority opinions they author. 26The data on language borrowing tendencies are clear: it is the exceptionrather than the rulefor a justice to fail to adopt language from lower court opinions.
To more systematically explore the factors that influence language borrowing tendencies in the UK Supreme Court, we estimate linear regression models.The estimates presented in Table 2 include three specifications.Model 1 includes justice fixed effects to account for within-group variation among the justices.Model 2 includes justice and year fixed effects to account for both within-group variation among justices and unobserved heterogeneity among Supreme Court terms.Model 3 includes justice and year fixed effects with block boot-strapped standard errors to preserve the original clustering structure while simultaneously accounting for the small number of group level units (Cameron, Gelbach, and Miller 2008).The results are robust across all model specifications.
Given the UK judiciary's unique institutional structure that emphasizes careerbased elevation and an apolitical appointment process, we hypothesized that ideological compatibility between opinion authors at the Supreme Court and lower court level should have little to no effect on opinion-borrowing tendencies, all else being equal.The coefficient estimate for Ideological Congruence is not statistically significant.Consistent with our expectation, we find that ideological congruence between higher and lower court opinion authors exerts no meaningful effect in either increasing (or decreasing) the extent to which Supreme Court justices adopt language from lower court opinions.In other words, we find no evidence to indicate that justices on the UK Supreme Court are borrowing language from lower court judges with whom they are ideologically aligned.The absence of ideological influences on language borrowing tendencies corroborates our intuition that a nonpolitical, careerbased judiciary seems to minimize the pervasiveness of ideological considerations in both the decision-making and opinion writing processes.The absence of ideological motivations suggests that judges and justices in the UK may prioritize legal or casespecific considerations over ideology, even at the highest tier of the judicial hierarchy.
Next, although we hypothesized that the justices should borrow more language from the lower court opinion when they are affirming that decision, we do not find evidence to support this expectation.The positive yet statistically insignificant Note: The unit of analysis is the UK Supreme Court-Court of Appeal opinion dyad.The dependent variable is the percentage of the lower court opinion that is borrowed by the UK Supreme Court opinion with the shortest phrase match set at ten words.Model 1 includes Justice fixed effects.Model 2 includes Justice and Year fixed effects.Model 3 includes Justice and Year fixed effects with bootstrap standard errors from 1000 block-bootstrap replications given the relatively small number of clustering units (Cameron, Gelbach, and Miller 2008).*p < .05(one-tailed tests).
coefficient estimate for Affirm suggests that Supreme Court justices borrow language from the appellate court regardless of whether the Court agrees with the lower court's outcome.This unexpected finding indicates that mere agreement with the legal outcome at the lower court level does not effectuate higher levels of opinion content adoption by UK Supreme Court justices.Instead, the result suggests that justices are just as likely to borrow language from lower court opinions whether they uphold or overturn the lower court decision.
Our final Supreme Court level attribute gauges the impact of a standard versus larger Supreme Court panel.The coefficient for Panel Size is statistically significant and signed in the positive direction.Recall that this variable is coded as 1 when the Supreme Court employs a standard panel configuration of five justices.The results indicate that opinion authors borrow more language from lower court opinions when the Supreme Court panel size is composed of five justices rather than larger panels of seven, nine, or eleven.The standard five-justice panel produces opinions where, on average, approximately 9.1% of the language is borrowed from the lower court opinion.Panels of more than five justices, however, produce opinions where approximately 6.8% of the language comes from the lower court opinion.Because larger panel sizes indicate that a case is of great public importance (i.e., salient), this finding is consistent with research on the American courts showing that Supreme Court justices tend to rely less on lower court opinions in cases that address salient issues (Corley, Collins, and Calvin 2011).
Turning to the Court of Appeal level attributes, we do not find any evidence that UK Supreme Court justices borrow more language from unanimous lower court opinions compared to opinions that are nonunanimous.The variable Unanimous has a positive coefficient, yet it is statistically indistinguishable from 0, meaning that agreement among the full panel of lower court judges regarding the outcome of a case does not motivate the justices to borrow more language from the lower court opinion compared to when the lower court decision includes a dissenting opinion. 27This result suggests that neither lower court consensus nor the availability of alternative rationales by dissenting judges has any discernible effect on language borrowing tendencies among Supreme Court justices in the UK.
We do, however, find evidence supporting our expectation that the length of a lower court opinion should affect Supreme Court language borrowing patterns.The estimate for Opinion Length is statistically significant and positively signed.Consistent with research showing that US judges borrow more language from longer lower court opinions, so too do UK justices (Corley, Collins, and Calvin 2011; Bowie and  Savchak 2022).Figure 2 illustrates the substantive effect of lower court opinion length on Supreme Court language borrowing. 28The results demonstrate that going from the minimum to the maximum value of the Opinion Length variable results in roughly a ten-percentage point increase in language borrowing, which represents a large effect.This result suggests that, ceteris paribus, lower court opinions codified in lengthy expositions and justifications of the outcome increase the tendency among Supreme Court justices to adopt the content of lower court opinions.
Finally, turning to our control variables, neither Supreme Court level controls -Freshman Justice and Caseloadreach conventional levels of statistical significance.These results indicate that justices who are newer to the Supreme Court do not experience acclimation effects such that they incorporate lower court language more extensively within their majority opinions compared to their more experienced peers. 29Relatedly, the number of cases that the Supreme Court reviews in a given term does not affect language borrowing patterns.This suggests that inevitable variations in the number of cases adjudicated by the Supreme Court, from one term to another, is not consequential in impacting the tendency of the justices to borrow language from lower court opinions.The two Court of Appeal-level controls are also statistically indistinguishable from 0 across all model specifications.Neither Reading Ease nor Passive Voice influence language borrowing patterns, suggesting that neither the textual clarity nor the readability of lower court opinions lead UK Supreme Court justices to borrow more or less language from lower court opinions. 30

Discussion
The practice of language borrowing is an area of study attracting increasing attention from judicial scholars, as the extent to which higher court justices use language from lower court opinions in crafting their own decisions offers important insights into factors affecting the behavioral dynamics of the judicial hierarchy on both individual and institutional levels.However, although the most prevalent study finds that the US Supreme Court borrows extensively from Courts of Appeal decisions (Corley,  Collins, and Calvin 2011), opinion borrowing tendencies have not previously been explored outside the American context.Given this apparent gap in the literature, our study contributes to the literature by systematically assessing whether opinion borrowing patterns are transferable across comparative settings, and if so, under what circumstances.By exploring how UK Supreme Court justices incorporate lower court language within their own opinions, we offer new insights into opinion borrowing tendencies and bottom-up influences, more broadly, from a comparative perspective.
Our results demonstrate that justices on the UK Supreme Court borrow language from lower court opinions at higher rates compared to the justices on the US Supreme Court.Yet, the factors that motivate opinion borrowing in the US judiciary are not always directly analogous to the UK judicial system, and for good reason.The UK judiciary emphasizes a career-based path for elevation and a strong deference to the principle of stare decisis, both of which form the basis of the theoretical framework through which we attempt to understand language borrowing practices by the UK Supreme Court.There are key institutional idiosyncrasies that increase language adoption tendencies in the UK, including the fact that elevation to the higher courts is predicated on lengthy judicial experience, the UK Supreme Court's ability to oversee a comparatively small Court of Appeal, the requirement for pseudo-random panels for review created from a Supreme Court with low membership stability, and the potential for forming strong networks of peer judges all working within the same city and building (Masood and Lineberger 2020).
At the same time, our results show that language borrowing tendencies do not always manifest in different ways in the UK compared to the US context.We find that justices in the UK are not particularly motivated to borrow at greater rates when the lower court issues a unanimous decision.And, simply affirming the lower court decision does not lead to larger amounts of language borrowing at the Supreme Court.Both of these results reflect findings from scholars researching language borrowing among the US federal and state courts.Our findings are also consistent with findings on the American courts because the justices in both countries distinguish cases of great importance by crafting original opinions in order to address the most pressing issues of the day.Despite differences in both the size of each country's judiciary and the extent to which politics may infiltrate the judicial process, case salience is uniformly seen as a priority among these two sets of common law justices.Meanwhile, Supreme Court justices across judicial systemsincluding those in the UKare more inclined to borrow language from longer lower court opinions.
Our analysis of opinion borrowing in the United Kingdom offers a number of contributions to the study of law and courts.Our work offers a unique account into bottom-up influences within comparative courts.In doing so, we believe that our results highlight the importance of continuing to explore language use in written opinions, as we find both similarities and differences in the factors motivating opinion borrowing tendencies among judicial systems.Second, we bring to light empirical evidence demonstrating that although opinion borrowing is common in the UK judicial system, what motivates language adoption is not the ideological factors that seem to define opinion borrowing in the US.Rather, factors grounded in the institutional norm of deference to existing precedent motivate language borrowing.This finding should provide an opportunity for future work to consider additional factors that affect borrowing throughout the UK judiciary in order to improve our understanding of how institutional norms and constraints might motivate higher court judges to adopt lower court opinion language.As such, our analysis should serve as a catalyst for further examinations on how factors such as judicial specialization, issue area, an opinion author's identity characteristics, and the availability of precedent may also influence language adoption from lower court decisions into higher court opinions.Finally, this study can provide a useful roadmap for scholars to analyze opinion borrowing and judicial behavior more generally across other countries beyond the US and UK.Despite important advances in comparative courts, there remain critical gaps in our collective understanding of institutional norms and judicial decisionmaking outside of the US courts.Our work demonstrates that it is critical for scholars to consider country-specific idiosyncrasies in research designs exploring hierarchical interactions within judiciaries.
From a comparative perspective, the data from our study indicate that opinion borrowing by the court of last resort is a practice ubiquitous within the United Kingdom.Yet, our work also demonstrates that there is a more nuanced distinction that must be drawn between the UK and other national judicial systems: namely, that the ideological predilections that incline justices to borrow language from the Courts of Appeals in the US do not motivate opinion borrowing in the United Kingdom.The lack of influence that ideology has on the likelihood for UK justices to borrow from lower court opinions has two important implications.First, it suggests that ideology does not play an outsized role in the decisionmaking or opinion writing process for judicial actors in the UK, as justices show no indication of extensively incorporating the language of lower court judges with whom they share similar ideological leanings.Second, the absence of meaningful support for ideological congruence corroborates our theoretical expectation that opinion borrowing in the UK is more prominent than it is in the US, as the presence of a career-based judiciary that relies on experience rather than partisanship or ideology as the key determinants for membership leads to a norm in which judicial decision-makers focus on adhering to legal doctrine rather than evincing tendencies of policy entrepreneurs.

Figure 1 .
Figure 1.Percentage of UK Supreme Court Majority Opinions with Borrowed Language, by Author, 2009-2019.

Figure 2 .
Figure 2. Impact of Opinion Length on Language Borrowing in the UK Supreme Court.Note: The effects are based on the estimates from Model 3. To plot these effects, we generate the predicted probability across all real values in the data.The solid line represents the predicted opinion percentage borrowed by the UK Supreme Court from the UK Court of Appeal.The shaded area represents the 95% confidence intervals.

Table 1 .
Language Borrowing in R v. Gnango

Table 2 .
Regression Models of Language Borrowing at the UK Supreme Court, 2009-2019

Table A3 .
OLS Regression Model with Raw Number of Words for Opinion Length :The unit of analysis is the U.K. Supreme Court -Court of Appeal opinion dyad.The dependent variable is the percentage of the lower court opinion that is borrowed by the U.K. Supreme Court opinion with the shortest phrase match set at ten words.Model includes Justice and Year fixed effects with bootstrap standard errors from 1000 bootstrap replications.*p < .05(one-tailed tests). Note

Table A2 .
OLS Regression Model with Six-Word Percentage Match Dependent Variable : The unit of analysis is the U.K. Supreme Court -Court of Appeal opinion dyad.The dependent variable is the percentage of the lower court opinion that is borrowed by the U.K. Supreme Court opinion with the shortest phrase match set at six words.Model includes Justice and Year fixed effects with bootstrap standard errors from 1000 bootstrap replications.*p< .05(one-tailedtests).Cite this article: Bowie, Jennifer, Ali S. Masood, Elisha C. Savchak, Natalie Smith, Bianca Wieck, Cameron Abrams, and Meghna Melkote.2023."LowerCourt Influence on High Courts: Evidence from the Supreme Court of the United Kingdom."JournalofLaw and Courts, 1-22, doi:10.1017/jlc.2023.18 Note

Table A5 .
OLS Regression Model with Wild Cluster Bootstrap : The unit of analysis is the U.K. Supreme Court -Court of Appeal opinion dyad.The dependent variable is the percentage of the lower court opinion that is borrowed by the U.K. Supreme Court opinion with the shortest phrase match set at ten words.Model includes Justice and Year fixed effects with wild cluster bootstrap followingCameron  et al. (2008).Model generates p-values instead of standard errors.*p < .05(one-tailed tests). Note

Table A4 .
OLS Regression Model with Alternate Measure for Ideological Distance : The unit of analysis is the U.K. Supreme Court -Court of Appeal opinion dyad.The dependent variable is the percentage of the lower court opinion that is borrowed by the U.K. Supreme Court opinion with the shortest phrase match set at ten words.Model includes Justice and Year fixed effects with bootstrap standard errors from 1000 bootstrap replications.The alternate measure for ideological distance is based on the absolute value of the difference between the party of the appointing Prime Minister and the ideological direction of the Court of Appeal decision *p < .05(one-tailed tests). Note