The Interpretation and Implementation of Supreme Court Decisions

Research on the impact and implementation of judicial policies spans more than a half-century (Peltason Reference Peltason1971; Baum Reference Baum1976; Canon and Johnson Reference Canon1984; Hall Reference Hall2010). Canon and Johnson (Reference Canon1984) provide a useful framework that conceptualizes the implementation of legal policies, offering four distinct “populations” that respond to judicial decisions, two of which are relevant to this study. The interpreting population consists of lower courts whose decisions broaden or restrict the application of the Court’s precedent. The implementing population, by contrast, consists of those government actors who have a responsibility to carry out policies and “apply the system’s rules to persons subject to their authority” (19). They frequently include prosecutors, bureaucrats, and public school officials, among others.

While a burgeoning literature exists on interpreting population responses to Supreme Court policies, research on the implementing population is limited by comparison. The examinations of the US Courts of Appeals (Songer, Segal, and Cameron Reference Songer1994; Brent Reference Brent1999, Reference Brent2003; Benesh and Reddick Reference Benesh2002; Haire, Lindquist, and Songer Reference Haire2003; Luse et al. Reference Luse2009; Westerland et al. Reference Westerland2010; Gizzi, Curtis, and Boldt Reference Gizzi2014) and state supreme courts (Martinek Reference Martinek2000; Benesh and Martinek Reference Benesh2002; Hoekstra Reference Hoekstra2005) have shown broad compliance and responsiveness regarding Supreme Court decisions with a context-dependent ability to resist change.

Most studies that are focused on the implementation of court decisions have evaluated controversial Warren Court cases involving such issues as school district desegregation, school prayer, and police interrogations; they have generally found that Court-issued reforms required significant actions by others (executives, activists, and bureaucracies) to have them implemented in a meaningful way (Dolbeare and Hammond Reference Dolbeare1971; Milner Reference Milner1971; Bullock and Rodgers Reference Bullock1976; Rosenberg Reference Rosenberg2008). The scope of these foundational studies may be constrained by the context of the decisions being analyzed, where resistance to their implementation was especially prominent. Other research has shown greater responsiveness and compliance by the implementing population. Spriggs (Reference Spriggs1996, Reference Spriggs1997) conducted examinations of the impact of the Supreme Court’s decisions on federal administrative agencies, finding no evidence of defiance or evasion in reaction to the Supreme Court.

Among the most far-reaching and generalizable analyses of judicial impact on the implementing population is Hall (Reference Hall2010). He asserts that two conditions explain the outcome of Supreme Court policies, whether the issues being handled are “vertical” or “lateral” in nature. Vertical issues like abortion or the right to counsel are those that need only the support of lower courts to carry out a shift in policy. Lateral issues, meanwhile, such as reapportionment or affirmative action, necessitate the actions of those outside the judicial branch to be successful. Hall finds that decisions in vertical areas of the law are often implemented effectively. Those in lateral areas rely more heavily on public opinion, requiring the citizenry to keep pressure on key implementers to conduct change.

Why Gant and Police Searches?

This study supplements the foundational qualitative judicial implementation work from the 1960s and 1970s that examined a few jurisdictions in great depth (e.g., Dolbeare and Hammond Reference Dolbeare1971; Milner Reference Milner1971) and the more recent innovative cross-sectional studies that test the influence of numerous court decisions on a variety of implementers (e.g., Spriggs Reference Spriggs1996; Hall Reference Hall2010). It focuses on a single court decision but, unlike the earlier research, uses large data sets that include approximately 34 million police-citizen interactions across thousands of state, county, and local law enforcement agencies in two large states. And, unlike the recent large-n studies, the nature of the data allows for extremely fine-grained analysis considering the impact of the Supreme Court in the time immediately following its decision.

Police organizational responses to changes in judicial policy by the Supreme Court present a unique context to view the power of the Court. Contrary to its hierarchical relationship with lower courts, the Supreme Court has no direct ability to supervise the decisions of law enforcement. The Court must rely instead on intermediaries and the rule of law to ensure its will is carried out. Many police searches occur outside the supervision of the judicial branch altogether unless their actions are contested as part of a criminal prosecution or suit involving police liability.

Prior research on the police and Supreme Court decisions suggests that the relationship is sometimes fraught with tensions. Skogan and Meares (Reference Skogan2004) reviewed the body of work on police compliance with criminal procedure. They explained that this literature shows that while law enforcement tends to obey the law, “they sometimes skirt constitutional standards because they want to deter crime by incarcerating the truly guilty” (73). Studies have provided evidence for major shifts in police practice following court opinions and patterns of resistance by police to limits on their authority.

Several studies argued that Miranda v. Arizona (384 U.S. 486 [1966]) substantially reduced rates of questioning, confessions, and case clearances, indicative of widespread implementation of the Court’s opinion (Cassell and Hayman Reference Cassell1996; Leo Reference Leo1996; Cassell and Fowles Reference Cassell1998). Yet, the results of other studies suggest that law enforcement did not always alter their behavior when questioning suspects. Reiss and Black (Reference Reiss1967) and Milner (Reference Milner1971) asserted that many officers observed during their studies did not issue required Miranda warnings.

In the area of search and seizure, several studies have examined whether police have followed precedents established by the Supreme Court. Following the application of the exclusionary rule to the states in Mapp v. Ohio (367 U.S. 643 [1961]), increases were found in the number of search warrant applications and motions to suppress in major cities, suggestive of the ruling’s implementation (Spiotto Reference Spiotto1973). Canon (Reference Canon1973) found declines in arrest rates in several major cities, implying that Mapp limited the ability of the police to gather incriminating evidence. However, one study conducted by Columbia University law students found that the proportion of narcotics cases in which officers claimed to have found evidence in plain view rose dramatically after the decision. This finding led the researchers to argue that law enforcement was “fabricating the grounds of arrest in narcotics cases in order to circumvent the requirements of Mapp” (Hoffman and Seltzer Reference Hoffman1968, 95). Fagan and Davies (Reference Fagan2000) and Gould and Mastrofski (Reference Gould2004) also found that large numbers of police searches in their studies were performed unconstitutionally.

The presence of noncompliance across a variety of prior studies suggests that this project could represent a hard test of the Supreme Court’s reach. Gant upended nearly 30 years of police search and seizure practices. The degree to which local police comply with a ruling that limits their discretion to obtain evidence and conduct criminal enforcement speaks directly to the ability of the Court to lead policy change. Further, the capacity of law enforcement to shirk, evade, or find alternative means to conduct searches may shed light on the ability of other government organizations to pursue their own interests after the Court has spoken.

The Development of Arizona v. Gant

The Fourth Amendment doctrine regarding warrantless searches incident to legal arrest has been a crucial tool used by law enforcement to conduct searches of automobiles for decades (see, e.g., Judge Reference Judge2009; Schott Reference Schott2009; Gizzi and Curtis Reference Gizzi2016). The Court has long permitted police to search the person of an arrestee and the area within their immediate control for officer safety and to prevent the destruction of evidence (Chimel v. California, 395 U.S. 752 [1969]; United States v. Robinson, 414 U.S. 218 [1973]). The subjectivity of the reaching-distance rule created questions for law enforcement about the scope of a search incident to arrest when the person arrested was in a vehicle. For police, these questions were resolved when the Court decided Belton. Police interpreted the decision as a per se rule that permitted the search of a vehicle’s entire passenger compartment incident to arrest.

As applied, the Belton rule eliminated the subjectivity of the Chimel reaching-distance test by permitting the search of a vehicle’s passenger compartment after every arrest regardless of the seriousness of the offense or how reasonable it was to expect to find evidence. It created a lower threshold to search for contraband than was permitted by the automobile exception, which allowed a warrantless search of a vehicle only when the officer had probable cause that the vehicle contained evidence of criminal activity (Carroll v. United States, 267 U.S. 132 [1925]; United States v. Ross, 456 U.S. 798 [1982]).

The true value of the Belton search came in waging the war on drugs. Police used the pretext of a minor traffic violation to conduct criminal investigations. If the officer suspected that the suspect’s vehicle contained contraband but lacked probable cause, an arrest for a minor offense permitted a vehicle search. The Drug Enforcement Agency trained more than 30,000 officers in how to make extensive use of the tools the Court provided (Gizzi and Curtis Reference Gizzi2016). Vehicle searches became a primary means by which the war on drugs was fought.

The Belton search became a central part of police training (Remsberg Reference Remsberg1995; Epp, Maynard-Moody, and Haider-Markel Reference Epp2014) and netted significant victories for law enforcement, but over time, some of the justices began to assert that its application was disconnected from the original rationale that was used to justify search incident to arrest: officer safety and the preservation of evidence.Footnote ¹ The majority in Gant found that most vehicle searches incident to arrest involved no officer safety concern after the arrestee was secured and removed from the vehicle, and for many types of offenses, there was no evidence to preserve.

The Supreme Court’s decision to grant certiorari in Gant ran counter to years of police practice (Judge Reference Judge2009). The case included a search that was a prototypical Belton search after an arrest for driving on a suspended license. Rodney Gant challenged the search, claiming there were no officer safety concerns or any need to search for evidence, given what he was arrested for (Arizona v. Gant, 556 U.S. 332 [Reference Judge2009], 335–38). In a 5–4 decision, the Supreme Court agreed with Gant, and Justice Stevens modified the Belton rule, replacing it with a two-pronged standard. Under the first prong, the Court held that incident to arrest, a warrantless search of the arrestee’s person was always permissible, but a vehicle search was allowed only under the rare circumstance in which the arrestee was unsecured and could gain access to the vehicle. Meanwhile, the decision’s second prong permitted a search of the passenger compartment only if it was reasonable to believe that further evidence of the crime of the arrest would be found. This meant that arrests for minor offenses like driving under a suspended license or driving without proof of insurance could no longer be used as the pretext to conduct a passenger compartment search.

The decision sent shockwaves throughout the law enforcement community. Police publications were vocal about the restrictions imposed by the ruling and about finding ways to adapt to those limitations to continue to conduct vehicle searches. The May 2009 issue of the Federal Law Enforcement Training Center’s Federal Law Enforcement Informer devoted 10 pages to the Gant ruling. In the article, an instructor at the center examined how Gant affected the doctrine of search incident to arrest and also discussed, in detail, ways in which officers may conduct searches by relying on other established exceptions to the warrant requirement. The author wrote, “Gant will have an immediate impact on the day-to-day operations of law enforcement officers in the field… . If an officer cannot justify a search of a vehicle incident to arrest under Gant, or is uncertain whether an SIA is warranted, other established exceptions to the search warrant requirement remain available to safeguard evidence and protect the safety of officers” (Solari Reference Solari2009, 4, 8). Despite noting the substantial limitations to conducting vehicle searches, the author indicated that many other warrant exceptions remained available to officers. Specifically, she said that officers might use protective frisks (Michigan v. Long, 463 U.S. 1032 [1983]), probable cause or the “automobile exception” (Carroll), consent of the driver to the search (Schneckloth v. Bustamonte, 412 U.S. 218 [1973]), or inventory searches (South Dakota v. Opperman, 428 U.S. 364 [1976]).

Similarly, the June 2009 issue of Police Chief, the official publication of the International Association of Chiefs of Police, featured a column on Gant by Lisa A. Judge (Reference Judge2009), who wrote, “To many state and federal courts and most of the law enforcement community, the decision marked a significant change to long-standing practices regarding the nature and scope of warrantless vehicle searches conducted incident to lawful arrest… . The Gant decision moves away from the easy-to-understand, bright-line rule of New York v. Belton, which was widely interpreted to authorize the routine warrantless search of a vehicle pursuant to the arrest of an occupant of that vehicle.” Judge continued by detailing a bullet-point list of exceptions to the warrant requirement police organizations may still pursue in addition to search incident to arrest. Included in the list were protective frisks, inventory searches, consent searches, and the automobile exception.

Research Design and Hypotheses

To examine the impact of Gant on law enforcement behavior, this study takes advantage of comprehensive traffic-stop data from two states. Statewide data sets from Illinois and North Carolina allow for a test of whether Gant actually limited searches incident to arrest and can shed light on whether alternative rationales were relied upon in order to conduct searches after the decision.

Several aspects of the case’s impact are of interest. The first is whether the Supreme Court actually changed officers’ reliance upon the search-incident-to-arrest exception to the warrant requirement when searching automobiles. To test this, the rates of search incident to arrest are subjected to time series intervention analyses to see whether the decision had a statistically significant effect. Gant allows for a search incident to arrest of a vehicle only when there is a reasonable expectation of finding further evidence of the crime of the arrest or when the arrestee is unrestrained and within reaching distance of the vehicle. While not eliminating the search-incident-to-arrest exception, this was a substantial shift from allowing a search after any arrest. It is expected that police largely complied with the restriction, thus reducing the rates of these searches. This expectation is stated in hypothesis 1. The strong reaction by law enforcement publications to the decision implies that police took its new limits seriously. While these articles were often critical of Gant, they instructed officers on how to follow the letter of the ruling. Further, the role of law enforcement organization supervisors and trial courts in monitoring police behavior seems likely to constrain officers from conducting illegal searches under the invalidated Belton rule. Not following the new search-incident-to-arrest standards could result in professional sanction or having evidence excluded by trial judges.

While it is anticipated that Gant reduced vehicle searches incident to arrest, it should not prevent the warrant exception from being used when the circumstances permit it. Officers routinely make arrests in circumstances in which it may be reasonable to find further evidence of the offense in the vehicle (e.g., drug possession, driving under the influence, illegal weapons possession).

Following searches incident to arrest, this study then considers a second aspect of Gant’s judicial impact. Police publications following the decision were vocal about finding other legal means to conduct vehicle searches. Accordingly, rates of alternative search types are examined and tested to see whether police sought other ways to continue doing searches in the new policy environment. Officers have several avenues available for conducting warrantless vehicle searches other than search incident to arrest. For example, if an officer knows that an arrest for driving without proof of insurance will not allow him to search the vehicle, he might ask for the consent of the driver (Schneckcloth) or call a police dog to sniff the exterior (Place). The officer could also elect to impound the vehicle to conduct an inventory search (Opperman). It is likely that even after their discretion has been limited, police officers are still motivated to conduct searches so that they may arrest suspected criminals and gather evidence for successful prosecutions. Thus, it is expected that alternative search rationales were increasingly relied upon after the holding. This expectation is stated in hypothesis 2.

If a reaction to Gant is found in any kind of police search, then several other questions need to be answered. First, how large of a change did Gant cause relative to before? Quantifying the effect of the decision will provide insight into how the language of the ruling translated into a substantive impact and, perhaps, to what degree officers are complying with the law. Second, how long does it take for officers to change their behavior in the face of a legal policy change? Whether any delay exists between the announcement of a policy and the adaptation of the policy provides valuable information about the process by which police learn about court decisions.

The Impact of Gant on Searches Incident to Arrest

Descriptive Results

Basic times series plots are displayed first to help demonstrate the purely descriptive evidence for the impact of Gant. The weekly series of searches incident to arrest for Illinois and North Carolina are displayed in figures 1 and 2. A dashed vertical line is placed the week following April 21, 2009, when Gant was decided. Purely upon visual inspection, it would be difficult to argue that Gant did not influence the rate of searches incident to arrest in traffic stops. Both series show a sudden and stable decline in the number of these searches the week of the decision.

Figure 1.

Illinois search-incident-to-arrest series

Figure 2.

North Carolina search-incident-to-arrest series

Statistical Time Series Results

Figures 1 and 2 provide preliminary evidence that Gant limited searches incident to arrest in these two states. To assess whether that effect is statistically meaningful, ARIMA/ARCH time series analysis models were fit to the weekly search-incident-to-arrest series in Illinois and North Carolina. Series were analyzed in first differences to account for nonstationarity, a common issue with time series.Footnote ⁸ The results of the time series intervention analysis model for the weekly search-incident-to-arrest series in Illinois is contained in table 4 and figure 3, while the corresponding model for North Carolina is in table 5 and figure 4.

Table 4.

ARIMA/ARCH Model, Illinois Search-Incident-to-Arrest Series

Note.—N = 415. Dependent variable is the first differenced weekly number of vehicle searches incident to arrest as reported by the Illinois Department of Transportation (http://www.idot.illinois.gov/transportation-system/local-transportation-partners/law-enforcement/illinois-traffic-stop-study). This model produces white noise residuals with a Ljung-Box Q test p-value of .26 at 53 lags. Seasonality fixed effects included. AR = autoregressive term. MA = moving average term.

*. p < .05.

Figure 3.

Predicted values from ARIMA/ARCH model, Illinois search-incident-to-arrest series.

Table 5.

ARIMA/ARCH Model, North Carolina Search-Incident-to-Arrest Series

Note.—N = 623. Dependent variable is the first differenced weekly number of vehicle searches incident to arrest as reported by the North Carolina State Bureau of Investigation (http://trafficstops.ncsbi.gov). This model produces white noise residuals with a Ljung-Box Q test p-value of .20 at 53 lags. Seasonality fixed effects included. AR = autoregressive term. MA = moving average term.

*. p < .05.

Figure 4.

Predicted values from ARIMA/ARCH model, North Carolina search-incident-to-arrest series.

Searches incident to arrest in Illinois decreased by approximately 688 per week as a result of the Gant decision, all else equal. This coefficient is significant, with a z-score of −14.78 and a p-value approaching 0. The average number of weekly searches prior to Gant in Illinois was approximately 1,685. Thus, Gant decreased the weekly rate of search incident to arrest in Illinois by nearly 41%.

These results demonstrate that in Illinois the decision imposed substantial limitations on police officers’ ability to perform searches incident to arrest after traffic stops. Interestingly, the estimate of the effect of Gant on the conditional variance indicates that the decision significantly decreased the volatility of the series. There is less variation in the number of weekly searches incident to arrest after Gant than before it. It may be that by narrowing the discretion with which officers could use the search-incident-to-arrest exception, the Supreme Court’s holding induced more uniformity in the number of these searches week to week.

The results from North Carolina are substantively similar to those from Illinois. The effect of Gant on searches incident to arrest in North Carolina is a decrease of approximately 192 per week, all else equal. This coefficient is significant, with a z-score of −19.11 and a p-value approaching 0. The average number of weekly searches prior to Gant in North Carolina was 322. Gant decreased the weekly rate of search incident to arrest in North Carolina by nearly 60%. Much like in Illinois, Gant substantially reduced the rate of searches incident to arrest as a result of traffic stops. The estimate of the effect of Gant on the conditional variance is similar to that of the Illinois model. Again, Gant appears to have decreased the variation in search incident to arrest in these data.

Overall, these results provide clear evidence of judicial impact on this implementing population. The Court’s modification to search-incident-to-arrest doctrine in Gant resulted in a strong and immediate (within 1 week) effect on the behavior of law enforcement. Police officers adapted to the new legal environment, quickly altering their search practices. This suggests that police organizations pay attention to and abide by changes to Supreme Court precedent. However, since the Gant ruling allows for searches incident to arrest when there is a reasonable expectation of finding further evidence of the crime of the arrest in the vehicle or when the arrestee may gain access to the vehicle, it does not eliminate the use of this search rationale in either state. The diminished presence of search incident to arrest after the ruling implies that officers are still conducting these searches in situations in which the Supreme Court has permitted them to do so.

While the immediacy of the effect is surprising, and one might expect at least a small lag in time for police to familiarize themselves with the new law, the sudden shift in behavior does make some intuitive sense. From the date the Supreme Court changed the Fourth Amendment doctrine, officers who do not learn and apply the new standard risk having evidence excluded in court. That could potentially result in accused criminals who may otherwise have been convicted having their cases dropped or their charges dismissed or being found not guilty. In addition to draining limited resources, such occurrences may hurt an officer’s professional standing or public reputation.

These results also shed light on how police learn about court rulings. In learning to abide by a new precedent, law enforcement can be proactive by applying the rule of their own accord, or they can be reactive by changing their behavior only after trial courts have sanctioned them for not following the decision. The immediate timing of the reaction to Gant strongly suggests the former, with law enforcement independently reacting to the changes to search-incident-to-arrest requirements. The consistency of the results across the two states, with large and permanent changes to search rates, also implies organizational systems for police to become educated about relevant changes to the law. If officers were expected to learn about the Gant decision individually, then one would expect a period of delay during which increasing numbers of police were becoming aware of the opinion. Instead, the decrease in searches incident to arrest is sharp.

The Impact of Gant on Alternative Searches

Descriptive Results

Law enforcement appears to have largely complied with the limitations expressed in Gant, as evidenced by the substantial decrease in searches incident to arrest. However, this does not exclude the possibility that officers looked for alternative ways to conduct searches using other established exceptions to the warrant requirement. Law enforcement publications written after Gant detailed several ways for police to circumvent the limitations it expressed and find alternative justifications for searches without relying on search incident to arrest (Judge Reference Judge2009; Solari Reference Solari2009). Thus, it seems reasonable to expect that many officers found ways to adapt to the new legal policy.

To determine whether police turned to other search rationales in the wake of Gant, rates of alternative searches are examined. The Illinois data include several alternative search categories: consent, K-9, other, probable cause, and reasonable suspicion. North Carolina officers record data on consent, probable cause, protective frisks, and search warrants. The plots of Illinois’s alternative search series are displayed in figure 5. In Illinois, the “other” search was not included as a category in the traffic-stop data until 2007, resulting in a shorter series. The plots of North Carolina’s alternative search series are contained in figure 6. Several of the series display visually striking changes after the week of April 21, 2009. In Illinois, the rate of the “other” category of searches increases the week of the decision and stays elevated until 2012.

Figure 5.

Illinois alternative search series

Figure 6.

North Carolina alternative search series

In North Carolina, probable cause searches and protective frisks have a sudden increase that occurs at the beginning of 2010. Consent searches appear to decline starting in 2011, roughly 2 years after Gant. Upon first encountering the changes in these three series, it might lead one to conclude that Gant had an impact on these kinds of searches in North Carolina. However, the notion that Gant caused these shifts can be discounted owing to their timing and functional form. It is much more likely that these changes are the result of some other unknown factor. Supreme Court rulings are binding from the date they are decided (in this case, April 21, 2009). Accordingly, the timing of any change to police behavior should begin at or very near that date. Instead, these changes occur 36 weeks after the decision in the probable cause and protective search series. The consent search series does not experience a structural change until 148 weeks later. Each of these shifts also occurs abruptly, with the seasonal pattern noticeably altered at a particular week. It is implausible that officers rapidly changed their search practices as a result of Gant months after the ruling. Time series models will provide further evidence of whether any of these changes is a result of the ruling.

Statistical Time Series Results

To determine whether Gant increased rates of alternative searches, time series models were fit to all but one of the remaining categories in the Illinois and North Carolina traffic-stop data. A model was constructed for each weekly series of searches, with the exception of search warrants in North Carolina, which were too rare to merit inclusion. Because the 2010 increase in North Carolina’s probable cause series and protective search move in the same direction as the hypothesized impact of Gant, they are accounted for through a distinct intervention variable to be more conservative.

Only the “other” search series in Illinois obtained significant results. The results of the intervention model on the “other” type of search from Illinois are reported in table 6, and the predictions from the model are displayed in figure 7.Footnote ⁹ For all remaining search types, the result of the Gant intervention variable was not significantly different from 0 (p > .05). The tables for these nonsignificant models are reported in the appendix.

Table 6.

ARIMA/ARCH Model, Illinois Other Search Series

Note.—N = 259. Dependent variable is the first differenced weekly number of vehicle searches incident to arrest as reported by the Illinois Department of Transportation (http://www.idot.illinois.gov/transportation-system/local-transportation-partners/law-enforcement/illinois-traffic-stop-study). This model produces white noise residuals with a Ljung-Box Q test p-value of .18 at 53 lags. Seasonality fixed effects included. AR = autoregressive term.

*. p < .05.

Figure 7.

Predicted values from ARIMA/ARCH model, Illinois other search series

The results from the Illinois “other” search intervention model also demonstrate an immediate adaptation to Gant. This suggests that officers “think on their feet” after a limiting court precedent, using their knowledge of search warrant exceptions to justify some searches by using alternative means. Since the “other” search is a catchall category for any search not included on the Illinois traffic-stop form, it may incorporate inventory searches, searches of persons under correctional supervision, or searches based on warrants. Whatever specific search criteria might fall under this category, when the opinion in Gant was announced, the rate of “other” searches increased by nearly 91 per week. The mean of this category of searches prior to the intervention was 82 per week. Thus, the rate of “other” searches increased as a result of Gant by approximately 111%. Interestingly, even though the rise of these searches is dramatic, the increase only partially counters the loss of searches incident to arrest in Illinois (about a 13% makeup effect).

This alternative search justification was used more heavily as a result of Gant in a way that is both visually striking and statistically significant. No significant change is found in any of the remaining alternative search series, yet it may be reasonable to expect increases that are idiosyncratic. Police using alternative rationales to conduct searches must justify their actions in a way that will satisfy others (e.g., supervisors, prosecutors, and judges) that they have met existing legal requirements. Simply because an alternative exists does not mean an officer can rely on it in every circumstance. Despite some evidence of police turning to alternative rationales, the finite number of search categories reported in these data sets, the lack of thorough information on what they specifically mean, and the fact that the two data sets do not contain the same search categories make it difficult to ascertain whether this result is truly reflective of police behavior. As such, this study offers what should be considered preliminary evidence on the question of whether police look for alternative justifications to conduct searches after the Supreme Court limits their discretion.