Abstract
States have increasingly taken the process of redistricting out of the hands of elected legislators and placed it with the public. The shift is in part driven by a concern that legislators are motivated to partition districts to advantage their own and their political party’s electoral prospects, whereas citizens are not. We know little, however, about the preferences of the public when it comes to redistricting. One party-based argument is that individuals should prefer to share a district with as many like-minded partisans as possible to maximize their legislative representation, whereas other arguments suggest that nonpartisan factors, such as sharing a district with their community, may be more important. Using a novel experimental design, we find that for most participants, the draw to share a district with copartisans is stronger than a preference for preserving a community (county) within the district even when participants are specifically instructed to attend to local jurisdictional boundaries.
In recent decades, states have increasingly taken the process of redistricting—that is, the drawing of congressional and legislative district boundaries, which follows the decennial U.S. Census—out of the hands of elected legislators (Altman and McDonald 2011; see also Altman et al. 2010). In 2010, California implemented a citizen commission to handle redistricting following the move of Arizona to an independent-based commission. That same year, Florida, via constitutional amendment, banned political gerrymandering by not allowing the use of partisan or incumbency information during the redistricting process. Other states, such as Texas, Georgia, and Virginia, have discussed ways to increase transparency and public involvement to combat political and partisan manipulation of congressional and legislative districts. The shift is in part driven by a concern that legislators are motivated to draw districts that advantage their own and their political party’s electoral prospects—a motivation that citizens ostensibly do not share. 1
These policies raise a compelling theoretical question about the extent to which elected leaders have different considerations about the redistricting process than citizens. Historically, redistricting has been an elite game. Despite the ramifications for the public, citizen awareness and knowledge of the process that determines their electoral boundaries are low. For example, a 2006 Pew Center Survey showed that half of registered voters did not know who was in charge of drawing district lines and knew nothing at all about the debates over how to draw boundaries. In addition, a full 70 percent had no opinion about the process used in their state. 2 No doubt, part of this is by design as political elites have traditionally completed the process behind closed doors away from public view as part of a “largely invisible matrix of election practices that influences our lives” (Overton 2006, 27). Although interested elites behave as if redistricting is a means to their electoral fortunes, the public generally ignore this process and have little knowledge of both the outcomes generated by and ramifications associated with redistricting.
This lack of public awareness raises concerns about whether elite-drawn districts reflect the same preferences citizens would use if they could design their own district boundaries. Although there is much debate among observers as to what the underlying goals in the redistricting process should be (and about the best method for achieving those goals), the preferences of the elected officials who draw the lines are largely clear: maximizing partisan seat shares and/or incumbent protection (Born 1985; Cox and Katz 2002). In the eyes of many, the gerrymandering that follows nefariously plays to the advantage of elites who control the process and the disadvantage of the citizenry. An implicit assumption of reform movements designed to give more power to the public in the redistricting process is that the public would make different decisions than elites if it had the opportunity to draw districts (Altman and McDonald 2011; Lowenstein and Steinberg 1985). Altman and McDonald’s (2011, 2012, 2013, 2015) work in this area, for example, along with their analysis of plans in several states, including Virginia and Florida, show that citizen-submitted plans often differ from legislative plans on important dimensions such as partisan competitiveness and respect for county boundaries.
In this paper, we use an experimental design to examine citizen preferences when given two relevant considerations for selecting their preferred district configuration. Specifically, we examine a traditional redistricting principle of protecting political subdivisions (counties in our experiments), along with the partisan makeup of the districts. 3 The results of two experiments indicate that use of county integrity as a criterion for selecting a preferred district configuration fades when people have access to partisan information, even when they are explicitly reminded to preserve geographical boundaries to the extent possible.
These experimental results offer an important contribution for understanding the views of citizens toward redistricting. As more states move toward greater inclusion of citizens in the process, from citizen-led commissions to public hearings and citizen submission of plans, we know very little about the outcomes citizens would actually prefer. This study takes a step in that direction by examining how one important consideration, whether to split a community of interest, fares when partisan considerations are brought to bear on decisions about district boundaries. The redistricting process is complex and multifaceted, and this study examines just two specific (but important) considerations for drawing districts and does not speak to preferences about the overall map outlining multiple districts or to perceptions of procedural fairness. Nevertheless, our results suggest that citizen-led processes may not foster greater protection of communities of interest and local jurisdictional boundaries within districts. Indeed, when given the opportunity to think about the implications of their district’s boundaries for its partisan composition, citizens appear to prefer districts where their party holds the majority over districts that recognize local jurisdictional boundaries.
Redistricting Principles and Goals
At its most basic level, the constitutional purpose of redistricting is straightforward and generally noncontroversial—to create electoral districts of equal population size for selecting representatives. In a series of cases from the reapportionment revolution of the 1960s, the Supreme Court established that the Equal Protection Clause of the U.S. Constitution requires electoral districts to contain equal population sizes in Congressional districts and roughly equally populated districts in (state) legislative redistricting. On its own, however, the equal population criterion allows tremendous latitude in the drawing of district boundaries because there are many ways a legislature or other body can draft maps satisfying this requirement.
This latitude opens the door for drafting district boundaries in accord with other criteria while still meeting the equal population criterion. The attention legislatures and other bodies charged with redistricting give to political considerations such as incumbency protection and partisan advantage is perhaps most widely recognized (Miller 2013). Yet, over the years, a set of important criteria, conventionally known as traditional redistricting principles, has emerged to guide decision making in accord with various understandings of fairness (see Levitt 2010 for an overview of these principles). Many of these principles are explicitly geographic, such as compactness, contiguity, and preservation of local political subdivisions or communities of interest. 4
Shared geography provides a goal of representation based on the commonality of where people live rather than overt political concepts. The potential for positive representational gains from this shared geography is a driving factor for pushing these geographic principles in the redistricting process (Bowen 2014; Butler and Cain 1992; Engstrom 2005). Furthermore, geographic boundaries are important because they structure the laws and customs of an area that can, both directly and indirectly, alter the activity and behavior of the individuals living within those borders (Elazar 1994; Gimpel and Schuknecht 2004). For example, jurisdictional boundaries from state lines to school districts directly influence the public goods and services provided to individuals and the taxes those citizens are required to pay to support the activities of government. In addition, these boundaries have a direct link to the electoral process as they determine the slate of candidates for which a person is eligible to vote for, which in turn influences the resources and goods provided by the government.
Although the Supreme Court ruled in Baker v. Carr (369 U.S. 186 (1962)) that states cannot base districts primarily on political subdivisions such as counties at the expense of equal population size, courts have subsequently recognized the importance of these geographic principles as secondary factors to equalizing population (see Miller v. Johnson, 515 U.S. 900 (1995); Shaw v. Reno, 509 U.S. 630 (1993); Carstens v. Lamm, 543 F. Supp. 68 (D. Colo. 1982)). In addition, these geographic criteria have been officially sanctioned by most states either as laws or as decision rules for drawing district maps. 5 Forty-four states have provisions either in their constitution, legislative law, or redistricting guidelines for protecting political subdivisions in the legislative redistricting process, while twenty-one also call for the protection of “communities of interest,” 6 and thirty-six include a compactness standard.
Of course, these geographic standards frequently clash with other considerations like partisanship and incumbency. For example, in the 2000 round of redistricting state legislative seats, just over 30 percent of counties were split into more districts than required under the equal population standard. Although rules emphasizing the importance of geographic integrity may have had some effect on redistricting outcomes—29 percent of counties were split into more districts than necessary based on the equal population standard in states with a specific rule against unnecessarily splitting political subdivisions, but 43 percent of counties were similarly split in states without such a rule—geography nevertheless appears to fall behind other considerations. In the 2010 redistricting of congressional seats (in which it is much easier not to split counties because there are fewer districts and larger district populations when it comes to Congress), 15 percent of counties were split into more districts than necessary, with 13 percent falling under states with a rule prohibiting such splits and 16 percent under states without such a rule.
In addition to policies directed at the outcomes of redistricting, reforms have also taken aim at the process. The 1980s saw a move toward more emphasis on the role of commissions to mitigate concerns about politicians determining their own district lines. Most recently, the reform movement has shifted more toward the role of the public in the process and bringing more transparency into an often-invisible institution. 7 In both the process-based movement toward commissions and the new open access movement, these traditional redistricting principles serve as important guidelines for those in charge. Concern over the political considerations of elected leaders has played a driving role in the call to limit their influence in the process (Donovan 2011).
Citizen Preferences in Redistricting
In general, citizen knowledge of the redistricting process is quite low (Fougere, Ansolabehere, and Persily 2010; Gaines 2011; Panagopoulos 2013). Most citizens are not aware of the intricacies of what is generally seen as a political insider’s game. However, citizens do show concern for both gerrymandering and uncompetitive elections, two of the major negative consequences associated with redistricting (Cain, Mac Donald, and McDonald 2005), and express support for traditional redistricting principles (e.g., compactness, protecting political subdivision boundaries, etc.) when given the information (Gaines 2011). The available evidence certainly suggests that there may be some disconnect between what people prefer and what elites draw. For example, there is some evidence that people find strangely shaped districts to be cognitively demanding and often deem them patently unfair (Ely 1997; Grofman 1993), whereas those involved in the process often dismiss bizarre or ugly district shapes as having no identifiable harmful effects on voters (Altman 1998; Karlan 1996). 8
More broadly, it remains unclear whether the primary considerations of the public differ from the primary considerations of elites who typically draw the maps. Individuals have psychological and social attachments to geographic places and spaces that define their community (Huckfeldt and Sprague 1995; Putnam 2000). 9 There is some evidence that citizens think geographically when evaluating redistricting plans. Gaines (2011), for example, shows that individuals emphasize geographic principles, but these tests are limited to evaluations of an overall plan and do not bring political considerations about one’s own district into play. This geographic thinking may not be as straightforward as linking district lines with specific boundaries as Christenson and Makse (2015) and Wong et al. (2012) show that individuals care about nonadministrative spatial boundaries that often do not overlap neatly with administrative boundaries, such as counties.
Moreover, there are reasons to suspect that preferences to keep geographic jurisdictions intact would fade when faced with explicitly political considerations. For instance, electoral losers prefer more competitive districts, whereas electoral winners hold a more favorable opinion of less-competitive elections (Brunell and Clarke 2012). Quite simply, individuals may be partisans first and members of their broader geographic community second. Therefore, these political preferences may trump any other concerns, much as we see with the elites when drawing the maps. Decades of research show the influence party identification has on political attitudes and behaviors (Bartels 2002; Campbell et al. 1960; Fiorina 1981; Gerber, Huber, and Washington 2010; Green, Palmquist, and Schickler 2002; Lupia 1992). In our view, whether citizens behave differently than elites when it comes to redistricting, especially in the face of partisan information, remains an open question.
Data and Experimental Design
We conducted two experiments that were similar in their design. 10 Each study was administered through Amazon.com’s Mechanical Turk (MTurk) interface. MTurk is a platform for recruiting individuals to work on specified tasks. The sample is not nationally representative but is more diverse than undergraduate student participant pools often used in experimental research and has become a common source of study recruits in social science research (Berinsky, Huber, and Lenz 2012; Buhrmester, Kwang, and Gosling 2011; Paolacci, Chandler, and Ipeirotis 2010). 11 Respondents who completed Study 1 were paid thirty-five cents to complete the survey, which was conducted in May 2014 (n = 1,428). Respondents who completed Study 2 were paid fifty cents to complete the survey, which was conducted in March 2015 (n = 1,363). In addition to taking part in the experiment, participants completed a short survey to collect demographic characteristics and political attitudes. In the analysis that follows, we focus our analysis on partisans (including independents who lean toward a particular party; final n = 1,148 for Study 1 and 1,132 for Study 2), that is, we exclude “true” independents. 12
The goal of each experiment was to have participants choose between two hypothetical district maps when faced with different information about those districts. Given that we are interested in identifying the (treatment) effect of information, we chose to use MTurk, despite its limitations in terms of generalizability (concerns that are present in any experiment that is not a randomized field trial) because the lab experimental design provides us with strong internal validity to identify the effect of information even if some concerns about generalizability remain.
In each experiment, all participants were shown two images in which hypothetical congressional district boundaries were overlaid onto a map of hypothetical counties (see the online supplemental material for all pairs of maps used in each study and complete question wording). Participants were asked to imagine that they lived at a specified location marked on the maps (the same location on both maps) denoted by a blue star and select their preferred district configuration between a pair of adjacent congressional districts. The two maps were identical except for the placement of this district boundary line, which was a dark red line made to stand out to respondents.
13
In one map (Option 1), the boundary between districts ran along county lines, leaving the respondent’s hypothetical county (Washington) entirely contained within a single congressional district. In the other map (Option 2), the boundary cut the participant’s hypothetical county in two, placing the participant in a different congressional district than in Option 1. This difference between the maps was made clear to participants in a number of ways. In addition to the dark red line denoting the different districts, which was also made more prominent in the question text by having the words “red lines” in red font, we also informed respondents in the question text that the red lines “represent the boundaries of congressional districts (these lines are different in the two maps, but each map would keep the same number of people in each district).” Furthermore, we made the decision they were to be confronted with explicit wording in the question text. For example, in Study 2, the text read as follows: Based on your location and the other information provided in the maps, which arrangement do you prefer—the one on the left (Option 1, in which Washington County is entirely contained within the 2nd Congressional District and your location [the blue star] is in the 2nd Congressional District) or right (Option 2, in which Washington County is split between the 2nd Congressional District and 3rd Congressional District and your location [the blue star] is in the 3rd Congressional District)?
In short, we made it as apparent as possible to respondents where the congressional district lines were and what was in each.
In terms of the specific experimental conditions, in Study 1, respondents were randomly assigned to one of five partisan information conditions. In our baseline (control) condition, participants received no additional information about the partisan makeup of the two congressional districts in each map. The remaining conditions provided participants with the partisan makeup of the two districts in each map. For example, in one condition, the district on the left side of each map was labeled as 30 percent Republican and 70 percent Democrat, while the district on the right side of each map was labeled as 70 percent Republican and 30 percent Democrat. In this condition, choosing to keep the county intact places the participant in the majority Republican district and splitting the county places the participant in a majority Democratic district. Another condition mirrored this one, with the Republican and Democratic shares reversed. The final two conditions are similar to these two, but the partisan breakdowns of the districts are 51-49/49-51 and 49-51/51-49, in an effort to tap into the influence of electoral competition along with partisanship. 14
Study 2 took largely the same form as Study 1, with two exceptions. First, instead of five partisan information conditions, we only used three: (1) no additional information about the partisan makeup of the congressional districts, (2) partisan breakdown of the districts are 70-30 favoring Democrats and 30-70 favoring Republicans, and (3) the reverse of (2) 30-70 favoring Democrats and 70-30 favoring Republicans. That is, because heightened competition did not moderate the effect of the treatment in Study 1 (as we show below in Figure 1 and Table A2 in the online supplemental material), we do not include additional partisan information conditions with varying levels of competition in Study 2.
Estimated probability of splitting county by treatment condition (Study 1).
Second, we presented some respondents with information about what commissions are supposed to consider, specifically highlighting the importance of protecting “communities of interest.” Respondents randomly assigned to this treatment condition read the following information prior to viewing the maps: As you may know, the boundaries of congressional districts have to be drawn every ten years to make sure that the same number of people live in each district. To the extent possible, districts are also supposed to be drawn in a way that keeps counties intact within one district. On the next page you will be presented with a pair of maps that present two different options for congressional districts in a hypothetical state.
Respondents randomly assigned to the control condition did not receive the middle, italicized sentence. Our expectation was that the information about keeping geographic units intact to “the extent possible” (wording that was borrowed from California Common Cause’s Redistricting CA Pocket Guide) 15 would, if anything, result in respondents being less willing to split their counties even if doing so would advantage their partisan interests. Thus, Study 2 was a 3 (partisan information conditions) × 2 (commission prime conditions) experimental design.
Results
Study 1
The results of Study 1 appear in Figure 1, which displays the estimated probability that an individual selects the district map that splits the county for each treatment condition (along with 95% confidence intervals). For instance, the top point in Figure 1 marks the probability of splitting the county (0.433) when receiving no information about partisanship. This indicates that people are a bit more likely to prefer a district that does not split the county when not provided with partisan information.
Comparing this point with the other conditions reveals the effect of partisan information. To observe preferences when partisans are provided with knowledge of both partisanship and county integrity, we pool participants across different conditions according to their party identification. The second estimate, labeled “other party 70 percent,” for example, captures Democrats and Republicans assigned to the condition in which choosing to split the county would place them in a district where the opposing party has a 70 percent majority. In other words, keeping the county intact also results in a partisan advantage, so most respondents choose that option. The probability of splitting is only 0.235, which is 0.198 less than the no information condition (a statistically significant difference at a standard α = .05 threshold). 16 The third estimate, “other party 51 percent,” follows the same logic for conditions where the partisan makeup was more competitive. The remaining two estimates, “own party 70 percent” and “own party 51 percent,” capture Democrats and Republicans in conditions in which they are faced with a trade-off: either keep the county intact and live in a district where the opposing party has a (70% or 51%) majority or split the county and live in a district where copartisans hold a (70% or 51%) majority.
The results of Study 1 suggest that the presence of partisan information affects people’s willingness to split their county. The likelihood of selecting a map that splits the county is 0.235 for “other party 70 percent” and 0.247 for “other party 51 percent” when doing so would situate the respondent in a district where the other party holds a majority. Both conditions then result in statistically significant reductions in the likelihood of splitting the county as compared with the no (partisan) information condition (p < .05). However, the likelihood of selecting a map that splits the county is 0.765 for “own party 70 percent” and 0.722 for “own party 51 percent” when a split would allow the respondent to share a district with copartisans. That is, compared with the no (partisan) information condition, the probability of splitting increases by 0.332 and 0.288, respectively, each of which is a statistically significant difference from the no information condition (p < .05). Indeed, partisans 17 are more likely than not to split the county in this scenario.
Figure 1 provides little evidence that individuals prefer competitive districts over noncompetitive districts. The probabilities of splitting the county are statistically indistinguishable (p > .05) when comparing “other party 70 percent” with “other party 51 percent” or when comparing “own party 70 percent” with “own party 51 percent.” Respondents appear to be just as willing to split the county whether it means they would be in a district where their party has a slim or sizable majority.
Study 2
Our second study parallels the first but permits an examination of the effect of a trimmed-down version of a redistricting commission’s call to consider geographic integrity of local jurisdictions and communities. Once again, partisan information matters a great deal. The probability that a respondent will choose to split the county in the absence of both partisan information and the commission’s directive is 0.351 (topmost circle in Figure 2). As in Study 1, in the absence of additional information, individuals are more likely to keep the county whole than to split it. When given partisan information, respondents behave in much the same way as in Study 1: they are less likely to split the county when doing so would place them in a district where the other party has a majority (0.197; difference of 0.155 from no information, p < .05) and more likely to do so when a split county places them in a district where their party holds a majority (0.655; difference of 0.303 from no information, p < .05).
Estimated probability of splitting county by treatment condition (Study 2).
The commission prime does dampen the effect of partisan information in the expected direction, but its effect is not statistically significant (p = 0.326) or substantively very large (the estimated probability of a split drops from 0.655 to 0.585 for participants in the “own party 70%” condition). In other words, the draw to share a district with copartisans overcomes a preference for preserving a county within the district even for respondents who are instructed to attend to local jurisdictional boundaries.
Self-Reported Considerations
The finding that partisan information matters for whether citizens choose to split communities of interest (counties) is likely not a surprising one in general; however, given other recent studies in this area, our findings do look different. For example, Gaines and Kuklinski (2011) find that people prefer maps with simple shapes and the most compact districts regardless of the partisan and competitive outcomes of those maps.18 They find even strong partisans prefer these maps to ones that would have their party win more seats. Our study accounts for the location of the individual in political space though, as we specify where a citizen lives on the map, whereas Gaines and Kuklinski offer alternative district plans without specifying where someone would live. In tandem, the results of our study and that of Gaines and Kuklinski suggest that individuals may be more likely to express support for traditional redistricting principles that support fair plans in the abstract but revert to partisan cues when confronted with information on both the geographic and political outcomes for their own districts.
Indeed, a follow-up question we asked after Study 2 provides some additional evidence that this interpretation is plausible. We asked respondents why they selected the map they did by selecting (as many reasons as they wanted) from the following choices: (1) competiveness of the districts; (2) keeping counties, cities, or other communities of interest intact; (3) partisan makeup of the districts; (4) shape of the districts; and/or (5) “other” (where they typed in their reason). Table 1 provides the self-reported reasons for map selection by the various treatment conditions. A few results stand out.
Self-Reported Reasons for Decision by Treatment Conditions (Study 2).
Percentages are provided, with standard errors in parentheses.
First, with no partisan information, respondents were overwhelmingly likely to say that keeping communities of interest intact was a reason for their selection. It is by far the most selected option for respondents who received no partisan information: 71.7 percent among those who received the commission information and 62.9 percent among those who did not. Indeed, across all three (partisan) experimental conditions, receiving the commission information results in more people selecting keeping communities of interest intact as a reason for their selection, but sizable percentages (the lowest is 33.3%) select this option across all conditions. Finally, the partisan makeup of the district was selected as a reason at similar levels to keeping communities of interest intact in some cases, but only when the respondents received partisan information: 8.2 percent to 12.4 percent selected it when no partisan information was given, and 39.6 percent to 55.0 percent selected it across the four conditions in which partisan information was available to respondents. That is, when confronted with partisan information, respondents were much more likely to state it as a reason for their selection. These findings support the conclusion that in the absence of partisan information, respondents may be more likely to turn toward geographic (or other) factors when making decisions about how to partition district lines. However, even when faced with a commission prime, with access to partisan information, respondents are likely to make use of that information.
Discussion
Our experimental designs are of course not without limitations. Although we contend that giving partisan information more accurately reflects the knowledge elites or a citizen commission would have on hand when making decisions about where to place district boundaries, we still must abstract away from the real world in our experimental setting. One such abstraction is that people were not asked to split their own county or other community of interest but rather a hypothetical one. This might explain why, even without partisan information, respondents did not overwhelmingly choose to live in an intact county. This may have resulted from respondent confusion (although, as we detail above, our instructions for reading the maps were made as clear as possible) or from some broader disconnect about the importance of geography for individuals.
A second abstraction is that we use counties to test for citizen preferences for respecting political boundaries. We use counties as our political boundary as most states with a principle for respecting political boundaries specifically mention the county as a primary unit of geography. This is an appropriate and useful starting point for this type of study, but it brings up an important question related to individuals and their spatial awareness. First of all, counties are more relevant in some states and areas than others. Therefore, we may be underestimating the potential role of geography as some respondents may not care as much about or identify as strongly with their county but may have a stronger geographic affinity for their city or neighborhood. More important is the connection between political subdivision boundaries and the broader context of a community of interest. There is an ongoing literature that is grappling with the question of how people conceptualize the maps in their heads and the importance of this spatial awareness on public opinion (see Christenson and Makse 2015; Enos and Wise 2012; Wong 2010; Wong et al. 2012). The concern is that how people envision their communities may not overlap with the political boundaries on the map. This is an important consideration for future work in this area.
A third abstraction is that we examine only preferences for individual district boundaries rather than preferences over the full output of the redistricting process, which is an entire map of districts within a state. Individuals may reveal different preferences when considering complete maps detailing boundaries for all districts together than when they consider only their own district’s boundaries in isolation. Would citizens give partisanship less importance when considering an entire plan rather than their own individual district? The role of partisan interest in redistricting is a concern courts, reformers, and practitioners must consider when evaluating the merits of competing plans and methods for designing those plans. At the very least, our experiments indicate a strong potential for partisan considerations to play into decisions over boundaries and open further lines of research into the question of whether those results hold when citizens evaluate entire maps covering all districts.
Despite these limitations, the results of our two experimental studies suggest that like any reform coming before it, a citizen-led process is unlikely to be a panacea that will end the ills of redistricting. Consistent with the evidence we present here, the 2010 California Citizens Redistricting Commission unnecessarily split, in terms of population, 35 percent of counties across the Assembly and Senate plans when given instructions to protect communities of interest; however, this was down from the 43 percent split by the legislature in the 2000 redistricting. The citizen commission results are consistent with our results when providing language highlighting the importance of communities of interest. This information may lessen but will not necessarily negate the importance of partisan information.
As states move toward more citizen involvement in redistricting, it is important to gain a better understanding of citizen preferences toward not only redistricting in general but also specific aspects of the process. We provide one of the first studies to test citizen views not toward the overall process but toward a particular trade-off between specific considerations that all mapmakers must face—whether and how to keep communities of interest intact in the face of information about partisan implications of a district’s boundaries. Hopefully, this opens the door for additional studies that can give us a more complete picture of how citizens may behave as some citizens become more involved in a process that traditionally kept the public on the outside looking in.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Materials
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
