Opinion
SJC-13237
06-14-2022
MARTIN EL KOUSSA & others [1] v. ATTORNEY GENERAL & others. [2]
M. Patrick Moore, Jr. (Sarah K. Grossnickle, of Maine, & Thomas O. Bean also present) for the plaintiffs. Jesse M. Boodoo, Assistant Attorney General, for the Attorney General. Thaddeus Heuer (Andrew M. London & Seth Reiner also present) for the interveners. The following submitted briefs for amici curiae: Sarah David Heydemann & Sunu P. Chandy, of the District of Columbia, Rebecca G. Pontikes, & Lori A. Jodoin for National Women's Law Center & others. Lydia Edwards for Matahari Women Workers' Center. Michael J. Holecek, of California, & Joshua S. Lipshutz for Jon Paul Prunier & others. John Pagliaro & Daniel B. Winslow for New England Legal Foundation. Gary J. Lieberman for Chamber of Progress. Kevin P. Martin & William E. Evans for Chamber of Commerce of the United States of America. Sally Dworak-Fisher & Matthew J. Ginsburg, of the District of Columbia, & Audrey Richardson for American Federation of Labor and Congress of Industrial Organizations & others. Michael J. Adame, of California, Elsa C.W. Haag, of Oregon, & Jonathan B. Miller for Public Rights Project & others. Shannon Liss-Riordan & Anastasia Doherty for Massachusetts Employment Lawyers Association. Adam Cederbaum, Corporation Counsel, & Randall Maas, Assistant Corporation Counsel, for city of Boston. Harold P. Naughton for Massachusetts Budget and Policy Center. Jennifer G. Miller for William Good & others.
Heard: May 4, 2022.
Civil action commenced in the Supreme Judicial Court for the county of Suffolk on January 18, 2022. The case was reported by Lowy, J.
M. Patrick Moore, Jr. (Sarah K. Grossnickle, of Maine, & Thomas O. Bean also present) for the plaintiffs.
Jesse M. Boodoo, Assistant Attorney General, for the Attorney General.
Thaddeus Heuer (Andrew M. London & Seth Reiner also present) for the interveners.
The following submitted briefs for amici curiae:
Sarah David Heydemann & Sunu P. Chandy, of the District of Columbia, Rebecca G. Pontikes, & Lori A. Jodoin for National Women's Law Center & others.
Lydia Edwards for Matahari Women Workers' Center.
Michael J. Holecek, of California, & Joshua S. Lipshutz for Jon Paul Prunier & others.
John Pagliaro & Daniel B. Winslow for New England Legal Foundation.
Gary J. Lieberman for Chamber of Progress.
Kevin P. Martin & William E. Evans for Chamber of Commerce of the United States of America.
Sally Dworak-Fisher & Matthew J. Ginsburg, of the District of Columbia, & Audrey Richardson for American Federation of Labor and Congress of Industrial Organizations & others.
Michael J. Adame, of California, Elsa C.W. Haag, of Oregon, & Jonathan B. Miller for Public Rights Project & others.
Shannon Liss-Riordan & Anastasia Doherty for Massachusetts Employment Lawyers Association.
Adam Cederbaum, Corporation Counsel, & Randall Maas, Assistant Corporation Counsel, for city of Boston.
Harold P. Naughton for Massachusetts Budget and Policy Center.
Jennifer G. Miller for William Good & others.
Present: Budd, C.J., Gaziano, Lowy, Cypher, Kafker, Wendlandt, & Georges, JJ.
KAFKER, J.
The plaintiffs, twelve voters registered in Massachusetts, challenge the Attorney General's certifications of two initiative petitions, each proposing "A Law Defining and Regulating the Contract-Based Relationship Between Network Companies and App-Based Drivers." The plaintiffs contend that these petitions violate the requirement under art. 48 of the Amendments to the Massachusetts Constitution that initiative petitions must contain only related or mutually dependent subjects. The plaintiffs also object to the Attorney General's summaries of the proposed laws, arguing that they are not "fair" for purposes of art. 48 because the summaries do not adequately explain how the petitions, if approved by the voters, would change existing law.
We conclude that the petitions contain at least two substantively distinct policy decisions, one of which is buried in obscure language at the end of the petitions, and thus fail art. 48's related subjects requirement. As such, the Attorney General's decision to certify the petitions was in error, and accordingly the petitions may not be placed on the ballot.
We acknowledge the amicus briefs submitted by the National Women's Law Center, National Partnership for Women and Families, and twenty-five additional organizations; Matahari Women Workers' Center; Jon Paul Prunier, Shepard Collins, Rachel Brown, Ever Barrera, and Octavio Mejia-Suarez; New England Legal Foundation; Chamber of Progress; Chamber of Commerce of the United States of America; American Federation of Labor and Congress of Industrial Organizations (AFL-CIO), National Employment Law Project, and thirteen Massachusetts worker centers; ten civil rights organizations; Massachusetts Employment Lawyers Association; city of Boston; Massachusetts Budget and Policy Center; and William Good, George Garcia, and Anne Luepkes.
Background.
In August 2021, two initiative petitions signed by at least ten registered Massachusetts voters were filed with the Attorney General. The Attorney General designated them as Initiative Petitions 21-11 and 21-12. The two petitions each propose laws that are identical, except that Initiative Petition 21-11 includes an additional section relating to paid driver safety training.
The declared purpose of the petitions is to "define and regulate the contract-based relationship" between a specified category of business entities termed "network companies" and a category of workers termed "app-based drivers." Network companies, as defined in the petitions, comprise "Delivery Network Compan[ies]" (DNCs), which maintain online-enabled applications or platforms that connect couriers to customers to arrange and provide delivery services, and "Transportation network compan[ies]" (TNCs), which are rideshare companies that use a digital network to connect riders to drivers to arrange and provide transportation services. The category of app-based drivers under the proposed laws covers those couriers for DNCs and drivers for TNCs who provide delivery and transportation services under certain specified conditions of independence from the network companies.
The petitions define TNCs according to the statutory definition provided in G. L. c. 159A 1/2, § 1.
In particular, to be a covered "app-based driver," couriers or drivers must not have their work schedule unilaterally prescribed by the network company, must not be directed to accept specific service requests on pain of having their contract terminated by the network company, must not be generally restricted from performing services through the digital networks of other network companies, and must not be restricted from working in any other lawful occupation or business.
The laws proposed by the petitions each contain a provision that would classify any covered app-based driver as "an independent contractor and not an employee or agent" of a network company "for all purposes with respect to his or her relationship with the network company," "[n]otwithstanding any other law to the contrary" (first classification provision) --that is, regardless of the classification of app-based drivers under existing law. The proposed laws also specify a minimum level of compensation that network companies must pay to app-based drivers, calculated based on the total amount of a driver's "engaged time" or time spent fulfilling delivery or transportation requests. The proposed laws further specify various benefits that network companies must provide or make available to app-based drivers, including a health care stipend for drivers who meet a certain minimum of average engaged time per week, earned paid sick time, contributions to drivers' coverage under the paid family and medical leave (PFML) program established by G. L. c. 175M, and occupational accident insurance covering drivers' medical expenses, disability payments, and death benefits. The proposed laws would also provide app-based drivers with some form of protection against invidious discrimination by prohibiting network companies from refusing to contract with or terminating the contract of a driver based on certain protected characteristics. Initiative Petition 21-11, although not Initiative Petition 21-12, includes additional provisions requiring network companies to mandate driver safety training for the app-based drivers who contract with them, while also requiring network companies to compensate drivers for the time taken to complete the training.
Finally, both the proposed laws include a provision, placed in their respective final substantive sections, directing that "[n]otwithstanding any general or special law to the contrary, compliance with the provisions of [the proposed laws] shall not be interpreted or applied, either directly or indirectly, in a manner that treats network companies as employers of app-based drivers, or app-based drivers as employees of network companies" (second classification provision). The same sections include a further provision instructing that "any party seeking to establish that a person is not an app-based driver bears the burden of proof" (burden-of-proof provision).
In September 2021, the Attorney General certified both petitions as compliant with the requirements of art. 48 and issued summaries of the petitions as required under art. 48, The Initiative, II, § 3, as amended by art. 74 of the Amendments. These summaries make no mention of the second classification provision or the burden-of-proof provision. By December 2021, the petitioners had timely gathered and filed sufficient signatures to require the Secretary of the Commonwealth to transmit the petitions to the Legislature, which the Secretary then did.
In January 2022, the plaintiffs commenced this action in the county court, claiming that the Attorney General's certifications of the petitions were in error because the petitions did not, as required by art. 48, contain only related or mutually dependent subjects. The plaintiffs also challenged the summaries issued by the Attorney General, contending that they were not "fair" for art. 48 purposes.
In February 2022, ten of the original signers of the petitions filed a motion to intervene as defendants, which the single justice allowed. On the joint motion of the parties and a statement of agreed facts, the single justice then reserved and reported the case to the full court.
Christina M. Ellis-Hibbet, Katherine Mary Witman, Abigail Kennedy Horrigan, Richard M. Power, Meghan J. Borkowski, Chad B. Chokel, Daniel Svirsky, Michael Strickman, Marcus Alan Cole, and James William Isaac Hills.
Discussion.
Before an initiative petition can be presented to the Legislature and then to the voters, the Attorney General must certify that it meets the requirements of art. 48. See Oberlies v. Attorney Gen., 479 Mass. 823, 829 (2018); art. 48, The Initiative, II, § 3, as amended by art. 74. We review de novo the Attorney General's decisions as to whether to certify an initiative petition. Abdow v. Attorney Gen., 468 Mass. 478, 487 (2014), citing Mazzone v. Attorney Gen., 432 Mass. 515, 520 (2000) .
1. Related subjects requirement.
Under art. 48, a measure proposed by an initiative petition must "contain[] only subjects . . . which are related or which are mutually dependent." Art. 48, The Initiative, II, § 3, as amended by art. 74. This related subjects requirement arises from a recognition that "a voter, unlike a legislator, 'has no opportunity to modify, amend, or negotiate the sections of a law proposed by popular [initiative].'" Anderson v. Attorney Gen., 479 Mass. 780, 786 (2018), quoting Carney v. Attorney Gen., 447 Mass. 218, 230 (2006), S.C., 451 Mass. 803 (2008). Because "a voter cannot 'sever the unobjectionable from the objectionable' and must vote to approve or reject an initiative petition in its entirety," Anderson, supra, quoting Carney, supra, the related subjects requirement serves to ensure that voters are not placed "in the untenable position of casting a single vote on two or more dissimilar subjects," Weiner v. Attorney Gen., 484 Mass. 687, 691 (2020), quoting Abdow, 468 Mass. at 499. See Carney, supra at 220 ("the aggregation of . . . two very different sets of laws into one petition that the voter must accept or reject would operate to deprive voters of their right under art. 48 to enact a uniform statement of public policy through exercising a meaningful choice in the initiative process").
We have interpreted the related subjects requirement to allow for an initiative petition to include multiple subjects, "provided that the joined subjects have 'a common purpose to which each element is germane.'" Carney, 447 Mass. at 225, quoting Massachusetts Teachers Ass'n v. Secretary of the Commonwealth, 384 Mass. 209, 221 (1981). But recognizing that "[a]t some high level of abstraction, any two laws may be said to share a 'common purpose, '" Weiner, 484 Mass. at 691, quoting Carney, supra at 226, we have looked to two further factors to determine whether the different subjects in a petition are sufficiently tied to a common policy scheme, taking care not to define the required degree of relatedness "so broadly that it allows the inclusion in a single petition of two or more subjects that have only a marginal relationship to one another," Weiner, supra, quoting Abdow, 468 Mass. at 499. First, we ask whether "the similarities of an initiative's provisions dominate what each segment provides separately so that the petition is sufficiently coherent to be voted on 'yes' or 'no' by the voters." Second, we consider whether "the initiative petition 'express[es] an operational relatedness among its substantive parts that would permit a reasonable voter to affirm or reject the entire petition as a unified statement of public policy." Weiner, supra at 691-692, quoting Hensley v. Attorney Gen., 474 Mass. 651, 658 (2016) . Determining whether a petition's provisions come together to present voters with a sufficiently coherent or unified policy proposal is the "crux of the relatedness controversy." Anderson, 479 Mass. at 786, quoting Carney, supra.
Under this approach to the related subjects inquiry, it is no bar to a finding of relatedness that a measure is "complex[]" and contains "numerous different provisions," as long as the various provisions constitute an "integrated scheme." Weiner, 484 Mass. at 693, quoting Hensley, 474 Mass. at 659. For that reason, we determined that a petition to adopt a scheme for the licensing of food stores to sell wine and liquor, with associated regulatory provisions to prevent improper alcohol sales, complied with the related subjects requirement. See Weiner, supra at 688-690, 693. Similarly, we concluded that a petition proposing a comprehensive scheme for legalizing the possession and use of marijuana and for licensing, regulating, and taxing the retail sale of marijuana contained only related subjects. See Hensley, supra at 653-659. We have also held that a "measure does not fail the relatedness requirement just because it affects more than one statute, as long as the provisions of the petition are related by a common purpose." Albano v. Attorney Gen., 437 Mass. 156, 161 (2002).
We have, however, rejected as containing unrelated subjects petitions that address "two separate public policy issues." Gray v. Attorney Gen., 474 Mass. 638, 649 (2016). Presenting voters with a petition that combines "substantively distinct" policy issues, thereby yoking together disparate policy decisions into a single package that voters are only able to approve or disapprove as a whole, is to engage in "the specific misuse of the initiative process that the related subjects requirement was intended to avoid." Id.
We must also be sensitive to the possibility of voter confusion caused by obfuscation. Article 48 was designed to guard against various abuses of the initiative process, including the packaging of proposed laws "in a way that would confuse the voter." See Carney, 447 Mass. at 228, citing 2 Debates in the Massachusetts Constitutional Convention 1917-1918, 131, 152, 495-496 (1918) (Constitutional Debates). The delegates to the constitutional convention who drafted art. 48 specifically denounced "the practice of 'hitching' alluring provisions at the beginning of an initiative petition and burying more controversial proposals farther down." Carney, supra at 229, citing Constitutional Debates, supra at 567. Concealing controversial provisions in murky language is another way of burying them.
2. The petitions fail the related subjects requirement.
We conclude that the initiative petitions at issue here each encompass at least two distinct public policy decisions. Most of the petitions' provisions are devoted to defining a new contract-based relationship between network companies and app-based drivers, including an associated wage and benefit scheme that the companies will provide to the drivers. In accomplishing this purpose, the petitions define the drivers as independent contractors, regardless of whether they would have been so classified under existing law, and provide drivers with the specified wage and benefit scheme, regardless of what they would have been entitled to receive in wages and benefits under existing law.
However, in vaguely worded provisions placed in a separate section near the end of the laws they propose, the petitions move beyond defining the relationship between app-based drivers and network companies and the associated statutory wages and benefits. These provisions extend the classification of app-based drivers as independent contractors rather than employees or agents to potential lawsuits involving third parties, including apparently the victims of torts committed by app-based drivers, such as those assaulted by drivers or injured in traffic accidents. These provisions would thus have the apparent effect that in any actions seeking relief for torts committed by app-based drivers, the drivers are to be deemed independent contractors and not employees or agents, regardless of how they would have been classified under existing law. This would narrow the tort liability of network companies for drivers' misconduct or negligence, whether on a negligent hiring or retention theory or on a respondeat superior theory.
The petitions thus violate the related subjects requirement because they present voters with two substantively distinct policy decisions: one confined for the most part to the contract-based and voluntary relationship between app-based drivers and network companies; the other -- couched in confusingly vague and open-ended provisions -- apparently seeking to limit the network companies' liability to third parties injured by app-based drivers' tortious conduct.
a. Provisions concerning the contractual relationship between network companies and app-based drivers.
The petitions propose laws that would classify app-based drivers as independent contractors "for all purposes" with respect to their relationship with network companies, while creating a statutorily defined compensation and benefit structure for these drivers. In so doing, the petitions seek to redefine the distinction between independent contractors and employees under multiple statutes. As we have previously explained, Massachusetts laws "impose[] differing, and not uniform, definitions of employees and independent contractors." Camargo's Case, 479 Mass. 492, 500 (2018). This "lack of uniformity" in classification criteria across different statutory schemes "reflects differences in the particular laws," with the "laws governing workers' compensation, unemployment insurance, minimum wages, and tax withholding serv[ing] different, albeit related, purposes." Id. at 500-501. Each statute and its associated definition of employees and independent contractors involve a distinct and "complex allocation of costs and benefits for individuals, companies, and State government itself." Id. at 501.
By imposing a blanket classification of covered app-based drivers as independent contractors that applies "for all purposes" relative to their relationship with network companies, regardless of how they would be classified under the panoply of existing laws, the petitions would adjust app-based drivers' eligibility for various benefits under these different statutes and thereby change the multiple existing allocations of costs and benefits for drivers, network companies, and the Commonwealth. The scope of these changes, which affect so many different statutory schemes and so many different stakeholders, is obviously wide-ranging.
That being said, so long as all of the changes have a common purpose, a proposed law does not fail the related subjects requirement simply because it has an effect on multiple existing statutes. See Weiner, 484 Mass. at 693, quoting Albano, 437 Mass. at 161 (explaining that initiative petition does not fail related subjects requirement "just because it affects more than one statute"). Here, the petitions' proposed changes to the law are for the most part connected to the common purpose of defining the voluntary relationship between network companies and app-based drivers, by specifying the scheme of wages and benefits to which app-based drivers who contract with network companies will be entitled.
Thus, whether these wide-ranging revisions of our independent contractor and employment laws are sufficiently similar or operationally related to form an integrated or coherent policy scheme that satisfies the related subjects requirement is a complex, multifaceted question. It is also, however, a question we need not answer to decide the related subjects inquiry, given that the petitions extend their reach well beyond the contract-based relationship between network companies and app-based drivers, addressing the distinct policy issue of network companies' liability to third parties injured by the tortious conduct of app-based drivers.
b. Provisions implicating network companies' tort liability to third parties.
As we have explained, the first classification provision in each of the initiative petitions classifies every covered app-based driver as "an independent contractor and not an employee or agent," and does so "for all purposes with respect to his or her relationship with the network company," regardless of existing law. The second classification provision, which is found in a section near the end of each of the initiative petitions, establishes that "compliance with the provisions" of the petitions "shall not be interpreted or applied, either directly or indirectly, in a manner that treats network companies as employers of app-based drivers, or app-based drivers as employees of network companies." This vaguely worded provision, like the first classification provision, overrides any conflicting laws, as it applies "[n]otwithstanding any general or special law to the contrary," which is the standard statutory language used to "displace or supersede related provisions in all other statutes." See Harmon v. Commissioner of Correction, 487 Mass. 470, 480 (2021), quoting Camargo's Case, 479 Mass. at 498. The same section near the end of the petitions includes the burden-of-proof provision, stipulating that "any party" that seeks to "establish that a person is not an app-based driver bears the burden of proof."
In order to determine whether, by including the second classification provision and the burden-of-proof provision, the petitions fail the relatedness inquiry, we must first discern what they mean, which is no simple task. In interpreting the meaning of these provisions, the plaintiffs insist that they would bar the classification of app-based drivers as employees of network companies even in lawsuits that arise not between network companies and app-based drivers but between network companies and third parties harmed by drivers, where the third parties seek to hold network companies liable for the tortious actions of the drivers who provide services through their platforms. The interveners suggest instead that the provisions at issue simply require courts to interpret all the provisions within the petitions consistently with the petitions' definition of app-based drivers as independent contractors and not as employees of network companies. The Attorney General argues in turn that "there is no reason to conclude that the proposed laws, if enacted, would have any effect on private tort litigation in the Commonwealth," and even if they do have such "secondary effects" on tort law, the Attorney General maintains that the challenged provisions are still related to the petitions' common purpose of defining and regulating the voluntary relationship between drivers and network companies.
The interveners do not deny that the petitions could have legal consequences for third parties; however, they claim that any such consequences will simply be follow-on effects of the petitions' central purpose of defining the relationship between app-based drivers and network companies, which includes classifying app-based drivers as independent contractors in relation to network companies.
In line with the interpretation suggested by the plaintiffs, we interpret the classification provisions and the burden-of-proof provision to require app-based drivers to be classified as independent contractors rather than as employees in third-party tort suits. The language in the burden-of-proof provision stipulating that "any party" seeking to establish that an individual is not a covered app-based driver "bears the burden of proof" seems to contemplate a lawsuit brought by third parties, not just litigation between drivers and network companies. Among such third-party suits would be suits by members of the public who have been harmed by the tortious conduct of app-based drivers, such as individuals injured in automobile accidents caused by drivers' negligence or assaults by such drivers. As we explain infra, in determining network companies' liability in such cases, either based on negligent hiring or retention or based on respondeat superior, the issue whether app-based drivers are independent contractors, or instead are employees or agents of the network companies, is a crucial question.
The first and second classification provisions in the petitions set out rules of interpretation to guide courts in analyzing this question. Under the first classification provision, regardless of any contrary existing law, an app-based driver is to be classified as an independent contractor, rather than an employee or agent, "for all purposes with respect to his or her relationship with the network company." The scope of this provision is somewhat uncertain: although "for all purposes" suggests an unlimited scope, that phrase is modified by "with respect to his or her relationship with the network company," which suggests that the provision may apply only to regulate the voluntary relationship between network companies and app-based drivers.
When the first classification provision is read in conjunction with the second, however, most if not all doubt is removed that the purpose and effect of the initiative petitions is to extend to third-party tort suits the classification of app-based drivers as independent contractors. We emphasize in particular that the second classification provision instructs that compliance with the proposed laws "shall not be interpreted or applied, either directly or indirectly," so as to treat network companies and app-based drivers as standing in an employer-employee relationship (emphasis added). There would seem to be no reason to include the word "indirectly" if the provision was intended to be limited in its application to defining the contract-based relationship between app-based drivers and network companies. Rather, its scope extends beyond that relationship, to encompass lawsuits brought by other parties and the indirect application in those suits of the classification of app-based drivers as independent contractors rather than as employees or agents of network companies. This interpretation of the classification provisions and the burden-of-proof provision supplies a reasonable meaning and purpose to all of the different words employed, some of which might otherwise be inoperative or superfluous.
See Commonwealth v. Woods Hole, Martha's Vineyard & Nantucket S.S. Auth., 352 Mass. 617, 618 (1967) ("It is a well established principle of statutory interpretation that none of the words of a statute is to be regarded as superfluous, but each is to be given its ordinary meaning without overemphasizing its effect upon the other terms appearing in the statute, so that the enactment considered as a whole shall constitute a consistent and harmonious statutory provision ..." [quotation, citation, and alteration omitted]).
Finally, for the purposes of the related subjects inquiry, any residual doubts about the meaning of an obscurely drafted petition must be resolved against the proponents of such a petition. Otherwise, we would be encouraging or at least condoning efforts to mislead and confuse voters by concealing controversial provisions in obscure language. That would cut impermissibly against the design of art. 48, which was constructed to include "safeguards against potential voter confusion in the initiative process." Carney, 447 Mass. at 230. See Anderson, 479 Mass. at 801, quoting Carney, supra at 227 n.20 (art. 48 designed to safeguard voters from being "misled" by efforts to "wheedle or deceive" them).
We therefore conclude that, by including the vaguely worded classification provisions and burden-of-proof provision, the petitions go well beyond the contract-based relationship between network companies and app-based drivers, and the compensation and benefits associated therewith. Instead, they mandate that app-based drivers may not be deemed agents or employees of network companies either directly or indirectly, that is, in lawsuits brought by third parties. In so doing, they apparently redefine the scope of tort recovery for third parties, including those who may have been injured in traffic accidents caused by the negligence of app-based drivers, or even sexually assaulted by them.
Precluding app-based drivers from being classified as employees or agents of network companies in third-party tort suits would narrow the scope of tort recovery for two reasons. First, under the doctrine of respondeat superior, network companies would be vicariously liable for the torts of app-based drivers committed within the scope of their agency or employment only if the drivers are classified as agents or employees of the network companies. See Merrimack College v. KPMG LLP, 480 Mass. 614, 620 (2018), citing Lev v. Beverly Enters.-Mass., Inc., 457 Mass. 234, 238 (2010) ("the tortious conduct committed by an agent in the scope of his or her agency will be imputed to the principal under a theory of respondeat superior"); Lev, supra, quoting Dias v. Brigham Med. Assocs., Inc., 438 Mass. 317, 319-320 (2002) ("Under the doctrine of respondeat superior, 'an employer . . . should be held vicariously liable for the torts of its employee . . . committed within the scope of employment'"); Restatement (Third) of Agency § 2.04 (2006) ("An employer is subject to liability for torts committed by employees while acting within the scope of their employment").
Second, even when an app-based driver's tortious conduct falls outside the scope of agency or employment, as with sexual assault for example, the network company might still be liable for negligently hiring or retaining the driver, provided that the driver is an employee. The doctrine of negligent hiring or retention provides that "an employer whose employees are brought in contact with members of the public in the course of the employer's business has a duty to exercise reasonable care in the selection and retention of his employees." Foster v. Loft, Inc., 26 Mass.App.Ct. 289, 290 (1988).
When an employee or agent commits sexual assault, he or she does not act within the scope of employment or agency because sexual assault is not "motivated by a purpose to serve the employer [or principal]," and it "do[es] not serve the interests of the employer [or principal]." Doe v. Purity Supreme, Inc., 422 Mass. 563, 568 (1996).
Under existing law, "the task of determining what constitutes an employer-employee relationship is fact dependent." Dias, 438 Mass. at 322. To determine whether an employer-employee relationship exists, Massachusetts courts consider a number of factors, including "the method of payment . . . and whether the parties themselves believe they have created an employer-employee relationship." Id., citing Restatement (Second) of Agency § 220(2) (1958) . Moreover, as explained by the plaintiffs without contradiction, the classification of app-based drivers as employees or agents, or as independent contractors, has been a contested issue in Massachusetts tort suits against the network companies Uber and Lyft. The vaguely worded provisions in the petitions would displace this fact-sensitive inquiry, barring courts from classifying covered app-based drivers as employees in third-party tort suits.
Based on this understanding of the meaning and legal effect of the obscurely drafted provisions, we return to the related subjects inquiry. We conclude that limiting the scope of third parties' tort recovery for injuries caused by app-based drivers is a substantively distinct policy issue from defining the wage and benefit structure of those drivers. Voters may support one and not the other. They may, for example, strongly approve of better wages and benefits for drivers struggling to make ends meet in the gig economy, but at the same time strongly oppose limiting their own rights to recover money damages from network companies if the tortious actions of drivers who provide services through those companies' platforms cause them injury. See Anderson, 479 Mass. at 799-800 (explaining how related subjects requirement not met where petition requires voters to cast single vote on different subjects on which they might make divergent choices).
In presenting voters with two distinct policy decisions packaged in a single petition, the petitions at issue here are similar to others we have previously determined to contain provisions addressing unrelated subjects. See, e.g., Oberlies, 479 Mass. at 835-837 (requirement that State-funded hospitals make comprehensive financial disclosures was separate policy issue from implementing mandatory nurse-to-patient staffing ratios); Anderson, 479 Mass. at 799-800 (tax increase to support two important but diverse spending priorities combined unrelated policy decisions); Gray, 474 Mass. at 648-649 (elimination of core curriculum content and publication of standardized testing were two separate public policy issues); Carney, 447 Mass. at 231-232 (enhanced penalties for animal cruelty and abolition of dog racing did not express uniform statement of public policy).
The defendants and the interveners argue nonetheless that the petitions' effect on third parties' scope of tort recovery is simply a downstream consequence of the petitions' purpose of defining the contract-based relationship between app-based drivers and network companies and the associated classification of drivers as independent contractors. To be sure, we have previously held that even if an initiative petition would have "consequences under an assortment of other statutes," that alone does not make the provision fail the related subjects inquiry, provided that these consequences are "logically related to the petition's aim." Abdow, 468 Mass. at 503-504. For example, in Albano, 437 Mass. at 161, we held that a petition to define marriage in Massachusetts as only between one man and one woman did not violate the relatedness requirement because "each statue affected creates a benefit or responsibility that arises from married status."
Here, by contrast, we are not just dealing with downstream consequences. The initiative petitions provide instructions and directions on how courts should interpret and apply the provisions of the laws they propose, notwithstanding any other laws to the contrary. By instructing or directing that covered app-based drivers are to be deemed independent contractors and not agents or employees, regardless of how they would otherwise be classified under existing agency or tort law, the petitions move well beyond the consequences of establishing a scheme of wages and benefits for app-based drivers as independent contractors. An express instruction or directive in an initiative petition is different from a consequential effect.
We also note that the classification and the burden-of-proof provisions regarding liability to third parties are not mutually dependent on the provisions defining the wage and benefit structure for app-based drivers. Regardless of whether the mutual dependence requirement is separate from or subsumed within the relatedness requirement, an issue that has not been definitively resolved by this court, it is not satisfied here. Compare Anderson, 479 Mass. at 790-791 ("As these cases demonstrate, the language 'or which are mutually dependent' . . . [has not been found to] impose a separate requirement that may be satisfied even if the subjects of a petition are not related"), with Oberlies, 479 Mass. at 835-838 (analyzing two petitions' provisions under both relatedness and mutual dependence inquiries and affirming Attorney General's decision declining to certify one petition because petition addressed subjects that were "neither mutually dependent nor related"). As we explained earlier, limiting the scope of third-party tort recovery for injuries caused by app-based drivers is a substantively distinct policy issue from defining the wage and benefit structure of those drivers. The petitions create a novel wage and benefit regime for app-based drivers while classifying them for purposes of their relationship with network companies as independent contractors, notwithstanding any existing law to the contrary. Although obscured by murky language, the petitions also redefine the scope of network companies' tort liability in relation to third parties injured by app-based drivers, notwithstanding any existing law to the contrary. Specifically, the petitions require the app-based drivers to be deemed independent contractors for third-party purposes, even if the drivers would not have been so considered under existing agency and respondeat superior principles. The new wage and benefit regime and the directive to define app-based drivers as independent contractors for purposes of network companies' tort liability are separate decisions that can "exist independently" of each other. Oberlies, supra at 837, quoting Gray, 474 Mass. at 648. See Oberlies, supra at 837-838 (financial data requirements not mutually dependent on enforcement of mandatory nurse staffing levels); Gray, supra (in evaluating whether common core curriculum requirements and release of diagnostic tests were mutually dependent, court concluded that "whether the diagnostic assessment tests are based on the common core standards or some previous set of academic standards . . . will not affect in any way the commissioner's obligation ... to release before the start of every school year all of the previous year's test items in order to inform educators about the testing process"). Here, although the new wage and benefit regime set out in the petitions would have had consequential effects on courts' analysis of network companies' tort liability to third parties under existing law, the petitions went well beyond such downstream consequences. Rather, the petitions appear to affirmatively instruct and direct courts to classify app-based drivers as independent contractors for purposes of determining network companies' tort liability under respondeat superior and negligent hiring and retention theories.
Finally, we emphasize that the petitions' redefining of the network companies' third-party liability in murky language, and their burying of these provisions in the final substantive section of the proposed laws, raise particular concerns from the perspective of art. 48. As explained supra, we are conscious that a "recurring topic of concern" among the framers of art. 48 was "the possibility that well-financed 'special interests' would exploit the initiative process to their own ends by packaging proposed laws in a way that would confuse the voter," in particular by prominently placing "alluring provisions" in the front of the petition while "burying more controversial proposals farther down." Carney, 447 Mass. at 228-229, citing Constitutional Debates, supra at 131, 152, 495-496. Indeed, the delegates to the constitutional convention expressed a more general concern that the initiative process might be abused by presenting voters with confusingly and misleadingly formulated petitions. See Constitutional Debates, supra at 12 ("measures initiated . . . may be as abstruse . . . [and] as full of tricks . . . as the proposers of the measures may choose"); Id. at 532 ("any one can frame any measure he chooses ... as trickily as he wishes"); id. at 537 (emphasizing need to ensure proposed measures are not "misleading in [their] phraseology"); Id. at 567 (expressing concern that initiative proponents may "wheedle or deceive" voters into enacting measures that Legislature would never permit).
Petitions that bury separate policy decisions in obscure language heighten concerns that voters will be confused, misled, and deprived of a meaningful choice -- the very concerns that underlie art. 48's related subjects requirement. Voters are not only unable to separate one policy decision from another; they may not even be aware they are making the second, unrelated policy decision. When even lawyers and judges cannot be sure of the meaning of the contested provisions, it would be unfaithful to art. 48's design to allow the petition to be presented to the voters, with all the attendant risks that voters will be confused and misled.
In sum, as these petitions reasonably appear to seek to limit network companies' liability for torts committed by app-based drivers, by barring courts hearing tort suits from treating network companies as employers of app-based drivers and drivers as employees or agents of network companies, and this is a separate, significant policy decision that has been obscured by murky language, we conclude that the petitions violate the relatedness requirement of art. 48.
As we conclude that the initiative petitions fail the related subjects requirement, we need not resolve the issue whether the Attorney General has provided fair summaries of the petitions. We do note, however, that the failure to even discuss the provisions narrowing third parties' tort recovery here would have rendered the summaries unfair.
Conclusion.
The matter is remanded to the county court where a judgment shall enter declaring that the Attorney General's certifications of Initiative Petitions 21-11 and 21-12 are not in compliance with the related subjects requirement of art. 48 and thus that the petitions are not suitable to be placed on the ballot in the 2022 Statewide election.
So ordered.