Crowdsourcing experiences in the lawmaking cycle have grown significantly in the past fifteen years. Crowdsourcing legislation practices aim at engaging the public in lawmaking to improve the legislative process and its outcomes, as well as enhance legitimacy and transparency in the conventional legislative process. Different efforts have been made to map cases, identify core characteristics, and assess results. Despite these contributions, a more comprehensive understanding of the universe of crowdsourcing legislation practices and core components is still lacking. This paper provides a deeper look into this phenomenon through a review of crowdlaw classification criteria, influential attributes mentioned by the CrowdLaw initiative and others identified in e-participation and crowdsourcing literature. Finally, it proposes a taxonomy with key features of crowdsourcing legislation initiatives and their respective variables through which these initiatives can be analyzed and compared.
Technology offers numerous opportunities to democratize the lawmaking cycle. It promises to improve
regulatory decisions and compliance, as well as decrease administrative burdens such as face-to-face
meetings. Furthermore, it provides more possibilities and new forms of public participation, through which
more effective, efficient and legitimate regulations, legislation, and constitutions can be created [
The term "crowdlaw" combines the ideas of "crowdsourcing" and "law". It encompasses more
institutionalized digitally enhanced initiatives which seek more than just opinions and aim at improving the
quality of the lawmaking process and its outcomes [
While crowdsourcing in public decision-making processes promises to convey economic, democratic,
and epistemic value [
Throughout the past years, crowdsourcing practices in lawmaking have become more popular. While some efforts were made to map, assort, and assess crowdsourcing experiences, the field would benefit from a closer look into the variety of their formats. This paper intends to make a contribution to this gap. In section 2, it identifies key elements and variables proposed by the CrowdLaw initiative in different publications and repositories. In section 3, these elements are supplemented by other relevant features identified in e-participation and crowdsourcing literature. Finally, in section 4, a taxonomy with the final criteria for the broader concept of crowdsourcing legislation is presented. The table is intended to be a useful tool to analyze and compare cases, determine typologies, and, in future research, identify influencing factors in crowdsourcing legislation processes and their results for future improvement.
The CrowdLaw project made two efforts to map and classify crowdlaw cases: the CrowdLaw Repository3 and the CrowdLaw Catalog4. The Repository, the first of the two, compiled twenty cases on Google sheets, while the Catalog contains over one hundred on its own webpage, displays links to more information on the cases, shares some details on the CrowdLaw initiative and the classification criteria used. The CrowdLaw Playbook5 is a more recent publication with selected cases that are further described and analyzed. Some criteria used to classify cases is common to all three, while other elements were only used in one or two of these efforts.
The Repository, Catalog, and Playbook all provide basic information on the cases, such as the name of the initiative, its owner, and its location. They also classify cases according to three attributes: the implementation level (local, regional or national), the stage of the lawmaking cycle (agenda-setting/problem definition, proposal-making/solution identification, drafting, implementation or evaluation) and the participant task (ideas & proposals, expertise, opinions, actions and evidence & facts). The variables used, however, slightly differ, e.g., the repository uses “agenda-setting” while the other two distinguish between the “problem definition” phase and the “solution identification” one.
Both the Repository and Playbook classify the cases according to the type(s) of platform used, web, mobile and offline. Finally, the Repository assesses whether they were subject to a formal legal process (legal framework), while the Playbook includes the method implemented by each initiative. There are six methods attributed to cases: collaborative drafting, open innovation, AI-based insight generation, pairwise wikisurveys, online brainstorming, prize-backed challenges and social auditing.
A publication released by the CrowdLaw initiative highlights the importance of assessing some of these
features in order to understand the impacts they may have on the results of crowdlaw practices [
The publication also highlights the relevance of studying the impact of different participatory tasks in the crowdsourcing exercise to understand the ones that function better. Finally, regarding the criteria applied to classify crowdlaw cases, Capone et al. believe it is important to know which kind of platform, from webbased to SMS-based, are more often used and what results they have prompted so far.
There are no explanations or more details regarding the value in assessing the implementation level, legal
framework or methods in these publications and repositories, although an important article by Alsina &
Martí [
The CrowdLaw initiative also defends the inclusion of other so-called “salient features of crowdlaw
initiatives” in the analysis of cases [
Firstly, they hypothesize that participation is mainly male and upper middle class if active measures to diversify the audience are not taken. Therefore, it is key to develop a form of assessing and connecting the demographics and features to the input of crowdlaw participants.
Uncovering the most effective incentives to engage the public in lawmaking is also crucial because
guidance (“clearly defined rules of procedure”), relevance (“an understanding of the relevance of one’s
participation to the ultimate outcome”), and impact (“the ability to make a difference”) can be key
motivators for repeated engagement [
The type of law in question, whether it is a regulation, legislation or constitution, can also have distinct impacts. Capone et al. propose that the results of the same participatory tasks for different law types be compared.
The topic being discussed matters because more controversial themes can generate different levels and qualities of engagement. The authors also stress the relevance of identifying if and how the parliament provides feedback to participants and its impact on participation and repeated engagement. They hypothesize that the more responsive and interactive a government is, the more it will increase participation and frequency.
Timing, referring to the time available for participation and the number of phases, can also be linked to the results of participation. Finally, Capone et al. sustain that providing training before the participation exercise can increase the quality of participation and usefulness of inputs received from the public.
Apart from the elements mentioned by the CrowdLaw initiative, I hypothesize that other design and process features can be relevant to a deeper understanding of the wide array of crowdsourcing legislation initiatives and their results. These are drawn from the literature on e-participation and crowdsourcing, and studies on crowdsourced policy and lawmaking cases.
First, the presence of crowdsourcing partners can impact participation and the legitimacy of the process.
Partners can be government officials in bottom-up initiatives, or well-known and respected academic and
civil society organizations, for example. The Center for Technology and Society (Fundação Getúlio Vargas),
a well-known think tank in the field of Law and Technology, collaborated in a federal government-led
consultation that drafted the bill for today's Brazilian Internet Bill of Rights and was key to promoting and
involving more participants in the process [
The frequency of the initiative is potentially connected to its degree of institutionalization. Single cases
tend to be experiences that often do not produce outcomes, while ongoing cases are more likely to do so.
The vTaiwan initiative, for instance, an ongoing multistakeholder consultation process which promotes
discussions between experts, government officials, stakeholders, and citizens to elaborate recommendations
for national legislation, had an 80% responsiveness rate by 2018 [
Another important element is the presence of moderation during discussions. While active moderation
can stimulate the debate and ensure respect and civility in online consultations [
Like other participation practices, some promote discussions, while others do not include instances for
deliberation. Deliberative formats can generate a series of benefits such as promoting epistemic gains for
participants and crowdsourcers [
Crowdsourcing legislation practices can include one or more channels of participation. They are also
sometimes connected, for example, to social networks, and frequently function as another source of
participant contribution [
Like other crowdsourcing projects, they can differ according to their IT structure [
Finally, crowdsourcing practices generally involve the collection of large quantities of data. The processing method involved, whether automatic, human or both, can have several implications, including whether the input is adopted by governments or not. In the crowdsourcing of the update to the Palo Alto city plan, Chen & Aitamurto find that civic data overload led the government to rely heavily on a human filter, the citizen advisory committee, to synthesize and absorb public input. They conclude that this hindered equality and inclusiveness in the process [13].
In order to propose a taxonomy of crowdsourcing legislation initiatives (Table 1), I first included the
features applied in existing crowdlaw case classifications and those hypothesized as valuable by Capone et
al. [
While new challenges of technology for policy- and lawmaking are constantly emerging, crowdsourcing legislation projects are growing and diversifying their formats. The purpose of this paper was to advance the criteria to better identify and analyze them. Eighteen key attributes were identified.
The table includes features that can be contrasted amongst different initiatives within different contexts. The specific institutional, legislative context the initiative is inserted within of course affects its dynamics, as well as eventual external happenings.
The criteria proposed do not configure an exhaustive list of features nor variables. Several variables could also present other, hybrid, alternatives. This paper will be further developed to apply the taxonomy to an in-depth analysis of crowdsourcing legislation cases. Hopefully, this will lead to a larger variety and more precise features and variables.
[13] Chen, K., & Aitamurto, T. (2019). Barriers for crowd’s impact in crowdsourced policymaking: Civic data overload and filter hierarchy. International Public Management Journal, 22(1), 99-126.