Given the vital role of parents in shaping young lives, one might think that national constitutions would seek to regulate parental behavior, or at least identify some principles. Quite a few do, as it turns out. 95 of the world’s 192 constitutions provide some regulation about parenthood, as I’ve learned from analyzing the text of world constitutions currently in force.
This research is part of an effort to expand the topic set of Constitute, a repository of national constitutions indexed by some 330 topics. With a grant from Google, Constitute was built by scholars to provide constitutional drafters a sense of options and sample text across these topics. Since its launch in 2013, Constitute has hosted some 7,000 visitors daily and has become the primary source of ideas and text for those writing constitutions. Most of those who write constitutions have never done so before and probably will never do so again; the point of Constitute is to assist their learning from prior drafters.
Perhaps not surprisingly, Constitute’s current set of topics requires periodic updating. Relatively modern issues such as internet privacy, climate change, and gender norms are starting to find their way into higher law and are not in Constitute. Our update of the Constitute topic set enables drafters to understand the full range of constitutional possibilities. Candidates for new topics often come from a reading of new constitutions (such as the Chilean draft currently under consideration) as well as from scholarly research. In this case, we have drawn from a set of candidate topics identified by Jill Lepore’s team at the Amend Project, an effort to collect data on each of the 12,000-or-so proposals to amend the U.S. Constitution since 1789. Amend Project researchers used the Constitute topic set to index these proposals and identified some 25 topics not represented in the Constitute set. These 25, therefore, represent a set of new topics with which to index world constitutions on Constitute, assuming that the world’s constitutions include such provisions.
“Rights or Duties of Parents” is one topic Lepore’s team identified. Most of the U.S. proposals seem to protect the right of parents to parent as they see fit, sometimes defining it as a “fundamental right,” such as this proposal from 2014:
The liberty of parents to direct the upbringing, education, and care of their children is a fundamental right … The parental right to direct education includes the right to choose public, private, religious, or home schools (S. J. Res. 37).
The U.S. proposals, all 16 of which have stalled in Congress, are largely the same as the example above. Some, however, are more restrictive, such as a 2012 proposal: “The article shall not be construed to apply to a parental action or decision that would end life” (S. J. Res. 37). Others are specifically inclusive, such as those that disallow the denial of parental rights “on account of disability” (H. J. Res. 91).
My objective was to identify any such provisions in world constitutions and understand how they vary.
Matching Constitution Sections to the Topic
To discover and index constitutional text related to parents, one could read each of the 193 constitutions currently in force. But given the scale of the task, it makes sense to leverage digital tools. Our approach is to match descriptions of the target topic to each segment of constitutional text in the corpus (roughly 150,000 segments) using Google’s Universal Sentence Encoder, which has become state-of-the-art in assessing “semantic similarity.” The technology assigns a similarity score between 0 and 100 for each of the millions of pairs. We generally focus on results with a score over 65, which are usually pretty close semantic matches, but one sometimes needs to adjust this threshold to include more or less content.
The first and most important step when searching for constitutional text under a topic is defining the search parameters, particularly the topic description. The topic description should be distinct enough from existing Constitute topics to minimize the number of overlapping categories. After all, the point is to add a new topic to the topic set. Moreover, the topic description must be specific enough that constitution sections identified by the tool stay centered on the substance of the topic. While the topic’s essence is “parental rights or duties,” that formulation is far too broad. As the tool has less to go on in that case (only three meaningful words), it detects semantic similarities in a wide range of constitution sections, including, for example, an excerpt from Cameroon’s 2008 constitution about “The fundamental rights, guarantees, and obligations of the citizen” (art. 26 sec. 2a). “Parental rights or duties” was, therefore, a good place to start, but I needed to specify what kind of rights and duties I was looking for to give the tool a better idea of the topic.
After some experimentation, the topic formulation I settled on was “the parental duty to provide for and shelter children, or the parental rights to make decisions about children’s lives.” Once I found a topic formulation that gave me the most relevant results, I adjusted the search threshold to fine-tune how broad I wanted the topic to be. The idea is to keep those constitutional sections above a certain level of similarity — usually around 65. Lowering the threshold ensures that relevant constitutional text is not omitted, while raising the threshold eliminates less relevant results. After finalizing the results, the next step was to manually evaluate each of the 180 excerpts to determine whether they belonged to the topic. While tedious, this process was necessary to eliminate any excerpts that were semantically related but did not fit with the topic’s intended meaning. In other words, human judgment was required to correct the “mistakes” made by the tool.
How Do the Rights and Duties of Parents Differ Across Constitutions?
Using this process, I identified 147 constitutional excerpts from 95 countries that dealt with “parental rights or duties.” The excerpts, as I indexed them, are listed on the Constitute site here. While each country has a unique way of dealing with the subject, they fall into several basic constitutional approaches (see table below). Among these, the most common category (50 out of the 95 countries) is the parental right and duty to provide education for children. For example, the Portuguese constitution (2005) states that “Parents shall possess the right and the duty to educate and maintain their children” (art. 36 sec. 5). Similarly, the Chilean constitution (2021) affirms that “Parents have the right to choose the educational institution for their children” (art. 11).
In addition to education, fourteen countries explicitly give parents the right to raise children in accordance with their religious and moral standards. Cuba’s constitution (2019) has a unique approach to moral education:
Mothers and fathers have essential responsibilities and roles in the holistic education and upbringing [of their children] as citizens with moral, ethical, and civic values in correspondence with life within our socialist society (art. 84).
Interestingly, Cuba is the only country whose constitution requires parents to educate their children morally and religiously following the state’s political ideology.
Some provisions are focused not on the care of children, but on that of the parents. In fact, 28 countries obligate children to care for their parents in old age, such as Ecuador (2021):
To help, feed, educate and raise one’s children. This duty is a joint responsibility of mothers and fathers, in equal proportion, and shall also be applicable to children when their own mothers and fathers need them (art. 83 sec. 16).
In addition to giving them care, several constitutions require children to respect their parents as well. Lithuania’s constitution (2019) states that “the duty of children is to respect their parents, to take care of them in their old age, and to preserve their heritage” (art. 38).
Concerning rights within a household, seventeen countries maintain that all children have the same rights, no matter their birth circumstances. Malawi’s constitution (2017) is one such example:
All children shall be entitled to reasonable maintenance from their parents, whether such parents are married, unmarried or divorced, and from their guardians; and, in addition, all children, and particularly orphans, children with disabilities and other children in situations of disadvantage shall be entitled to live in safety and security (art. 23 sec. 4).
Furthermore, seven countries give individuals in a married couple equal rights to determine the lives of their children, one of which is Romania (2003):
The family is based on a freely consented marriage by the spouses, their full equality, and the right and duty of the parents to raise, educate, and instruct their children (art. 48 sec. 1).
Seventeen countries make provisions for the adoption of children due to parental neglect or abuse, including Ireland’s constitution (2019):
Provision shall be made by law for the adoption of any child where the parents have failed for such a period of time as may be prescribed by law in their duty towards the child and where the best interests of the child so require (art. 42a sec. 2.2).
Finally, Colombia’s constitution (2015) provides that “the couple has the right to decide freely and responsibly the number of their children” (art. 42). A few countries have historically limited the number of children a couple can have, such as China, Iran, and Vietnam, but Colombia is the only modern constitution to address the issue of child number.
Evidently, many constitutions do account for parents, both in protecting their right to raise their children in some reasonable fashion and also in protecting children from some known harms. The kinds of provisions that I describe here represent my sense of how they vary. But there is no reason to follow that categorization. Now that we have indexed constitutions on Constitute with the parental tag (and listed here), readers can read and organize them as they like. The hope is that constitutional drafters can make a more informed decision about whether to address this important issue and how.
Lucas Elkins is a Research Associate at the Comparative Constitutions Project. He writes about various aspects of world constitutions, among other topics across the social sciences.
* This material is based upon work supported by the National Science Foundation under Grant Number 2315189. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.