UN Report: Welfare, Control, and Paternalism Critique

Human Rights Watch welcomes the opportunity to provide input to the UN Special Rapporteur on Extreme Poverty and Human Rights' thematic report on "Welfare and Control: The Paternalism of Support," shedding light on some of the various forms of monitoring and control that people in poverty are subjected to, especially those receiving social security.[1] This submission responds to questions about three issues raised in the call for inputs: child protection services, conditionalities associated with cash transfers, and duty to accept "suitable" work. It also highlights how many social security systems that primarily rely on poverty targeting, which limits benefits to the poorest members of society, use surveillance and extensive data collection.

Drawing on Human Rights Watch research conducted in Germany, Jordan, Mauritius, Nepal, the United Kingdom (UK), the United States (US), and Australia, this submission illustrates how government policies can reinforce stigma against social security recipients. It also presents examples of positive measures that mitigate these harms while upholding the right to social security.[2]

1. Child protection services in the United States and Australia

A 2022 report by Human Rights Watch and the American Civil Liberties Union (ACLU) titled "If I Wasn't Poor, I Wouldn't Be Unfit," documents how the United States' child welfare system disproportionally affects families living in poverty.[3] The government subjects families to surveillance, investigations, and even the removal of their children, often because of poverty-related circumstances. At the same time, the government does not provide social security adequate to the task of alleviating those circumstances.

In the US, one in three children will be part of a child welfare investigation before the age of 18.[4] Black children are almost twice as likely to experience investigations as white children and are more likely to be separated from their families.[5] These investigations can be highly intrusive and destabilizing, involving home inspections, school visits, body checks, and intrusive questioning of parents and children.

As a result, every three minutes, a child is removed from their home and placed in the foster system.[6] More than 200,000 children enter foster care annually, often not due to physical abuse, but because their families are living in poverty.[7] Removing a child from their parents' care, even for a short period of time, is a drastic measure that can cause profound harm.

Surveillance and Policing of Poverty

In the US, child welfare involvement typically begins with a report to a government-run hotline, where suspected child abuse or neglect can be reported. In 2019 alone, 3 million reports were made about nearly 8 million children.[8] Over 80 percent of the children whose cases were investigated were found not to have been abused or neglected.[9]

Low-income families are far more likely to be reported to child welfare authorities. This is partly because they more frequently receive poverty-targeted social assistance programs, including to pay for food, housing, and medical care, and frequently interact with professionals who administer them. In the US, many of these professionals are legally required to report suspected child neglect or abuse-again, under definitions of those terms that often include circumstances which are a function of poverty. As a result, families in poverty face heightened scrutiny, even when their circumstances stem from financial hardship rather than parental wrongdoing.

Data analysis by Human Rights Watch shows a correlation between poverty and the rate of maltreatment investigations: counties with higher numbers of families below the poverty line have a higher rate of maltreatment investigations, compared to counties with higher family incomes that have lower rates of investigations (see Figure 1 in Annex).[10]

Punishing Poverty

Not only are families in poverty more likely to be investigated, but poverty-related circumstances can also trigger a maltreatment case due to "neglect" allegations.[11] The definition of neglect varies between states but often includes factors inextricably linked to poverty, and in practice opens the door to unjustified intrusion by penalizing the kinds of compromises people living in poverty often have to make. Such factors include the failure by a parent of caregiver to provide adequate:

• housing,
• food,
• clothing, or
• medical care, or
• supervision for the child in ways that threaten their well-being.

In fact, the primary reason welfare agencies remove children is neglect. In 2019, nearly 75 percent of confirmed child maltreatment cases involved neglect as defined by state statutes.[12]

Many people Human Rights Watch interviewed described how circumstances related to poverty, including housing instability and low incomes, were used as evidence of parental unfitness, either to support neglect allegations or justify family separation or termination of parental rights. Parents said they were required to complete "service plans" to reunite with their children, but the services ordered were often not tailored to the family's actual needs. For example, one parent of adolescent children was assigned parenting classes for newborns.

Once a child is placed in foster care, reunification can be extremely difficult, particularly for parents with low incomes, unstable housing, or limited legal resources.[13] As a result, conditions of poverty can prevent family reunification.

Human Rights Watch has recently conducted similar research in Western Australia, which is due to be released in March 2025. Here too, Human Rights Watch found that child protection authorities treated conditions related to poverty, including homelessness, as neglect. This contributed to the over-representation of Aboriginal children in the Western Australian out-of-home-care system. Human Rights Watch's findings are similar to those of Aboriginal controlled organizations such as the Secretariat of National Aboriginal and Islander Child Care (SNAICC), which has said that the socio-economic disadvantages disproportionately faced by Aboriginal families is often misinterpreted by child protection authorities in Australia as neglect or maltreatment.[14]

In some cases examined by Human Rights Watch, the state had failed to provide adequate support to parents needing emergency housing, including domestic violence victims, which ultimately resulted in the state's child protection authorities removing children from their parents. Human Rights Watch recommends the state of Western Australia update its child protection policies to ensure that circumstances relating to poverty should not justify child removal and where a family is facing homelessness, the onus is on the state to provide accommodation, not remove children from their parent's care. Human Rights Watch also recommends Western Australia ban evictions from public housing where an eviction would result in children and their guardians becoming homeless.

Recommendations:

The US and Australia child welfare systems should shift from punitive family separation toward family support and poverty alleviation. Federal, state, and local authorities should:

1. Ensure that poverty-related circumstances are not grounds for family separation. Strengthen due process protections for parents, including universal access to legal representation for parents engaged at any stage of a child welfare proceeding.
2. Build towards a universal social security system, including by progressively expanding programs that support families with children, such as the child tax credit in the United States, to all families.
3. Conditionalities associated with cash transfers and duty to accept "suitable" work in the United Kingdom and Germany

Some social security programs impose "conditionalities," meaning benefits are paid to recipients only if they carry out certain activities such as attending health clinics, preparing job applications and attending interviews, or accept work deemed by the authorities to be "suitable" for them. This section examines the conditionality regimes in Germany and the United Kingdom.[15]

In Germany, a major social security reform took place in 2005, when a new, secondary unemployment benefit called Arbeitslosengeld II (ALG II), was introduced. After their period of eligibility for the primary unemployment benefit, ALG I, individuals could apply for the ALG II benefit, a lower but permanent benefit tied to strict conditionalities linked to job-seeking. Under this system, authorities could withhold benefits from claimants who did not comply.

At the end of 2022, ALG II was replaced by the Citizen's Income (Bürgergeld), a program intended to be more generous and less punitive. However, its implementation fell short of these ambitions, particularly regarding conditionalities. Initially, the government pledged a one-year moratorium on sanctions for non-compliance with job-seeking requirements. However, the moratorium lasted only six months, and sanctions were reduced, but not eliminated. The Bürgergeld system still enforces job-seeking obligations, including a "duty to cooperate" and a "duty to report."[16]

As of March 2024, new legislation reintroduced the possibility of "total sanctions," allowing authorities to withhold up to 100 percent of benefits for two months if a claimant fails to comply with job-seeking obligations. While the legislation includes a provision allowing discretion in cases of "exceptional hardship," it marks a return to punitive policies.[17]

The United Kingdom's Universal Credit program, first introduced under the Welfare Reform Act 2012 and later amended, is the country's primary means-tested social security program. It was designed to replace six previously existing means-tested benefits and consolidate them into a single program.

The government at the time, and its successors, have justified Universal Credit as a simplified welfare system intended to incentivize employment. In practice, however, it imposes strict work-based conditionalities on claimants. The Universal Credit system has attracted extensive criticism, including from the previous UN Special Rapporteur on extreme poverty and human rights, the UN Committee on the rights of persons with disabilities, and the UN Committee on economic, social and cultural rights.[18]

At the center of these critiques is the use of sanctions, which allow the Department for Work and Pensions (DWP) to take deductions from social security payments from claimants who fail to meet their job-seeking obligations.[19] Under Universal Credit, at present, claimants are assigned to one of four work-related groups, depending on their circumstances:

1. No work requirements (e.g., individuals with severe disabilities).
2. Work-focused interviews (e.g., parents of young children).
3. Work preparation (e.g., people expected to work in the near future).
4. All available work (e.g., unemployed claimants required to seek full-time work).[20]

Each claimant must sign a "claimant commitment," a contract outlining work search expectations based on their classification. Failure to meet these conditions-including missing job interviews, failing to apply for enough jobs, or not increasing earnings while already employed-can result in sanctions. These can be applied at two rates (a 'high' rate which is the standard, and a lower rate typically where a person has child caring responsibilities), and for up to six months depending on how the breach of conditionality is categorized.[21]

The six-month maximum period reflects a decision in 2019 by the then Secretary of State for Work and Pensions to respond to criticism of its previous policy allowing sanctions for up to three years, and recognizing the harm that extended sanctions could have on some people.[22]

The Child Poverty Action Group found that in the first half of 2024, more than 583,000 sanctions were imposed, with the vast majority issued for failing to attend a work-focused interview.[23]

Sanctions worsen inadequacy of benefit levels

There are also broader concerns about the inadequacy of Universal Credit payment rates to guarantee the right to an adequate standard of living and other economic, social and cultural rights (ESCR), including as a result of the falling real and nominal value of a maximum cap on benefits since its introduction in 2015.[24] Benefit inadequacy is exacerbated by the negative impact of sanctions, which further reduce already strained income levels.

Human Rights Watch and other human rights organizations have backed the Essentials Guarantee campaign, led by the Joseph Rowntree Foundation and the Trussell Trust, drawing attention to this issue. The campaign calls on the government guarantee in law a level of support from Universal Credit which would cover the requirements all people need to live in dignity.[25] It calls for an independent process to determine social security payment levels, including by consulting people receiving social security support, and to ensure that the government does not make deductions, including for "work-based conditionality" sanctions, from monthly payments to leave any person receiving benefits below the established minimum level.

Recommendations:

1. Shift social security provision away from a reliance on work-based conditionality and punitive measures such as sanctions for non-compliance.
2. If applying sanctions or deductions as part of social security policy, to ensure that in no event is a claimant left with an amount below what is necessary to enjoy the right to an adequate standard of living.
3. Surveillance and Control of Poverty-Targeted Social Security

Social security systems that primarily target the poorest members of society, rather than providing universal coverage, come with significant drawbacks that tend to undermine the right to social security.[26] They are often prone to high error rates, reinforce stigma against recipients, and create bureaucratic hurdles. These systems also require extensive personal data to determine whether a claimant meets the predetermined poverty criteria, often leading to intrusive surveillance and exclusion errors.

Human Rights Watch research has found that the growing reliance on personal data to administer poverty-targeted social security programs poses serious risks to the right to privacy.[27] Over the past decade, governments worldwide have expanded their surveillance capabilities, aided by biometric recognition technologies, spyware, data brokers, and tech companies that store vast amounts of user data.

Many poverty-targeted social security programs use social registries, which are databases that collect and analyze personal information to determine eligibility for social security. These registries often support a Proxy Means Test (PMT), which is an algorithm-based system that ranks households from poorest to richest based on indicators such as:

• Employment and education history of household members
• Type of dwelling and living conditions
• Ownership of assets, such as vehicles or businesses
• Household utility consumption (e.g., electricity and water usage)

Only households meeting a specific ranking or score receive social security. This method flattens the economic complexity of people's lives into a single numerical value, often leading to arbitrary exclusions.

Between 2013 and 2022, the number of countries using social registries tripled, from 23 to 60.[28] Many of these systems were introduced with financial and technical support from the World Bank. A Human Rights Watch analysis of World Bank loan documents found that in the Middle East and North Africa, eight out of ten borrowing countries have taken out loans to create or expand social registries and related poverty targeting operations.[29]

In Jordan, the Unified Cash Transfer Program (commonly known as Takaful), uses an algorithmic ranking system to determine "economic vulnerability" and allocate benefits, drawing data from a social registry.[30] Based on interviews with claimants, government officials, activists, and an analysis of World Bank documents, Human Rights Watch found that the system denies benefits to many due to flawed and outdated data. The ranking also fuels social tension and perceptions of unfairness.

Similarly, in Lebanon, the largest social assistance program determines eligibility through a social registry and PMT, assigning households a score based on over 40 variables, including income, assets, housing conditions, and demographic characteristics.[31] Human Rights Watch found that this system has largely failed to reach its intended beneficiaries in anything like a consistent or even a coherent and predictable manner.

The World Bank acknowledges that social registries carry high risks of errors and exclusion but argues that "modernizing" them with real-time data collection and automated decision-making will improve efficiency.[32] However, Human Rights Watch research has found that the shift toward dynamic social registries heightens privacy risks and further restricts access to social security. Expanding data collection and monitoring exposes beneficiaries to greater surveillance and data vulnerabilities without necessarily improving outcomes.

In addition, more data does not always mean better decisions. People may be wrongly excluded because their circumstances do not fit predefined eligibility criteria. For example, a household that owns a car or a small business may appear more economically stable, but this fails to account for whether they can afford petrol or if their business is burdened by debt.

For these and other reasons, the Global Coalition for Social Protection Floors has deemed social registries as incompatible with the right to social security.[33]

Unlike poverty-targeted programs, universal social security schemes do not require intrusive data collection or means-testing. They are more likely to respect privacy rights while also ensuring that people receive uninterrupted access to social security.

For example, Mauritius and parts of Nepal have implemented universal child benefits, requiring only a birth certificate for enrollment rather than extensive income verification.[34] In Nepal, this benefit has not only increased birth registration but also improved access to food and reduced the likelihood of child labor.[35]

Recommendations:

1. End reliance on Proxy Means Testing (PMT) and other algorithm-based targeting methods and work towards replacing them with universal, rights-aligned social security.
2. Shift financing and technical assistance away from expanding social registries and instead invest in universal social security infrastructure, such as population registries and vital statistics systems, while ensuring compliance with privacy and other human rights law and standards.