More than a century after public policies dramatically reduced child labor in the United States, the topic has burst back into the headlines in an unlikely manner: President Trump’s now-withdrawn nominee for the Federal Reserve Board, Stephen Moore, came under fire for suggesting the repeal of child labor legislation. During a panel on the minimum wage at the 2016 GOP convention, Moore said: “I’m a radical on this, I’d get rid of these child labor laws. I want people starting work at 11, 12.”

As a self-described “radical,” Moore has been treated as an outlier on child labor. But that’s not the case. During his 2012 presidential campaign, former House speaker Newt Gingrich advocated for relaxing “truly stupid” child labor laws and employing children as school janitors. In 2016, the Acton Institute, a conservative nonprofit that received donations from Education Secretary Betsy DeVos, called child labor “a gift our kids can handle.” Earlier this year, an Indiana Republican state senator wrote a bill to weaken child labor regulations, while the Trump administration recently moved to weaken protections of child laborers working in agriculture.

In short, there is a significant push on the right to resurrect child labor. That push is based upon two key myths that do not square with historical reality: First, that markets, not the government, wiped out child labor in the early 20th century, and second, that child labor was beneficial to children’s work ethic, education and employment prospects. The history of child labor belies these assertions — and their rise in high political circles threatens to resurrect a problem that the United States largely eradicated a century ago.

The figures on child labor are notoriously unreliable. Census data, from which historians’ estimates are derived, are best understood as a minimum and show that child labor peaked between 1890 and 1910 with at least 18 percent of all children between 10 and 15 years old in gainful employment. By the 1930s, this figure had dropped to about 5 percent.

A complex range of forces, including legislation, led to this dramatic decrease in child labor. Some of the earliest child labor opponents were unions. Cognizant that competition from children forced down wages for adults, labor unions were instrumental in removing children from many industries.

Changing norms about children’s role in the family also created an increasing stigma on child labor for working families. Families began opting for mothers rather than children as the first choice for a second wage-earner. Reforms such as mother’s pensions and workers’ compensation laws reduced pressure on working families in times of crisis and extreme poverty, which also decreased their reliance on child labor.

In addition, courts decided that contributory negligence in workers’ compensation cases could not apply to children. In practice this meant that when a child was injured at work, employers could not avoid paying damages by blaming the accident on the child, which indirectly raised the cost of child labor.

Meanwhile, a nationwide network of reformers ensured the passage of compulsory education laws, as well as state and local bans on child labor. By 1912, every state and the District had passed child labor legislation. The Southern states, unwilling to support schooling for black children, resisted compulsory education laws, but by 1918 even Florida, Georgia, Louisiana, South Carolina, Texas and, finally, Mississippi had passed some level of school attendance requirement for children.

Although enforcement of local regulations remained patchy and employers of children could often circumvent them, at a minimum these regulations drove down the official numbers of children working by obliging many employers and parents to conceal child labor.

Put these factors together, and it becomes clear that governmental action of every sort and at every level was instrumental in the drop-off in child labor.

Even so, all these changes did not eradicate the practice. Far from it. As children were pushed out of mills, mines and factories, they continued to work in more informal settings, such as on the streets, at home and, particularly, in agriculture — something that continues to today. Recent studies have shown the majority of children have done paid work by age 16. This work usually consists of activities such as babysitting and yard work, but an estimated half-million children younger than 18 still work in agriculture in the United States.

In spite of this clear history, conservative thinkers today often dismiss the government’s role in the reduction of child labor. They cite Congress’s passage of the 1938 Fair Labor Standards Act, which enacted a permanent prohibition on employing children, long after child labor had started to decline, as evidence that markets rather than legislation ended child labor. By the time Congress acted, rising wages, the argument goes, had enabled families to keep children out of work and in school.

But the history of child labor reveals that it is more likely that earlier government actions that increased the cost of child labor, reduced the number of working children and improved education led to an increase in adult wages.

Regardless, new child labor advocates go further, rhapsodizing about purported benefits of the practice, such as teaching children work ethic, providing superior training for work than schools offer and improving young people’s entry into the workforce. They say the practice is a win-win: good for the economy and good for children who will find work more exciting than the classroom.

These arguments are tired and recycled: Advocates of child labor in the early 20th century made the very same points only to be contradicted by facts. While work ethic was particularly difficult to evaluate, the data was unambiguous. Education rather than labor proved a clear winner for children: improved lifetime earnings, employability and productivity, and reduced illiteracy rates. Child laborers gained skills, but they were of the specialized sort that were tied to menial roles and not readily transferrable. Child nut shellers, to name but one example, impressed and appalled reformers with their speed and dexterity at cracking nuts with fingers and teeth, skills that were unlikely to lead to higher-paying jobs down the line.

Child labor was also linked to poor public health. Children faced higher rates of accidents than adults in industrial jobs. Child laborers were also shorter and weighed less than their peers who did not work, both signs of poverty. At best, child labor was a symptom of poverty; at worst, it was a cause.

If anything, then, instead of resurrecting child labor, we need to strengthen protections for child workers. Our focus should be on regulating harmful labor practices, improving education and communally assuming the burdens of developing human capital, which have increasingly fallen on families, thanks to rising student loan debt, unpaid internships and underfunded public schools.

Most children working today aren’t doing backbreaking menial tasks at the expense of being in school, but more than half of the 452 work-related deaths of children in the United States between 2003 and 2016 were in agriculture. We also need to broaden our definition of labor, especially in the face of children earning money on the Internet, and strengthen protections for the many children who continue to work.

Rather than being ideal training that prepares the next generation for success, the history of child labor exposes it for what it is: exploitative, harmful to adult wages and resulting in less healthy, less educated youth who were less likely to escape poverty.