When it comes to ban-the-box legislation, there are winners and losers.

Winners are people with criminal records who now have a better chance of getting work because employers can't ask prospective employees to check the job application box that asks about arrests and convictions.

Philadelphia's law was passed in 2011 and was strengthened in March 2016. Employers can  inquire about criminal backgrounds only after a conditional offer has been made.

Losers are women, primarily African American women, who have been crowded out of the labor market, and young workers "without college degrees, who see their employment reduced," said Stan Veuger, a resident scholar with the American Enterprise Institute, who used census and employment data to study the labor market impact of the legislation.

Veuger's work comes at a time when there is increasing focus -- by policymakers, politicians, advocates, scholars, and criminal-justice professionals -- on the impact on the nation's economy and social fabric when an estimated 65 million Americans have criminal records.

According to his research, one in three African American males and one in six Hispanic males have criminal records.

Veuger said his research shows that ban-the-box legislation has improved employment prospects for people with records by about 4 percent, a number that he calls "modest, but meaningful."

An unintended consequence of ban-the-box legislation may be a negative employment impact on minorities overall if employers reason that more African Americans and Hispanic people have records, and so decide to avoid hiring people whose names or faces reveal their racial background.

Employers have long been accustomed to sorting applicants for low-wage, low-skilled jobs by whether they have criminal records as "one signal" of employment suitability, he said. Scholars and advocates worry, Veuger said, that they'll use "race and gender as signals to deduce criminal records."

That may explain why employment of African American women has declined in neighborhoods where there are many African American males with records.

Veuger said there are no actual data that measure the employment situation of people who have come out of prison, so it's all a matter of inference. He drew his by looking at employment patterns in high-crime neighborhoods, inferring that those neighborhoods would also be home to people with records. Veuger examined census data-tracking commuting patterns from high-crime neighborhoods to neighborhoods covered by ban-the-box legislation.

Veuger also analyzed job postings. They indicate employers are countering ban-the-box by "upskilling," adding increased educational and experience levels to job requirements, assuming many people with criminal backgrounds lack education and experience.

That's what knocks out young people, particularly African American young people. They, like inmates in prison, haven't had the time or opportunity to acquire either education or experience, Veuger said.

Veuger's bottom line? Yes, ban-the-box legislation is helpful, but it does have consequences. "While these measures are praiseworthy and well-intentioned," he said, "to add value, we need to reduce the number of people who go through the criminal justice system" in the first place.