Hillary Clinton has called for “ending the era of mass incarceration,” joining a chorus of politicians—both presidential candidates and members of Congress—highlighting the need for reform. Her plan includes reforms to mandatory minimum sentencing and increased alternatives to incarceration for low-level nonviolent drug offenses. In April, Clinton delivered a major policy speech outlining her plans to tackle mass incarceration, and has been outspoken on the campaign trail about the importance of comprehensive criminal justice reform. Unfortunately, this rhetoric does not match the reality of her campaign proposals.

As we have demonstrated with the state Prison Population Forecaster, tackling mass incarceration is a challenging task that will require ambitious reforms. While addressing low-level drug offenses is important, it is only the first step. There is little consensus as to what “low-level” means—if you ask 100 people in Washington to define “low-level,” you will probably get 150 different definitions. Where people draw the line on the seriousness of the current offense coupled with prior criminal history to still qualify as “low-level” will vary so widely as to make the terminology useless.

However, let’s ignore the “low-level” exception and just estimate the impact of changes to admissions for all drug offenses, which is a proposal far more ambitious than that proposed by Clinton or any other candidate. At the state level, cutting admissions for drug offenses in half will reduce the prison population by 7 percent by 2021 compared with the baseline projection.

But remember, there are 1.5 million people in prison. A 7 percent reduction is a far cry from “ending mass incarceration.” Obviously, the impact would be far less when limited to “low-level” offenses.

Even in the federal system, where people convicted of a drug offense comprise half of all those in prison, the impact of limiting policy reforms to drug offenses is substantial. Halving admissions for drug offenses, which is incredibly ambitious, would only reduce the federal prison system by 17 percent.

These reductions are not trivial, but rolling back 40 years of prison growth will require reforms to admissions and length of stay for all offenses, including violent crimes. In the states represented in our forecaster, reducing admissions for all offenses by half will cut the prison population by 37 percent by 2021 compared with the baseline projection.

Another way to reduce the scale of mass incarceration is to address proportionality in length of stay. In the states, cutting length of stay for all offenses by half will cut the prison population by 39 percent by 2021.

Achieving the goal of reducing mass incarceration is an important policy objective that will require reforms that reach far beyond low-level nonviolent drug offenses. Clinton and other presidential candidates who have echoed a desire to address mass incarceration will need to set their sights higher.