Elected local judges in the U.S. hand down harsher sentences for crimes like rape and murder when they are up for reelection — but only in states where those elections are competitive, according to a new study that provides the broadest look to date at how America's unique system of electing judges affects sentencing.

While previous studies of so-called "electoral sentencing cycles," through which judges impose harsher sentences when up for reelection, had individually examined the states of Pennsylvania, Washington, North Carolina and Kansas in great detail, the new study in the Journal of Comparative Economics added data from seven previously unexamined states, complicating researchers' understanding of sentencing cycles. 

"If previously people thought there were electoral cycles in sentencing — no, that is not always the case," co-author Mikhail Poyker of the University of Nottingham told The Academic Times. "This is all very state-specific."

Poyker wrote the paper alongside Christian Dippel of the University of California, Los Angeles. The pair spent three years on the paper, spending months obtaining sentencing data from states and talking to legal experts across the country. 

While Poyker and Dippel confirmed the presence of electoral cycles in Pennsylvania, Washington and North Carolina, they did not observe similarly strong effects in the other states they examined, which included Virginia, Colorado and Georgia.  

Virginia, where judges are appointed by the state legislature rather than through voting or by the general public, served as a "quasi-placebo" for the study, the researchers wrote. There, the researchers observed that judges actually handed down slightly shorter sentences ahead of state elections, though the results were not statistically significant. 

In other states, like Tennessee, judges are first appointed by the governor and then must stand for reelection later on. In some states with this system, entrenched political norms can discourage serious electoral challengers, possibly due to potential challengers angling instead to be appointed directly by the governor, the researchers said. 

Poyker and Dippel found that judges in states like these generally exhibit little or no change in sentencing patterns around elections. 

"In approximately half of the cases [in the study], judges do not face any competition," said Poyker. "In those states, there's zero probability change in sentencing." 

In another group of states, however, judges are elected directly and different political cultures can mean elections are more competitive and judges hand down statistically significant longer sentences for severe crimes prior to elections. This is especially true in states like Pennsylvania, where judges openly associate with the Democratic or Republican parties rather than running as nonpartisan, the researchers found. 

"In some states, competition is so tough, and if you show mercy to a defendant, your competitor will go to a local newspaper and complain," said Poyker, who earned his Ph.D. at UCLA. "They have to become tougher if they want to be reelected." 

In addition to measuring electoral competitiveness by whether or not candidates are associated with political parties, Poyker and Dippel also used political donations as a measure of competition. 

Millions of dollars can flow into campaigns in states with partisan judicial elections. In 2015, spending around Pennsylvania's supreme court race topped $15 million, for example. That court played an important role in defeating former President's Donald Trump's attempt to overturn the 2020 presidential election results through baseless accusations of widespread voter fraud. 

Because the first studies on electoral sentencing cycles focused on the states of Pennsylvania, Washington, North Carolina and Kansas, Poyker believes that some academics, journalists and pundits mistakenly believed that the pattern necessarily extended to all U.S. states where judges are elected. 

"All of the other papers were looking at one state at a time. This happened because it was very difficult to get data, even for one state," said Poyker. "Because we have many states, we could try the same empirical situation for every state and for all states put together." 

Obtaining data for 10 states was difficult, Poyker said, because each keeps records differently and public records systems can be wildly inefficient. For example, while other researchers have examined electoral sentencing cycles in Kansas, Poyker and Dippel excluded it from their study because the state wanted to charge them $8,000 for sentencing data. 

"Out of 50 states, some of them gave us data, some of them we paid, some of them gave us for free," he said. "In some cases it was a couple of weeks [to get the data]. In some cases it was a year." 

The U.S. is essentially alone in giving elections such a large role in the judicial system, a process that political scientist Alexis de Tocqueville wrote in 1835, "will sooner or later lead to disastrous results. … It will become clear that to reduce the independence of magistrates in this way is to attack not only the judicial power but the democratic republic itself." 

Poyker said that while his work shows that elections do not necessarily significantly make sentencing harsher in every U.S. state, the system can still inject unjust incentives into the legal system. 

"I think generally that electing judges is a little bit weird because when you introduce elections, you also introduce concerns for reelection that may be affected by other things rather than just your work," he said. 

The study, "Rules versus norms: How formal and informal institutions shape judicial sentencing cycles," published in the Journal of Comparative Economics on March 8, was authored by Christian Dippel, University of California, Los Angeles; and Mikhail Poyker, University of Nottingham.