No More Field Sobriety Tests in Michigan . . . for now

Listen to my testimony before the Senate Judiciary Committee on Field Sobriety Tests in Michigan!



Field sobriety tests are no longer admissible at trial in Michigan drunk driving cases.  I explained in a related article, Preliminary Roadside Analysis: Free Motion to Suppress FSTs, how MCL 257.43a and MCL 257.625a were amended last year resulting in exclusion of these tests at trial.  The drunk driving laws were amended to make field sobriety tests mandatory (which, I explained, is unconstitutional), but the new law excluded these tests from evidence except for probable cause determinations.  I distributed a free motion to suppress these tests at trial, and my friends in the defense community are reporting that judges across the state are granting the motion.  Prosecutors are complaining that this law was passed by mistake.  The Michigan Senate agrees, explicitly admitting that the law was passed by mistake, but they hope to quickly fix it.


Field sobriety tests should not be admitted at trial as evidence of impairment.  These tests are highly misleading and prejudicial, and the data behind field sobriety tests proves that these tests should not find their way before a jury. 


It is important to understand that certain types of evidence are admissible for different reasons and at different stages of a criminal case.  Currently, field sobriety tests are admissible at a preliminary examination and at a motion hearing challenging the arrest.  A person cannot be arrested without a warrant under the 4th Amendment unless an officer has "probable cause" to believe that the accused was committing a crime such as drunk driving.  If the accused claims that the officer had no justification for the arrest, the prosecutor can admit the field sobriety tests and any preliminary breath test to prove probable cause.


PBTs, little handheld roadside breath tests, are not admissible at trial because they are deemed too unreliable and too inaccurate. For example, several people have been reportedly arrested for drunk driving after consuming Fisherman's Friend cough drops. After the arrest, evidential breath testing devices, which are much larger and work on different scientific principles, revealed that these motorists had not consumed any alcohol. The falsely arrested motorists cannot sue the police for false arrest, however, because probable cause existed to support the arrest based upon the preliminary breath test. 


A prosecutor must be able to establish that evidence is both relevant as well as reliable before it may be admitted in court before a jury. PBTs are certainly relevant, since the breath test occurs at a time much closer to the time of driving than the evidential breath test taken at the police station or a blood test hours later at a hospital. But no matter how much they might want to admit this relevant evidence, prosecutors cannot prove that the PBT is reliable enough to pass muster.  With a few narrow exceptions, this is longstanding black letter law in Michigan.


Prosecutors have never been compelled to prove that field sobriety tests are reliable. Judges have typically ruled that these tests are simple observations made by police officers in the field, akin to a police officer describing red glassy eyes and the odor of intoxicants. Unfortunately, while coyly admitted as simply observations, these tests are anything but simple.  The testing protocols require elaborate instructions, demonstrations and a systematic scoring system.  In all reality, these tests purport to be the pinnacle of science, and juries will frequently return guilty verdicts based upon these tests, standing alone, without any breath or blood test. 


The problem is that these tests are complete and utter bullshit. If we want to compare the reliability of a PBT to field sobriety tests, the PBT wins hands down. Field sobriety tests are so inaccurate, as a matter of fact, that the statistics are mind boggling. Take for example the HGN eye test.  This test has been repeatedly described as the best and most accurate field sobriety test by the federal government and courts across the nation.  In 77% of cases where a person is intoxicated, the officer will make the correct arrest decision by arresting the motorist.  But a recent study by the federal government itself revealed that officers will falsely arrest a sober person 77% of the time.  The ugly truth is that this means that people will fail the HGN test 77% of the time, whether they are sober or intoxicated.  


Exactly what does the HGN test prove?  In the case of Joanne McKown, an Illinois driver accused of drunk driving, it proved that she was drunk, beyond a reasonable doubt.  HGN was the single test admitted against Ms. McKown at trial, and the case made its way all the way to the Illinois Supreme Court to determine whether the HGN test was admissible in Illinois.  The Illinois Supreme Court held that, despite all of its failings, if the test was properly conducted by a trained officer, the test was admissible. Fortunately for Ms. McKown, however, the test was not conducted properly, and her case was reversed. The same study that determined that sober people will fail the HGN test 77% of the time when conducted properly found that when the test is improperly performed, sober people are falsely arrested 92% of the time!  Say what you will, but PBTs do not falsely arrest people 77% or 92% of the time.


Here are just a few more summary points on field sobriety tests.  Keep in mind that the standardized field sobriety tests are far more accurate and more reliable than nonstandard tests, and these statistics become shocking.  (And also keep in mind that most of the statistics regarding these tests are not published.  This data was harvested by reviewing what has been released.)

  • In a 1977 study, the first study to ascertain which tests were best suited to use as field sobriety tests, 47% of sober motorists were falsely arrested using the three test battery.
  • In a 1981 study, 32% of sober motorists were falsely arrested.  18% of those falsely arrested had absolutely no alcohol in their system.
  • A 1986 study regarding the HGN test by Good-Augsberger determined that the eye test was 92% accurate. This means that drunks were correctly arrested 92% of the time. Unfortunately, however, 82% of sober people also failed the test.
In 1995, the National Highway Traffic Safety Administration sponsored a study in Colorado. NHTSA claimed that the field sobriety test battery was 93% accurate. But:
  • The average blood alcohol content of studied motorists was .15, nearly twice the legal limit. 
  • 12% of people under .05 failed the HGN test. 


In 1997, NHTSA conducted a Florida study. They claimed that the field sobriety tests were accurate 95% of the time. But:

  • The average blood alcohol content of studied motorists was .15, nearly twice the legal limit. 
  • Over 50% of people below .08 failed the HGN test. 
  • Sober motorists who were falsely arrested averaged 3.6 clues on the walk and turn and 2 clues on the one leg stand.
  • More than 70% of sober motorists failed the walk and turn test.
In 1998, NHTSA conducted a study in San Diego California.  Correct arrest decisions were made in 91% of cases. But:
  • The average blood alcohol content of studied motorists was .15, nearly twice the legal limit. 
  • 37% of sober motorists failed the HGN test. 
  • 53% of sober motorists failed the walk and turn test. 
  • 41% of sober motorists failed the one leg stand test.
  • A sober motorist was six times more likely to be falsely arrested than a false release, where an intoxicated motorist is not arrested.


SB 207 was proposed last Thursday, March 12, 2015.  It is designed to "fix" the current law.  The bill was fast-tracked for hearing before the Senate Judiciary Committee on Tuesday, March 17, 2015, and it was expected to pass out of committee for a vote as early as the next day. I rushed up to Lansing on St. Patrick's Day, tossing the beautiful green jacket and tie combo I had picked out for the holiday, swapping it for a conservative blue suit and red power tie. I didn't think that the Senate Judiciary Committee would appreciate the frivolity of a holiday that involves drinking green beer while addressing the topic of drunk driving in Michigan.


It doesn't surprise me that Michigan's full-time legislature could pass a law by mistake.  While Michigan lawmakers put in more time than a part-time legislative body, lawmakers focus on passing laws as a measure of success. As a result, we have more criminal laws than all of the other states in the Midwest. A recent study by the Mackinac Center for Public Policy found that Michigan’s 918-section criminal code is more than twice the size of Ohio’s and Wisconsin’s. Michigan has at least 3,102 crimes, and adds 45 to its criminal code each year.


I hope that this new bill does not pass, but you can listen to the comments and questions from Sen. Rick Jones and Sgt. Perry Curtis from the Michigan State Police.  Both believe, based upon their years as law enforcement officers, that the alphabet and counting backwards test are somehow statistically relevant and reliable tests, even though those tests are less relevant and less reliable than the standardized field sobriety tests. As Sgt. Curtis admits, these tasks have never been validated.  What he doesn't admit is that these tests were explored, tested and excluded from the standardized test battery. Since 1977, there has never been any effort to revive these so-called tests.


The problem is that law enforcement officers tend to rely wholly upon their experience and have little formal training in statistics.  Because Sen. Jones and Sgt. Curtis have seen drunk people fail to properly recite the alphabet, this makes it a "good" test.  While it is absolutely true that some drunk people will fail to properly recite the alphabet, many sober people will fail too because they are nervous or haven't recited the alphabet in 20 or 30 years. At the same time, some intoxicated people will correctly recite the alphabet.  This is a common misunderstanding, rooted in confirmation bias and outcome bias. During the course of a law enforcement career, a police officer will recall times that a drunk person failed the alphabet, and the arrest decision was justified. This provides a cognitive basis for believing that the test is valid, even though it isn't. The officer will tend to place less emphasis, cognitively in terms of memory, on those instances where sober people failed to properly recite the alphabet when the officer released the sober subject and on instances where intoxicated people passed the test, explaining it away without actually acknowledging the statistical significance of this event. 


The science behind field sobriety testing is not easily grasped by most people.  We all tend to believe that drunk people lack coordination because we have seen drunk people; therefore, a test that reveals a lack of coordination proves that a person is drunk.  The problem, as detailed by Greg Kane, MD, is that:


Mathematically the accuracy of the SFST test at identifying drunkenness is the same as (actually less than) the accuracy of an airport security detector test at identifying guns. It makes as much sense to arrest people for being uncoordinated as it does to arrest people for making a metal detector beep. When travelers set off a metal detector, we don't arrest and convict them of gun smuggling and terrorism. Everyone with a gun sets off the metal detector, but most people who set off the metal detector do not have a gun. Metal detectors are highly accurate at spotting people who do have a gun, but they are not especially good at telling the difference between gun and no-gun. On innocent people metal detectors often give the wrong answer. They ring for keys and phones and belt buckles, for lots of stuff. SFSTs work the same way. It is true that everyone who is drunk is uncoordinated. But it is not true that everyone who is uncoordinated is drunk. If you arrest people for being uncoordinated, you'll end up arresting, and convicting, lots of people who are innocent.