G.S. 20-138.1(a)(2) prohibits a person from driving a vehicle upon a highway, street or public vehicular area after having consumed sufficient alcohol that the person has, at any relevant time after the driving, an alcohol concentration of 0.08 or more. S.L. 2006-253 amended this subsection to provide, effective for offenses committed on or after December 1, 2006, that “[t]he results of a chemical analysis shall be deemed sufficient evidence to prove a person’s alcohol concentration.” The meaning and constitutionality of this latter provision were considered and upheld by the court of appeals in State v. Narron, 193 N.C. App. 76 (2008) (discussed in this post). Narron held that the “shall be deemed” provision, contrary to the defendant’s argument, did not establish a mandatory presumption. Instead, it merely set forth the prima facie standard for proof of impairment under the per se prong of G.S. 20-138.1. The court explained that the “[r]esults of a chemical analysis are sufficient evidence to submit the issue of a defendant’s alcohol concentration to the fact-finder” who “may find it adequate proof.” Id. at 81, 84.
Let’s suppose you are a defense lawyer (a proposition that is hardly a stretch for many readers) representing a defendant charged with impaired driving based on a reported alcohol concentration of 0.08 resulting from a breath test administered in compliance with all statutory and regulatory procedures. On what basis might you argue to the judge in district court or the jury in superior court that the alcohol concentration result introduced into evidence by the State fails to prove the defendant’s alcohol concentration beyond a reasonable doubt?
One argument made with some frequency is that the defendant showed no outward signs of impairment. She drove well and satisfactorily performed field sobriety tests. This, the defense attorney argues, casts doubt on the veracity of the alcohol concentration results reported from the chemical analysis. There’s nothing at all improper about this argument framed in this way. Determining whether the evidence establishes the defendant’s guilt beyond a reasonable doubt unquestionably is the province of the finder of fact. See, e.g., State v. Finch, 244 P.3d 673, 679 (Kan. 2011) (stating that proof of the elements of a per se criminal statute will not compel conviction as a matter of law as “[t]he defense may still attack the State’s proof and attempt to discredit its witnesses, their machines, and their methods during the State’s case-in-chief or later” and “[t]he jury may finally agree that reasonable doubt prevents a conviction.”) Moreover, “[t]he State’s introduction of evidence supporting the statutory elements in a per se criminal statute does not endow the evidence with infallibility.” Id.
There’s a variant on this argument, however, that is not supported in law. Defendants sometimes argue that the State’s evidence is insufficient as a matter of law to establish impaired driving under the per se prong unless, in addition to proving the defendant’s alcohol concentration, the State also proves that the defendant was appreciably impaired. This argument may be hold-over from the statutory scheme that preceded the Safe Roads Act of 1983, which defined a per se violation of the impaired driving laws as a lesser included offense of driving under the influence and under which the results of a chemical test yielding a result of 0.10 or more created a presumption that the person was under the influence. See G.S. 20-138 (Cum. Supp. 1981); G.S. 20-139.1 (Cum. Supp. 1981); State v. Shuping, 312 N.C. 421 (1984). Whatever its origins, this argument reflects a misunderstanding of the impairment element of impaired driving as a single element that may proved in any one of three ways. See State v. Coker, 312 N.C. 432, 440 (1984); Narron, 193 N.C. App. at 79. As the court of appeals clarified last month in State v. Arrington, __ N.C. App. __ (August 16, 2011), “it is not necessary for the State to prove that the defendant was appreciably impaired, uncooperative, or driving in an unsafe manner in order to prove that defendant is guilty of a violation of N.C. Gen Stat. § 20-138.1(a2). To prove guilt, the State need only show that defendant had an alcohol concentration of 0.08 or more . . . .”
Another argument relied upon by defendants is that the “margin of error “ for the breath testing instrument renders the State’s proof of per se impairment based on a breath alcohol concentration of .08 unreliable. The argument generally points to one of two sources for the margin of error. First, administrative regulations deem a breath-testing instrument to be accurate if the control sample used to verify instrument accuracy before the defendant’s test measures at the expected result of 0.08 or 0.01 or less than the expected result. See 10A NCAC 41B .0101. Second, G.S. 20-139.1(b3) deems admissible results of a chemical analysis consisting of “two consecutively collected breath samples [that] do not differ from each other by an alcohol concentration greater than 0.02,” and provides that “[o]nly the lower of the two . . . can be used to prove a particular alcohol concentration.” G.S. 20-139.1(b3). Under the first basis, the margin of error is 0.01 (though, ironically, any such variance engenders a lower alcohol concentration result than actually is present); under the second, the margin of error is 0.02. While alleged unreliability based upon a margin of error, like other questions about the reliability of a reported alcohol concentration result, is fair game for the fact-finder’s consideration, see, e.g., Finch, 244 P.3d at 679, an alleged margin of error does not render the State’s evidence of impairment insufficient as a matter of law. See Shuping, 312 N.C. 421 (1984) (rejecting defendant’s challenge to the sufficiency of the evidence based on an alleged margin of error and characterizing the 0.01 deviation allowance below the expected reading as “a safeguard to ensure that when the actual test is subsequently run, any possible error during actual testing is in favor of defendant”); Arrington, __N.C. App. at __ (rejecting defendant’s contention that since his reported alcohol concentration of 0.08, the result from both breath tests, was the lowest for which he could be convicted of a per se violation, the “‘margin of error’ of the [instrument] should be taken into account to undermine the State’s case against him”; determining that the testing satisfied statutory requirements, was reliable, and accurately identified the defendant’s level of impairment.)
If there are other common bases on which alcohol concentrations from chemical analysis are alleged to be unreliable, or other relevant grounds on which proof of per se impairment is alleged to be lacking, please send in a comment to let me know.
I don’t know of any “margin of error” but rather an actual breath composition result. The instrument measures the result, displays it, and then prints the results. The .02 as frequently reference is due to variables from the individual due to the breath composition. Since the body metabolizes at an average rate, we are left with what is expelled from the lungs. NC law indicates the instrument not to display/record the third value unlike in other states aka truncation. The .08 may very well be .089 or could be .080 but, we will never know. But here is one final thought: the breath result is actually LOWER than the blood result. The defendant really is given more than a benefit of the doubt, he/she is provided with a LESS than actual alcohol concentration result.
NC law does not indicate the instrument to record the third value. Even though it may not be displayed, it is recorded. Frequently the different results may be indicative of variables noted but may also be attributed to the inaccuracy of the machine as well. I think an argument could be made that due to “truncation” and results differing by reported vales of 0.02 may be actually results of more than 0.02 ,i.e. 0.08 and 0.10 reported “results” may actually be 0.080 and 0.109. This MAY reflect an inaccurate and/or inadmissible result. I would further say that usually blood results are higher but would NEVER concede that they are “actually” higher
This is not true. The value is not less, but more the was they are calibrated. Kinda like he evidence lab problems.
The 2100:1 blood to breath partition ratio is probably one of the most successful points I have made in front of juries. Juries come to understand that the conversion is based on the average human being and that not all people have the same partition ratio. I have had jurors tell me that they are reluctant to convict a .08 if the appreciable case is thin and the State fails to adequately explain how the instrument makes the conversion from blood to breath. Many times the State simply relies on the test ticket and completely fails to show the jury why they should find the result RELIABLE beyond a reasonable doubt.
Some food for thought: in light of Melendez-Diaz, is the statute that allows preventative maintenance to be judicially noticed unconstitutional? If the State wants to tell the Court that preventative maintenance was performed within the time limits, which is a foundational requirement to the admissibility of the breath test, and is certainly offered for the truth of the matter asserted and is done in anticipation of litigation, should not the forensic alcohol person who performed the maintenance be required to testify live in order not to violate the Confrontation Clause. Would love to see a blog on this topic.
I think it is terrible to drive drunk, but I have a problem with the methods used to test and maintain the breathalyzer machines. My son was charged with DWI and I examined the methods used to test and calibrate the machines and it is my belief they are sending many potentially innocent people to jail, taking their cars, and causing them to pay high insurance. And the sad fact is that you can’t question their methodologies in court as long as their incorrect methods are law.
I have a Bachelor of Science degree and have experience auditing Fitness for Duty and DOT testing programs and the methods used by North Carolina would not be acceptable for evidential grade equipment under those programs. I called the manufacturer of the equipment to ask them questions about how to perform testing on the machines and they would not talk to me and I called the NC state people who maintain the machines and they were not knowledgeable at all about the QA requirements required for testing the breathylzer machines. I had to get the US Department of Transportation to send me the manufacturer’s QA plan for testing the machines to meet evidential grade requirements since the manufacturer would not. And believe me the state is not close to meeting the QA requirements. They told me the machines used in NC did not have to meet DOT requirements even though the manufacturer submitted a QA plan to DOT which was the minimum requirements for testing the machines to be evidential grade. These machines could be very inaccurate in the non conservative direction the way they are tested, but you can’t challenge this in court. I have a multi page report that proves these machines are not accurate the way they are tested. The state of NC holds commercial drivers to the DOT requirements, so they are required to follow their QA plan requirement, but they don’t. Many innocent people are being convicted of DWI in NC.
I would be glad to meet with you and explain why these machines are not calibrated correctly with an external standard. Just e-mail me and I will be glad to come present my evidence. Some one who reads .08 could have an actual reading or .06 the way the machines are tested. The state seems to want to test the machines in a way that cost the least money instead of doing in a manner that meet evidenciary requirements. The test they perform every 3 months is a waste of time. This test is performed each time they test someone.
I would love to see this research, I have started down this rabbit hole.
The reduction of freedom is offered up to us as protection. You will be unlikely to find someone who would say driving drunk is a good idea. The machine on which BAC is determined are notoriously faulty and the last round of DWI law reform places people making a mistake in a position of a willful felon. Considering the inaccuracy fo these machine, some people mistakes may not have been actually over the limit.
I have been defending DWI for more than a decade now. Although I don’t personally drink, it is a passion of mine protecting the accused from this system that is in need of a major upgrade.
One reason the machines are inaccurate is that they are not tested with an external standard. LE uses the internal standard used to perform original calibration. This standard is a gas standard that has an accuracy of plus or minus 2%. This means the actual concentration could be 0.0784. The acceptance criteria is plus 0.01 and does not have a limit for minus direction. Let’s say they calibrate the machine with a 0.0784 to be 0.08. They do not test to 3 digits so the third digit is truncated. This means a 0.0784 gas canister could be reading 0.089 and still pass at 0.08. That alone could allow the machine too read any alcohol level above 0.784 as 0.09. That would also mean someone with an actual breath alcohol level as low as 0.067 would read 0.08. There are a lot more variables that are wrong. The acceptance criteria should be plus or minus 0.004 and an external gas cylinder should be passed through the breath tube which is not done. This does not consider different elevations for which corrections are supposed to be made. From looking at PM records the machines read a 0.08 gas standard the same in Boone as at the coast. Unless the machines compensate for elevation differences this could cause major errors. I noted on the Arkansas website their machines read different at various elevations, which would indicate the machine does not compensate for elevation. There is much more deficiencies in the calibration of the machine. This info is just a start.
Both weather conditions and operating at elevations other than sea level will change the absolute pressure from 760mm of Hg and cause the expected value for the dry gas standard to change. It is important to account for changes in absolute pressure when running accuracy checks and calibrations. The Intox EC/IR contains a precision pressure sensor which monitors the absolute pressure and automatically corrects the expected value of the dry gas standard for the current pressure at the moment. The value is displayed before the accuracy check and is printed on an accuracy check printout. (Example: In Sata Fe New Mexico at 7000 ft. elevation, given normal atmospheric conditions, using an approved dry gas standard labeled with a gas value of 0.038 grams of alcohol per 210 liters of breath, the INTOX EC/IR will display the expected dry gas value as 0.029. If an accuracy check is run when the expected value of the dry gas standard is 0.029, the tolerance requirement is met if the accuracy check result does not differ by more than plus or minus 0.005 grams per 210 liters of breath of the expected (0.029) value.) North Carolina does not test to three digits and their acceptance criteria of 0.01 is twice that required to be an evidential grade instrument as required by the manufacturer. Looked at recent accuracy check from Sylva NC which is 2500 ft above sea level and the accuracy check indicated 0.08 which is no different from reading in Wilmington at sea level. Boy there sure are a lot of questions here. Why does the state not elect to calibrate machines in accordance with manufacturer’s recommendations to meet evidential requirements?
Now let’s apply the possible error to the acceptance criteria allowed by law. The acceptance criteria is 0.08 plus 0.01, so the machine could be reading 0.09 during the initial calibration with a new dry gas standard calibration and be acceptable and we know this standard is supposed to be 0.08 plus or minus 2%, so it could be as low as 0.0784. If a 0.0784 is calibrated and the machine reads 0.09 it is acceptable by the PM standards and since the machine truncates the third digit it could be reading 0.099. That would make the error 0.0266 and anyone whose alcohol concentration was over 0.0784 would read a 0.010. Why is a red flag not being thrown for this? And since the state does not use a separate dry gas standard to perform PM and accuracy check every 3 months, they would never know it was off this much. It is a huge problem that the state does not test to the third digit and use a separate dry gas standard to perform PM accuracy checks through the breath tube. And when you add the cumulative error of dry gas error due to elevation changes the machines could be inaccurate in the non conservative direction more the 0.03-0.04. Why is the state allowed to write their own QA program and not follow the QA Program specified by the manufacturer for DOT testing that is the standard used by all federal agencies? Innocent people are having their driving privileges taken, their cars taken , their freedoms taken and in some cases going to jail based on instrument reading that are not reliable as evidence in court.
Please disregard my contention that the accuracy check is 0.08 plus 0.01 in earlier posts. The acceptance criteria is actually 0.08 minus 0.01.
So the test performed by the State of NC allows twice as much error as the manufacturer recommends in the negative range. They do not allow any variance in the positive range, so the acceptance criteria is 0.08 -0.01 grams per 210 liters of breath. However there is no way the machine can be verified to not have error in the positive direction since they only test to two decimel places. The machine could be reading 0.089 during calibration or accuracy checks, but would only read out as 0.08 since the machine truncates the third decimal point. So the acceptance criteria is actually 0.08 +0.09, -0.01 gram per 210 liters of breath. This means there is no way the machine can be verified in accordance with NC requirements because it cannot read a third decimal point as other states do.
This also presents the scenario where the machine could be calibrated using a 0.080 standard and pass while actually measuring 0.089 and reading out0.080. This would mean the entire range of the machine would have the +0.09 bias and someone blowing 0.071 could read out as 0.08 and fail the test even though they are below 0.08.