Attacking Blood-Alcohol Evidence
Speaker: LAWRENCE TAYLOR

Click Here to Listen 

Continued from Page 2...

In any event, what is their truth? How good is this machine? Well let's look at the warranty. You didn't know it had a warranty? Got a warranty. Look in the manual. Should have gotten through in discovery. Try to. Just read something from the Intox as the Intoxilyzer 5000, "There are no other warranties, express or implied, including, but not limited to, any implied warranties of merchantability or fitness for a particular purpose." Okay, you got the expert up there on cross. "So what is the particular purpose, Mr. Jones, of this machine?" "What do you mean, what –." "Isn't the purpose of the machine to measure alcohol in the human breath?" "Yes." "The manufacturers say this isn't warranted to do that." That's what it says. "Mr. Jones, do you have a toaster?" "Yeah." "Does it have a warranty? Is it warranted to toast toast? Toast bread?" Repaired components are warranted for a period of ninety days from the date of repair and that warranty is subject to the same limitations as this warranty. This machine breaks down. When it does fall apart and IS shipped back to the factory, it goes back parts and labor for ninety days. Your toaster has a better warranty than that. Bring it down to the understanding of a jury. They understand machines, they understand warranties, they understand things breaking down. They understand ninety days parts and labor.

Okay, well, how critical is it this machine has to be scientifically precise? You want to explain to them what an unreliable machine this is and you also want to point out how critically important precision is in this case. One way to do it is to try to graphically get the jury to understand what we're dealing with. The Intox 5000 captures about, I believe it's around 70 cc's of human breath. On the partition ratio, we're talking about a very tiny amount. About a half a drop of blood in equivalency. Yeah, it's trying to measure the amount of alcohol in a half drop. You know, how much alcohol is that? Well, let's take a, for example, a .10 reading. Five one-hundredths of one percent for the sample. An amount that is invisible to the human eye. This machine is trying to measure an amount of alcohol equivalency invisible to the human eye. Now the jury is beginning to get the picture of what has to be done and what this is.

So how accurate are these machines? Well, in almost every jurisdiction they recognize something called inherent error in dealing with these machines. The expert, if you want to call him that, person from their crime lab or whoever gets up there to lay the foundation will admit, as he must, that there is something called inherent error. In many states this has been recognized by statute and/or case decision. In California, for example, its .01. In most states it's going to be .01. Any credible scientist will recognize at least .01 percent in inherent error and Hawaii is a .015 and so on and so on. Well, let's take it. What does inherent error mean? And you've developed this through your cross-exam, assuming you've got an expert up there. If not, you should have subpoenaed them and/or gotten your own defense expert. Inherent error means if everything is working perfectly, if it's calibrated, maintained and it's operated correctly. If everything—and there are no problems—none in your client with his physiology or with the machine, there's still going to be a .01 range of error. That means if you have, for example, a .10, it could be anywhere from .09 to .11 and still be deadly accurate. Think about that on cross-exam. "Are you telling me, Mr. Jones, that with this machine functioning perfectly, accurately, if everything is right, no problems—in this case it could have been anywhere from an .09 to a .11?" "Yeah." Okay, break that down mathematically. What is that? 20% range of error. From a .09 to a .11 at a .10. Now change the numbers according to your situation. 20% range of error if everything is working correctly. "Mr. Jones, would you get in an airplane, a 747, if the pilot had a 20% range of error in his flying?" "No."

Now it can actually get worse. In most states, there is duplicate analysis. There are still some that have only a single breath sample analyzer. Most require two. In those states almost always the two have to be within a certain range. Typically, as in California, it is .02 percent. That is if you come up with a .10, the next test has to be either .08, .09, .10, .11, .12. Alright? Wow! Now we have 40% range of error! You bring this out for the jury. The prosecution is offering evidence which, if everything is working correctly, has a recognized and acceptable range of error of 40%. And we're talking about proof beyond a reasonable doubt.

One method I kind of like—I don't use it—John Tarantino has something he calls a spiked aspirin. John is up from uh, Rhode Island, a very well-known DUI defense attorney and author in the field. John will get the expert to admit that there is—that these things are not foolproof. And his method is, "Well, you know Mr. Jones, it's—it—this is really, is only about 90% accurate, isn't that correct?" "No, no, better than that." "95%, 97?" Almost always the guy—he has to admit that of course it's not 100% or he wouldn't—he'd lose credibility as a scientist. So he'll say its 99% effective. 99% accurate. So then what John does, he leaves it alone and then in closing argument he gets a little bottle of aspirin. There are 100 white pills in this bottle of aspirin and he shows it to the jury. He says, "Ladies and gentlemen you've got a headache. I've got a bottle of aspirin here. There are 99 tablets of aspirin and one of arsenic. You've got a headache. Are you going to take it?" See, again, you have brought to the jury, graphically, some way they can understand, what we're talking about. Let's take that inherent error, for example. You want to bring it down to the jury's level. Ten, uh, we've got a 20% error factor here. With a .10 to .09 to .11. You go to a car mechanic. The guy says the bill's going to be $100 to fix your car. You come back ready to pay the bill and he gives you a bill for $110. "Wait a minute! Wait a minute! This bill’s $110. You told me it was going to be $100." The car mechanic says, "Hey, inherent error." Juries understand $110 versus $100.

Okay, this is the kind of machine that we've got and by the way, understand something else that is critically important for you to impart to the jury. Not only are these things inaccurate, not only does the manufacturer have no confidence in them, but they don't measure alcohol. Right. They don't measure alcohol. Please, by a show of hands, how many people here know what I'm talking about? Ladies and gentlemen, this machine does not measure alcohol! What it measures is any compound that has the methyl group in it. It's a stupid machine, okay? Their expert will call it a smart machine—state-of-the-art. Smart, meaning it's self-diagnostic and so on. It's a stupid machine. It does not recognize alcohol when it sees it. It does not measure alcohol. It measures any chemical compounds that contain the methyl group. Dick Jensen has a really neat little way of explaining that for juries. He has little models of molecules, atoms and so on. A very graphic way of doing it. There are thousands of compounds out there in the universe that contain the methyl group. More importantly, there are a lot of studies that have been done and you can see them in the various DUI books that deal with the subject. One study finds 102 compounds on the human breath that contain the methyl group. Another study, almost 200 compounds on the human breath that contain the methyl group, meaning the machine will record it as alcohol. What's worse, it is cumulative. If you have 14 different compounds, it's gonna, machine's gonna add them up, because if you know the concept, is there's a breath chamber, infrared light shoots through, the more infrared light that's absorbed by the methyl group, the less gets through on the receptor and the higher the blood-alcohol level goes, right? It is cumulative. The machine adds all those up and reports it as alcohol.