By Mike Dorf
Last week, Dan Alonso, the chief assistant district attorney in New York County, gave a talk at Cornell in which he promoted the recommendations of a task force (described here) that looked into reforming state laws governing the prosecution of white collar crime. Although he embargoed full disclosure of those recommendations, today (and possibly already by the time this post runs), Alonso's boss, DA Cyrus Vance, Jr., will unveil the full package of proposed reforms at a press conference. Here I want to discuss one such proposed reform that is no secret because Vance and others (including Governor Andrew Cuomo) have made the case for it already: Replacing the automatic transactional immunity that grand jury witnesses receive with use-and-derivative-use immunity.
First, the background. A witness called to testify before any kind of court has a Fifth Amendment right against self-incrimination. Yet witnesses to crime are often criminal suspects themselves. In order to prosecute the principals, the government sometimes strikes deals with lesser offenders, whereby the latter agree to testify in exchange for leniency. But even if a prospective witness does not agree to testify, he can be required to testify if the government gives him immunity from prosecution. The immunity replaces the Fifth Amendment privilege.
What kind of immunity must the government provide? In the 1972 case of Kastigar v. United States, the SCOTUS held that the government fully satisfies the Fifth Amendment by providing the witness with so-called "use and derivative use" immunity. As the term suggests, when the government provides such immunity, it may not use the witness's statements against him in the pending or any subsequent prosecution (use immunity), nor may the government use any evidence derived from the immunized testimony against him (derivative use immunity).
Although the Constitution allows for use-and-derivative-use immunity to substitute for the Fifth Amendment privilege, New York is more generous. Under New York criminal procedure law, a witness called to testify before a grand jury automatically receives immunity for her testimony, and the immunity the witness receives is "transactional immunity"--that is, she cannot be prosecuted (under state law) at all for any of the incidents about which she testifies.
So, let's suppose that a NY grand jury is investigating an alleged burglary of the home of Harry by suspect Sam. The prosecutor calls Harry as a witness and asks whether Sam had permission to enter Harry's house. Harry says no. Then the prosecutor asks whether Harry recognizes the allegedly stolen goods that were found on Sam's person, including a diamond necklace. Harry says yes. The prosecutor asks how. Harry says "because that's my diamond necklace. It looks exactly as it did the day I took it off of the body of Victoria right after I killed her." The government may not now--or ever--prosecute Harry for the murder of Victoria because by testifying about the killing in response to the question, he immunized himself. Crazy, no?
In case you think this sort of thing never happens, think again. Real cases--detailed in the Task Force Report as well as here--follow the pattern just described. There is, to be sure, one loophole for the prosecution: The automatic transactional immunity does not cover a witness who simply blurts out "I committed the following crimes" in the course of completely unrelated grand jury testimony. But even that exception has been narrowly construed, so that a prudent prosecutor must be very very careful about what witnesses she calls before a grand jury and what questions she asks.
Why, one might wonder, does NY give automatic transactional immunity? The short answer is that the statutory provisions I've just described were enacted before Kastigar, when it was arguable that the Constitution permitted compelled testimony only on the provision of transactional immunity. In Kastigar itself, Justices Douglas and Marshall took this position, and Justice Brennan likely would have as well, but he did not participate in the case.
So, why wasn't the statutory provision of automatic transactional immunity replaced with use-and-derivative-use immunity after Kastigar? It has not been for lack of trying. Both Democratic and Republican state and local executive officials have long sought to eliminate transactional immunity, but each time they have been frustrated in the state legislature. I don't know enough about the interest group politics at stake to say whether this time will be different, but I do want to close by suggesting that less may be at stake than meets the eye.
Under the post-Kastigar case law, the prosecution must show that any evidence used against a defendant who previously received use-and-derivative-use immunity derived from an "independent source." The high-water mark for defendants was the ruling by the DC Circuit in the Oliver North case, which found that despite heroic efforts by independent counsel Lawrence Walsh to isolate his prosecutorial team from any taint due to North's immunized Congressional testimony, the prosecution failed to satisfy the burden of proving independent source. To be sure, most other courts apply a somewhat less stringent standard for showing independent source, but the task is difficult enough that, in general, even in a use-and-derivative-use jurisdiction, a prudent prosecutor will do everything she can to avoid immunizing a witness she thinks she may later want to prosecute (except for perjury, which is outside the scope of immunity, but presents its own prosecutorial challenges).
So in practice there may not be that much difference between transactional immunity and use-and-derivative-use immunity. Which way does that cut?
It's not clear. On the one hand, proponents of the change can argue that use-and-derivative-use immunity is plenty protective. On the other hand, opponents of the change can say that if use-and-derivative-use immunity is de facto almost as much of an obstacle to prosecuting immunized witnesses, then there's no need to bother changing the law.