« ForrigeFortsett »
neither is reported to the Administrative Office. It was therefore necessary to make an educated guess at the former, with no information at all as to the latter.
Those disclaimers are still in order, particularly as to the amount of electronic surveillance involving agents wired for sound. As to these, all we have are informed guesses that this is very widely practiced on both the federal and state levels.
Thanks to data obtained and published by Senator Edward F. Kennedy, we now know something about the national security surveillance, however, and some of that will be discussed here. In addition, there have been some crucial legal developments in this area and they will also be reviewed.
This paper will consist of the following: (1) an introductory section containing a summary of the various findings, as well as certain methodological cave. ats and qualifications; (2) a discussion of national security surveillance; (3) a summary of the statistical data on court-authorized electronic surveillance that is contained in the annual reports from 1968–71, broken down into:
(a) The scope and variety of electronic surveillance, e.g., number of installations, people and conversations overheard, length of time of overhearing, and types of offenses involved ;
(b) the costs of such surveillance, both totally and broken down by offenses;
(c) the results so far in terms of convictions, arrests and incriminating conversations.
Throughout, information gleaned from the many court decisions already handed down, will be referred to where relevant; relevant information used in the 1971 report will also be worked in.
I. SUMMARY AND METHODOLOGY
1. Amount of surveillance
a. Court-ordered. There is a vast amount of electronic surveillance of the American people, on both the federal and state levels. A great deal of this is performed in the name of national security, and is completely uncontrolled. Much of this surveillance lasts for very long periods of time; on the state level, it is concentrated in just two states.
The bulk of this wiretapping and bugging is now used for gambling offenses, despite the original claims that it was necessary primarily for serious crimes like homicide, kidnapping and espionage: in three years, there has been only one federal device installed for kidnapping and none for either homicide or espionage; gambling installations accounted for about 90% of all the federal installations in 1971.
The figures are as follows:
In 1969–71, federal officials overheard 29,513 people in 442,137 conversations over 491 installations, as follows:
In 1968–71, state officials overheard 47,714 people in 676,755 conversations over 1,348 installations. The overwhleming bulk of this was in New York and New Jersey, with the major share in New York. To some extent, this is an estimate, but only to a very minor extent.
1. These figures differ somewhat from the ACLU 1971 Report because they are based on calculations derived from the reports of the individual installations and authorizations, rather than from the overall averages and summaries as in the 1971 Report.
2. Although the federal average was some 13.5 days per installations, a very high percentage of the New York installations lasted for many, many months.
3. Although the states, (which means largely New York and to a lesser extent, New Jersey) originally used the technique largely for drugs and nongambling offenses, by 1971, the states had shifted to overwhelming concentra, tion on gambling. Despite the pleas of need for serious offenses, the federal usage concentrated on gambling right from the start, and indeed increased that concentration at the expenses of drugs and others. The breakdowns are as fol. lows:
1 Includes attempts, threats, solicitations and conspiracy to commit homicide (including manslaughter) as well as a few occasional instances of consummated murder.
4. During the four year period, only 2 applications were denied, (in 1969); there is also some independent evidence of judge-shopping. For this and other reasons, there is reason to doubt that the court-ordered system is imposing meaningful controls, particularly on the state level.
b. National security.–Figures released by Senator Edward F. Kennedy indicate that at least the following number of national security wiretaps and bugs were installed : June 12, 1968 (50 taps and 6 bugs) 1969 (81 taps and 13 bugs).
94 1970 (97 taps and 16 bugs).
1. On the basis of classified information provided the Senator, his staff has calculated that the average national security installations lasted from 78.3–209.7 days, or about 6 to 15 times the court-ordered variety of 13.5.
During 1969–71, the latter caught an average of 56 perople and 900 conversations per installation lasting an average of 13.5 days. If the court-ordered averages are roughly comparable to the national-security type--and there is no reason to think otherwise-then the 100 national security surveillances annually may well have been intercepted: From 31,000 to 84,000 people per year; from 546,000 to 1,350,000 conversations per year.
2. Electronic surveillance for domestic security purposes allegedly accounted for very little of the national security total since the Department of Justice claims it found it necessary to shut off less than 10 devices as a result of the Supreme Court's decision banning such electronic surveillance without a warrant. Thus, this vast amount of surveillance is likely to continue.
3. This known surveillance may not be all there is, since it does not include electronic surveillance by the Army (which was revealed last summer) or possible surveillance by the CIA and other agencies, or interception of teletype messages.
4. According to Ramsey Clark, such surveillance rarely produces anything of value. He has testified that if all were shut off, "the impact on our national security ... would be absolutely zero." 2. Costs
a. Court-ordered surveillance.-The State figures are approximations, but are fairly close since the state cost reporting approached but was not equal to 100%.
b. National security.-Using the same 6–15:1 time ratio of national security: court ordered surveillance as earlier, and on the premise that most of the costs are manpower costs which vary with time, one can roughly estimate the costs of an average national security installation as about 6–15 times the $9,500 of the 1969-71 federal installations, or about $47,000-$142,500 per national security installation.
Since there have been about 100 national security installations each year, a rough estimate of the annual costs of national security surveillance is from $5.7 million to $14.3 million annually.
C. Comments.--1. These figures do not include the very substantial amount of judges' and lawyers' time necessary to prepare the court-ordered applications, and the lawyers' and FBI time for the national security authorizations by the Attorney General.
2. (a) There are inexplicable differences among the different types of offenses for which federal surveillance is used, particularly between drugs on the one hand, and gambling and the other offenses on the other. Moreover, there are startling fluctuations in costs from one year to the next for, e.g., the drug surveillance: from $61,825 per installation in 1969 to $26,035 in 1970 and $12,772 in 1971. The gambling average remains relatively stable and relatively low, whereas the “Other" category drops steadily from $9,212 in 1969 to $5,794 in 1971.
(b) The state figures vary and fluctuate so greatly, that no pattern is even discernible. Thus, the gambling average rises slightly, but drugs, and Other rises very steeply, while Homicide jumps up and down and kidnapping falls sharply. These fluctuations are so inexplicable that the only conclusion is that the reports are unreliable. 3. Results
Because it takes close to two years for cases to be disposed of, at least on the federal level, the figures for convictions are still incomplete, except perhaps for 1969 convictions and the 1969–70 arrests; the incriminatory conversations are likely to be fairly complete. So far, the results seem to be as follows:
1 Includes all types of homicide-related offenses, including attempts, threats, solicitation, conspiracy, manslaughter, etc. 3 May be the result of incomplete reporting
So far, only 194 out of the 1,448 installations have been directly associated with convictions; a small additional percentage may have been related thereto. For 1969, the year where most results are probably in, 55 out of 260 installations were associated with a conviction. As noted below, it is uncertain when and whether “associated with" involved a causal connection.
These arrests are probably fairly complete for 1968–1970.
Incriminating conversations, based on prosecutors reports
1968 1969 1970. 1971.
22 28 30 53
In 1969, 12 of the 30 installations were associated with a conviction : 9/20 of the gambling installations, 2/4 Drug, and 1 out of 5 Other.
The only year for which figures seem relatively complete is 1969. The overall reported cost for this surveillance-omitting the unreported lawyers' and judges' time costs—was $440,287 or about $3,500 per person convicted, and $37,000 for each of the 12 installations with which convictions were associated.
PERSONS ARRESTED, BY OFFENSE
For 1969 and 1970, the figures are probably close to complete. During this period, 92 out of 140 gambling installations were associated with an arrest, 26 out of 43 drug installations, 0 out of 1 kidnap installation, and 13 out of 26 Other, for a total of 131 out of 210 total.
The overall federal figures as reported are quite high : 82% in 1969, 70% in 1970 and 71% in 1971. But the non-drug and non-gambling installations produce very few incriminating conversations.
C. Comments.- 1. There is good reason to conclude that the electronic surveillance was not necessary to at least some of the convictions obtained ; indeed, in some cases, courts and prosecutors found or admitted as much.
2. The arrest figures are of little significance to this issue. Not only are arrests subject to manipulation, of which there is some evidence, but wiretaps and bugs may not be installed unless there already is probable cause to arrest at least someone, and the reported court cases confirm this.
3. Much of the federal wiretapping and bugging is on small-time gamblers; Justice Department figures and reported cases so indicate. Despite claims of high value and effectiveness or organized crime, law enforcement authorities have sharply reduced their usage for every offense but gambling since 1968.
4. At least in one case a court has found that the federal government's reports of "incriminating" conversations was overstated many times. In any event (1) the definition and application of "incriminatory'' is highly subjective, rarely testable, and often self-serving; (2) also, there is a good deal of evidence that very little effort to minimize non-"incriminatory” interceptions is made on the state level, and not too much more on the federal level.
5. Few of the major industrial states, which could be expected to "need” this authority, have adopted it. Indeed, most state (and much federal) wiretapping is in New York, and New Jersey, California, Illinois, Pennsylvania, Ohio and other states with major crime problems have not even bothered to give their police this authority; many states which have created such authority don't bother using it.
6. There is no indication that the heavy law-enforcement effort has substantially, reduced the drug or gambling problems, or that it ever can. Causality is particularly difficult to trade for the increase in electronic surveillance has been accompanied by a very heavy increase in men and money for all other ļaw enforcement operations.
B. SAME METHODOLOGICAL NOTES
The figures for court-authorized surveillance that are analyzed in this study are drawn from the Annual Reports issued by the Administrative Office of the United States Courts for the years 1968–71. These Reports contain data relating only to court-authorized surveillance; so-called national security and oneparty consent surveillance are not done pursuant to court orders and therefore are not included in the Annual Reports. Whatever information we have as to the national security surveillance and consent eavesdropping is from other sources.
The figures in the Annual Reports on judicially authorized surveillance are broken down into three sections:
(1) Summary tables, which contain both summary figures and averages; for some reason, totals for certain categories such as total persons or conversations overheard, are not published, perhaps because the averages are based on less than the total number involved since in several instances, the prosecutor or court omitted the relevant information. The Tables are identified by numbers.
(2) An Appendix containing individual date for each installation, relating to identity of judge and prosecutor, type of offense, type and place of surveillance, length of surveillance, number of persons and conversations overheard, costs and results; these appear in the Appendix in Tables A (judge reports) and B (prosecutor reports).
(3) Follow-ups on surveillances of prior years, set out both in summary form and individually (Table C) in the Appendix.
The 1971 ACLU report depended largely on the summary tables for its overall figures. The averages were multiplied by the total number of installations