I HAVE NOW completed 22 months as a federal prisoner. Just before Christmas I was released from the Federal Correctional Institution at Seagoville, Texas, to the Community Treatment Center in Dallas.

But it does not take a year and a half's exposure to the federal penal system to conclude it is a dismal failure. The figures speak for themselves. We keep prisoners locked up longer for comparable crimes than any other country in the civilized world. At the same time, our rate of repeat offenders, 70 percent, is the highest in the civilized world.

The single most striking impression the system has given me is the depth of vindictive bitterness it creates in the inmates. There are about 30,000 federal prisoners, and there is considerable turnover. Depending on how you calculate, we spend about $15,000 a year to keep a person locked up. But we do not turn out better citizens; we develop worse ones, vengeful ex-offenders with little respect for either the establishment or for law and order.

When I was first locked up, I quickly concluded that the fault for this dismal state of affairs lay with the Federal Bureau of Prisons. Most Americans despise the role of the jailer, and the bureau does not attract the best of the bureaucracy. So it was easy for me to believe the bureau was the prime malefactor and to accept the conventional wisdom that rehabilitation efforts were a waste of time. I was wrong.

The real reason for this bitterness is the U.S. Parole Commission. The parole system's bureaucracy, its regulations, policies and procedures, combine to foster attitudes that turn little criminals into big ones and today's convicts into tomorrow's repeat offenders.

There is a saying in prison: "If you can't handle the time, don't commit the crime." Prisoners, surprisingly, do not seriously object as a rule to serving time for the crimes they were convicted of. But the parole system punishes them by extending their imprisonment for crimes they have never been charged with and of which they have never been convicted. This happens over and over again, as a matter of deliberate Parole Commission policy and practice.

In other words, while the popular myth is that the parole system results in too much leniency for criminals, sending them back to the community too soon, the reality today is just the reverse. The case of Juan Gonzales

Suppose, for example, you were arrested for unlawfully carrying a gun. You pled guilty to this crime under a plea bargain. But when you went to your parole hearing, the decision was not to hold you for the gun violation, but to require you to serve an extra two years because the Parole Commission held you for trafficking in heroin -- even though the Parole Commission admitted you may well never have heard of the heroin transaction.

Would you be bitter? My young friend, Juan Jose Gonzales, was. This is precisely what happened to him.

Gonzales is in his mid-20s, married, and has young children. He is bright and full of energy. When I first knew him at Seagoville, he was a cheerful and effective participant in programs dealing with children.

In early 1978, Juan hired on as a bodyguard to a man who made a trip to collect a sum of money. When Gonzales and the man reached their destination, they were arrested. Gonzales, who was carrying a pistol, was charged with unlawful transportation of a firearm. The government offered a plea bargain and Gonzales accepted. He pled guilty to violating the Firearms Law and was sentenced under the Youth Act.

In theory, young men are to be released under the Youth Act as soon as they are "rehabilitated." In practice, however, the Parole Commissio makes no inquiry as to "rehabilitation." It releases prisoners only by inflexibly adhering to a chart of "guidelines" -- so much time must be served for whatever crime was committed.

The commission's guidelines for violating the Firearms Act call for serving 9 to 13 months. Gonzales had a clear record in prison and, under ordinary circumstances, would have been released at the bottom of these guidelines -- in about nine months.

But the commission decided to hold Gontales for 30 months -- nearly two years more than be should have been expected to serve. Why? Instead of using the firearms guidelines, the Parole Commission applied the heroin trafficking guidelines.

The man who hired Gonzales -- not Gonzales himself -- had been involved in a heroin transaction. Gonzales denied any knowledge of the man's business, and the perole penel that heard the case conceded there was no evidence that Gonzales had been involved in or even know of the heroin sale. The panel's summary states:

"In view of his background, his being a first offender, and perhaps not really knowing of the heroin transaction, the panel recommends a release at the bottom of guidelines [for trafficking in heroin]."

In effect, the Parole Commission violated the bargain made by the government prosecutor, held its own trial, judged Gonzales and sentenced him. Gonzales appealed to the regional office and to the national appeals board. He lost both appeals, no reasons given. He initiated court action; his case has not yet been put on the docket.

Just after Christmas, I learned that Gonzales had escaped and become a fugitive.I don't know what happened, but I can guess. I heard one rumor that he was gunned down by the federales in Mexico and another that he is in a Mexican prison. Arbitrary and illegal

The Parole Commission operates, in effect, as the hidden judiciary -- in a way that is arbitrary and utterly unlawful, a violation of the basic constitutional right of due process. It is also counterproductive -- sowing bitter reasons for some citizens to continue regarding the law as a fraud.

Most people, including lawyers and judges, think prisoners serve a third of their sentences and are then released for good behavior. They also believe the time prisoners serve is related to the crime of which a court has found them guilty.

These are misconceptions. The sentencing court does not control final punishment. The ultimate authority is the Parole Commission.

In 1976, Congress passed the Parole Commission and Reorganization Act to standardize the period served in federal prisons for various criminal offenses. Under this act, the commission adopted the "guidelines" against which its hearing panels were to set parole release dates. The problem was that for stealing a car, a criminal might be sentenced to five years in one district and to one year in another. These disparities have been reduced, though not eliminated, and that's good. But other injustices have been created.

The guidelines consist of a chart listing crimes and the period to be served, adjusted by a prisoner's "salient factor" score. This rating is designed to gauge a prisoner's prospect as a parole risk: "Very Good," "Good," "Fair," "Poor."

The practical effect is to vest in the commission the ultimate decision on the time a prisoner serves. The commission as a matter of practice ignores court recommendations. Prison officials dislike the system as well because it effectively eliminates any incentive for good behavior.

The Parole Commission goes beyond the limits of the guilty verdict and, instead, bases its guidelines on what it calls the inmate's "total offense behavior." Its instruction manual says: "In applying the guidelines, the offense serverity rating shall reflect the overall circumstances of the precent offense behavior... Information in the file describing offense circumstances more savere than reflected by the offense of conviction may be relied upon..."

This is a frightening statement. The commission is saying it will not confine itself to those matters charged against prisoners, or to the offenses of which they have been found guilty. Its parole panels use any "information" contained in the file to establish guidelines, and the Juan Gonzaleses are forced to endure the injustice that results.

This "offense behavior" is established at a two-person panel hearing that rarely lasts more than 20 minutes. If the prisoner doesn't like the outcome, he can appeal to the regional office that sent the panel. The last appeal goes to the national appeals board in Washington, an adjunct of the U.S. Parole Commission, which rums the whole system. The broken plea bargain

One of the rankling elements is that the prosecutor making the original deal with the offender represents the Justice Department. Then the parole system that ignores the plea bargain and makes its own decisions also represents the Justice Department.

James L. Smith, a young banker from Austin, Tex., for example, was indicted for 12 counts of falsifying bank records. He made a plea bargain with the prosecutor on one count, involving a loss to the bank of $227.98, the only crime of which he was convicted and one that would call for eight months to a year in prison. But the Parole Commission decided he must serve three to four years -- by holding him responsible for charges that were never proved against him.

The commission decided -- despite the plea bargain -- that Smith was guilty of property embezzlement exceeding $500,000. It arrived at that erroneous figure by adding in a civil claim against the young man, whose troubles included personal bankruptcy. How can the government keep a man in prison for a civil claim in which no criminal conduct was ever charged? The parole examiner said this distinction makes no difference because Smith was "responsible." Meanwhile, the civil suit was dropped.

In any case, the commission's arithmetic was faulty. A probation officer had erroneously added up the same $50,000 mentioned in three counts and thus tripled the bank's alleged loss. Because of the mistake, Smith faced the loss of an additional 10 to 16 months of his life.

Smith's case is not unusual. I helped prepare hundreds of cases with similar complaints about parole decisions. These men are regularly punished for crimes of which they were never tried or convicted. Probation reports, including hearsay evidence or false information, frequently add months or even years to a man's time in prison.

Charles W. Washington was charged with misprison -- having knowledge of a felony. In this case, it was knowledge of the theft of drill pipe. But the was forced to serve an extra 1 1/2 years because the Parole Commission held him for actually stoaling the pipe -- a crime the prosecutor never charged him with.

To make it worse, the commission relied on a pre-sentencing report that valued the stolen pipe at more than $100,000 when the prevailing market price indicates it actually was worth about $50,000.

The prison staff at Seagoville recommended parole for Washington six months before his hearing, but the recommendation was ignored. We appealed and demanded an explanation, but we never got any.

I served with Washington on the Warden's Council at Seagoville for many months. He is one of the kindest men I have ever known. His case was a long string of bad news, but his only comment was: "We done our best. Them mothers will never do me right." Misconstrued cases

One of the most damaging legal jumps the Parole Commission frequently makes is to misconstrue the meaning of conspiracy cases -- assigning to one prisoner offenses he didn't even know about, much less commit.

Albert Alfaro was indicated in a drug case as a "mule," a delivery man but not a principal organizer. The indictment said he personally handled about 30 pounds of marijuana. Others in the conspiracy handled a total of 175 pounds. The Parole Commission chose another figure -- four to six tons -- that came from a pre-sentencing report. Alfaro was given 45 months, roughly three years longer than he had a right to expect.

To make it worse, the commission included nine months because it said Alfaro committed "multiple separate offenses." This amounts to double jeopardy -- adding up the overt acts contained in a conspiracy indictment and counting them again as new crimes. These "separate offenses" are actually part of the same, single, overall crime of conspiracy. In effect, prisoners like Alfaro are sentenced twice for the same offense.

The Parole Commission is not a prisoner of consistency. Arthur Perez was charged as a "mule" in the same conspiracy and handled a larger amount of marijuana than Alfaro -- but Perez got a year less in prison, no reason given. His explanation: "These are some kind of crazy gringos."

Meanwhile, Oscar Trevino, another defendant in the same case with a virtually identical role, filed an appeal, making the same arguments made unsuccessfully by Alfero. Only Trevino got his sentence reduced.

One of the most notorious inconsistencies occurs when the commission makes changes in its sentencing guidelines. When it raises the length of time to be served for a particular crime, it applies the new guidelines, even though the prisoner was convicted before they were in force. When the commission lowers a guideline, it uses the old one. This is ex post facto justice, as any law student will recognize, but the Parole Commission's hearing officers are not lawyers.

My friend, Chad Corbett, convicted on a firearms case, applied for a reduction after the commission lowered its guidelines for that offense. Reduction denied. The difference is two years. Yet Jim Smith was "rated" under the new guidelines for property offenses -- guidelines raised twice after his sentencing -- which meant additional time in prison for him.

I can go on citing outrageous examples, almost indefinitely, because the errors and distorted interpretations are the pattern in hundreds of cases. I hear frequently from inmates at Seagoville and other federal prisons who have the same grievances and find it difficult to get knowledgeable legal help. The extra time crowd

Chris Bullard got an extra year on a drug charge because a pre-sentencing report said the conspiracy "was capable of distributing methamphetamines in 10-pound quantities or more." The report didn't claim this actually occurred, only that the defendants were capable of doing it. We finally won that one on appeal -- but not before Bullard served a substantial portion of extra time.

Frank Ward, a middle-aged Kansas farmer, was convicted on a bankruptcy fraud involving $28,000. The Parole Commission held him responsible for two additional charges in which he was not involved, raising the total to $110,000, even though the appeals court had already thrown out one of them. The difference was 10 to 16 months in prison.

Ken Crumpler, a Texas produce broker, was convicted of possessing illegal firearms -- a detonating cap, a blast simulator and a length of primer cord. The Parole Commission counted each of these as separate weapons -- though none would function alone as a weapon -- and gave Crumpler an extra four months.

Ben Garza will serve an extra 1 1/2 years on a drug charge because the commission holds him responsible for "large" amounts of hard drugs and "automatic weapons" mentioned in a pre-sentencing report that covers other defendants. Garza was never linked in court or in the report with any of those items.

Jeffrey Tuttle was given an extra six months because of an error in the parole files alleging that his offense was committed while he was on probation. The dates were wrong.

Stephen Land was a prison manipulator (I have to say I was grateful for the steak and fresh fruit he somehow obtained) who was convicted of driving a stolen car across state lines. Land was sentenced to two years, but the prosecutor recommended immediate perold for a number of reasons. Land had been coerced by threats made against his brother, who was held hostage. He had received to money, only the safety of his brother. He had testified against the principal in the case, who was convicted, but placed on probation. Furthermore, be was raped in jail after his arrest and his struggle with his six assailants cost him the sight in one eye.

But because Land had a prior conviction and because of a mistake in calculating his "salient factor" score, the commission held him to the maximum sentence. Some reduced times

In fairness to the parole system, after a lot of correspondence and appeals, Chairman Cecil C. McCall made an encouraging reply to the batch of cases I sent him. He promised relief if our claims held up. He has been as good as his word. Times to be served have been reduced -- not everything I felt the men deserved, but at least something.

But the parole guidelines and procedures that produce so many of these injustices are still in place, and until Congress or the Supreme Court do something, the same pattern of injustices will persist.

The new conventional wisdom is that rehabilitation doesn't work for these men and that criminals should be kept off the streets as long as possible. That is the only consistent motivation, I think, for these parole procedures -- stretching out sentences for offenders in any way possible. They're afraid -- afraid of criticism if something goes wrong on a parole.

A civilized society that believes in the law does not allow a qussi-judicial system to mete out punishment in violation of constitutional principles of fairness.

People who are not moved by the argument of equal justice should consider this one: Sooner or later, the nation's defendants and defense lawyers are going to realize that plea bargains in federal courts become broken promises once the defendant is before the Parole Commission. Since more than three-fourths of federal cases are settled by guilty pleas, this could really clog the U.S. courts. If defendants have no incentive for a plea bargain, they might as well go to trial and take their chances.

I have never heard a prisoner complain that the guidelines set for any partcular crime are too high. They know society must do something to punish them. But prisoners also have an acute sense of what is right and wrong. They know full well that the rules of the parole game are utterly unfair. CAPTION: Picture 1, L. A. Nikoloric, pictured here with his wife Mary, is a corporate lawyer who spent 22 months in federal prison for wire fraud involving an overseas mutual fund. A Yale Law School graduate and Phi Beta Kappa alumnus of Princeton, Nikoloric helped other inmates with their parole appeals and this is his report on how that system works.; Picture 2, A couple of nights before he left Seagoville Prison, the author writes, some of the inmates had a goodbye party at which a few Poloroid photos were taken. In this one, he is standing fourth from left.