Plea Bargaining's Triumph

Beginning of article

The Yale Law Journal Volume 109, Number 5, March 2000

Though originally an interloper in a system of justice mediated by courtroom battles, plea bargaining now dominates American criminal justice. This Article traces the path of plea bargaining's progress. The forum is a single Massachusetts county in the nineteenth century, but the lessons of the work extend to America as a whole.

The path of plea bargaining's rise was in great part a function of the powers and interests of individual courtroom actors. Although criminal defendants play a distinct part in this story, the most important actors prove to be prosecutors and judges. In the early decades of the nineteenth century, plea bargaining was the work of prosecutors, who found natural incentives in the quick and easy victories it gave them. But because judges--and not prosecutors--held most of the sentencing power and therefore most of the plea-bargaining power, plea bargaining could spread no further than those few cases in which prosecutors happened to hold the balance of sentencing power. Only in murder cases and in cases arising under the Massachusetts liquor-licensing law did prosecutors have substantial sentencing power, and so plea bargaining remained largely confined to these narrow categories.

Plea bargaining could not spread across the entire criminal spectrum until judges became partners in the cause. Through the first three-quarters of the nineteenth century, judges showed far less enthusiasm for plea bargaining than did prosecutors. After all, judges lacked prosecutors' natural incentives to plea bargain. Plea bargaining did not assure them "victory," and they did not suffer under the workload pressures prosecutors faced. Moreover, judges had principled objections to dickering for justice and prideful objections to sharing sentencing authority with prosecutors in the course of plea bargaining. Only in the last quarter of the century did these various obstacles to judicial participation begin to give way. Among the most important catalysts of change was the revolution in tort law brought on by the mechanization of manufacture and transportation. A flood of complex civil cases persuaded judges of the wisdom of efficient settlement of their criminal cases.

Claiming now the patronage of the two principal power-brokers in the criminal courtroom, plea bargaining quickly grew to be the dominant institution of American criminal procedure. Not only did it become the primary mode of settling almost all criminal cases, but it also became the principal arbiter of the success of other new institutions of criminal procedure. Plea bargaining's role in fostering the rise of probation was quite direct: The procedural formula that later took the name "probation" was popularized in the first half of the nineteenth century by Massachusetts prosecutors who sought a plea-bargaining technique they. could employ without judicial cooperation. The rise of public defenders' offices owes a similarly direct debt to plea bargaining: Early proponents of the public defender attempted to win support within legislatures and the established bar by arguing that the new defenders would be ready participants in plea bargaining--a strategy that became prophecy. Conversely, the interests of plea bargaining helped to defeat the indeterminate-sentencing movement. A true indeterminate-sentencing scheme, in which judges had no power to dictate sentences, would have brought plea bargaining nearly to a halt. So those judges and lawyers who valued plea bargaining made sure that such a scheme never came about, and they worked to frustrate any sentencing innovation that advanced the indeterminate ideal.

In the end, the most vivid testimony to the dominance of plea bargaining within American criminal procedure is this simple fact: It is difficult to name a single procedural innovation of the last 150 years that has been incompatible with plea bargaining's progress and has survived.

There is no glory in plea bargaining. In place of a noble clash for truth, plea bargaining gives us a skulking truce. Opposing lawyers shrink from battle, and the jury's empty box signals the system's disappointment. But though its victory merits no fanfare, plea bargaining has triumphed. Bloodlessly and clandestinely, it has swept across the penal landscape and driven our vanquished jury into small pockets of resistance. Plea bargaining may be, as some chroniclers claim, the invading barbarian. But it has won all the same.

The battle has been lost for some time. It was lost at least by the time prominent observers of the 1920s first lamented our "vanishing jury," and in some places it was lost decades before then. In the years since the jury succumbed in a war few knew it was waging, advancing waves of who-lost-the-jury scholarship have tried to retrace the path of defeat. Much of this work has charged that the ever-weightier burden of modern jury trials overbore the system, which surrendered to the plea bargain's efficiency to avert collapse. Another, newer collection of studies looks beyond the courtroom to spy out broader social forces that aided plea bargaining's cause. But in the breadth of its scope, this work overlooks the first principle of conflict: that victory goes to the powerful. And so while we can find many worthy accounts of why the jury fell, we must ask as well why plea bargaining triumphed.

Like most of history's victors, plea bargaining won in great part because it served the interests of the powerful. In the battlefield of the criminal courts, the kind of power that mattered most was the authority to dictate sentences, which judges held in great measure and prosecutors in a compromised but still substantial way. To track the course of plea bargaining's rise, we must discover how prosecutors, who had an almost inherent interest in plea bargaining, secured the power to make it happen, and why judges, who inherently had the power to make it happen, began to see it as in their interest. In this account of plea bargaining's rise, legislators will play a large role, for their power to allocate sentencing authority between prosecutor and judge tilted the terms of battle. And criminal defendants, who held a nominally absolute power to plead or not to plead but who often found themselves hopelessly undefended, must play a real if complicated part.

This is not, then, a "social history" of plea bargaining's rise, if by that we mean a story that highlights the play of social forces and minimizes the role of human actors. Although recent scholars have put forth appealing accounts of the electoral pressure of new immigrants or the analogical force of new industry in the creation of "assembly-line justice," these theories either stand without evidence or fall beside the point. Plea bargaining's triumph was manifestly the work of those courtroom actors who stood to gain from it. To the extent that external actors, especially legislators, played a conscious role in the struggle, they mainly took sides with the ultimately defeated jury trial. And to the extent that broader social forces, such as immigration and industrialization, influenced the outcome of the conflict, they did so by increasing the criminal caseload on the one hand and the civil caseload on the other, thereby altering the interests and options of prosecutors and judges.

And yet in one striking way, the story of plea bargaining's rise is bigger than the actors who made it happen. Plea bargaining, once it took hold, possessed a power of its own. That power derived ultimately from the individual power of those whose interests plea bargaining served, but in its collective form that power made plea bargaining a dominant force in the evolution of modern American criminal procedure. Sometimes, plea bargaining's influence appears on the face of the historical record. This is most tree in the case of probation, one of the most enduring nineteenth-century contributions to our penal mechanisms. I will argue that the birth of probation was in some part, and perhaps in large part, the work of prosecutors who sought a new way to expand their power to bargain for pleas. Once in place, probation grew in symbiosis with plea bargaining and became in time one of the most useful tools of lawyers cutting deals.

Plea bargaining's role in shaping other procedural institutions appears more subtly in the historical record. But the unmistakable correlation between those procedural changes that have survived and thrived and those that aided plea bargaining's cause compels a conclusion that plea bargaining has so fast a grip on our institutions of justice that antagonistic institutions cannot survive. The demise of the indeterminate sentence, one of the most promising of the late nineteenth century's progressive brainchildren, bears the mark of plea bargaining's malice. A true indeterminate sentence, in which only prison officials and parole boards could set the length of criminal sentences, would have stripped both judges and prosecutors of the power to bargain over the length of terms and would have hobbled the plea-bargaining regime. But though widely promoted, tree indeterminate sentences never emerged, and even the half-measure that we have come to know as parole found its development stunted when it threatened plea bargaining's dominance.

Probation's rise and the indeterminate sentence's fall are but two of plea bargaining's victories. Before canvassing the rest, we need to follow the course of plea bargaining's ascent to learn the source of its strength. I will tell this story as it unfolded in America, for although the earliest instances of plea bargaining may well have happened elsewhere,(1) and although plea bargaining in time would spread across the common-law world and beyond, it triumphed here first. Within America, I will focus on Massachusetts. Massachusetts is the birthplace of probation, and though I had little suspicion when I set out l hat plea bargaining and probation would prove so closely linked, it seemed likely that innovations would find a common home. Moreover, Massachusetts was among those states that acted most vigorously in the nineteenth century to suppress the sale of alcohol, and there was reason to suspect that enforcement of liquor laws played a part in plea bargaining's early rise.(2)

Within Massachusetts, I will focus mainly on its largest county--Middlesex(3) --which spreads south and east from the New Hampshire border to Boston's western edge, taking in the pioneering mill town of Lowell, the hamlets of Lexington and Concord, and cosmopolitan Cambridge. Middlesex's diversity avoids the possible distortions of Boston's idiosyncrasies, which particularly disrupted that city's enforcement of the liquor laws.(4) Moreover, Middlesex was the practicing ground of Asahel Huntington, perhaps the first prosecutor in history to be called up on charges of plea bargaining. And Middlesex is where I practiced as a prosecutor, an experience that left me familiar with the ways of its courts and perhaps more aware of the ways in which things have changed.

Within Middlesex County, I will look most closely at the middle tier of the county's judicial system, which had jurisdiction over all but the most serious crimes. A succession of courts occupied this middle tier during the nineteenth century--the Court of General Sessions of the Peace until 1804, the Court of Common Pleas from 1804 until 1859, and the Superior Court after 1859.(5) At the four annual sittings of the Court of General Sessions of the Peace, local justices of the peace presided jointly over jury trials.(6) At both the Court of Common Pleas and the Superior Court, jury trials took place before professional judges who rode circuit throughout the state, sitting in each county three or four times a year and ensuring some statewide uniformity of practice.(7) These middle-tier courts promise to be better sites for this study than the courts of the lower tier. The bottom tier of the state's judiciary(8) lacked two institutions--jury and prosecutor--whose role in the criminal system looms far too large for us to think that much can be explained without them.

My examination of the upper tier, occupied by the Supreme Judicial Court, will not range as far because the criminal trial jurisdiction of that court was quite narrow and grew narrower as the nineteenth century wore on. Like judges of the middle tier, justices of the high court rode circuit across the state to preside over criminal jury trials.(9) At the century's start, their jurisdiction included an array of serious crimes, but soon shrank to include only capital crimes, a category that, in turn, dwindled by mid-century to include only murder. In 1891, the Supreme Judicial Court lost even murder trials to the courts of the middle tier.(10)

Rarely handling more than a handful of criminal trials a year, the Supreme Judicial Court would not serve well as a focus for a study of plea bargaining's rise. Still, it is important to include its business in this study, both to assure that the practices I observed in the middle tier were not somehow peculiar to less serious criminal cases and to take advantage of the sometimes superior records of Supreme Judicial Court business and of the deeper insights they afford. As we will see, the rise of plea bargaining before the Supreme Judicial Court tracked roughly the same course as in the courts of the middle tier. Though it began perhaps a little later, plea bargaining in capital cases soon caught up with and at times overtook the development of plea bargaining in the courts below. More importantly, we will see in both tiers the same dynamic at work: The powers and interests of individual courtroom actors appear to explain the course of plea bargaining' s rise.

The records of the middle-tier courts for the years before 1859 have survived in almost full form, and those for the years since have suffered only minor losses and deletions. I have found no important gap in the records of the Supreme Judicial Court.(11) Complete as they may be, these records are hardly expressive of what went on in the courts of nineteenth-century Massachusetts and are maddeningly silent about the motivations behind most guilty pleas. The clerk's vacant formulas usually erased the tracks of any express agreement by prosecutor or judge to grant the defendant a concession in exchange for a plea. In most cases, therefore, we must use other methods to distinguish those defendants who pled guilty out of remorse for their crimes or to shorten their engagements in court from those who did so as part of a plea bargain, expecting to win some concession in exchange. We must employ still other devices to distinguish explicit plea bargains, in which the parties spoke out loud the terms of their exchange, from implicit bargains, in which the defendant pled guilty for an unpromised but hoped-for reward. My sampling of some 4000 cases(12) often permits statistical conclusions about the frequency and nature of plea bargaining even when individual case reports are mute. And newspapers and other contemporary sources help decipher the clerk's cryptic codes. These sources also make plain that the lessons of Middlesex County apply generally throughout Massachusetts and, with allowances for statutory and regional variances, in several other jurisdictions as well.

This story of plea bargaining's rise begins in the opening decade of the nineteenth century. One could begin earlier, as there is solid evidence of plea bargaining in earlier centuries, but that evidence is of isolated, episodic plea bargaining, whereas this is the story of plea bargaining's triumph as a systemic regime. Told chronologically, the story divides fairly neatly into two parts. During the first three-quarters or so of the nineteenth century, plea bargaining in Massachusetts advanced mainly in the realm of liquor-law prosecutions and murder cases, where for reasons we will explore, prosecutors had the power to negotiate pleas without any participation by the judge. These early deals took the form of charge bargaining--that is, in exchange for the defendant's plea to one or more of several charges, the prosecutor dropped the others or (in the case of murder) reduced the charge to a lesser offense. In the last quarter of the century, as judges converted to the cause, plea bargaining most often took the form of sentence bargaining, in which the defendant' s plea won a reduced sentence. Backed by judges as well as prosecutors, plea bargaining now broke the narrow hold of liquor and murder prosecutions and conquered the whole penal territory--so that by century's close, guilty pleas accounted for some eighty-seven percent of criminal adjudications in Middlesex County.(13)

As useful as this simple chronology may be, I at times will stray from the rather clean story line it provides. My aim is to highlight how the changing powers and interests of each courtroom actor played dominant roles in shaping the course of change. These two ways of telling the story are not always at odds: I will begin with the role of prosecutors, whose influence emerged most clearly in the liquor-law prosecutions of the century's first half, and I will conclude with judges, whose changing behavior in the last quarter of the century assured plea bargaining's triumph. At times, though, the focus on individual actors will require us to leap ahead or to fall back.

So I will begin with prosecutors. That prosecutors take up three parts of this eight-part study is not because their power to bargain was greatest of all. In fact, the judge held the largest share of the power to dictate sentences and therefore to plea bargain, while the prosecutor's power was almost always subordinate to his. I linger on the prosecutor's role precisely because the prosecutor managed to elude the judge's control only rarely--and then only by virtue of extraordinary legislative grant or procedural ingenuity. Part I will address the first of these extraordinary legislative grants, made in the context of liquor-law prosecutions and, with some differences, in capital cases. My research in Middlesex County confirms earlier findings of a strikingly high rate of plea bargaining in Massachusetts liquor-law prosecutions in the early nineteenth century. Various good theories might explain a link between liquor prosecutions and plea bargaining, but the evidence overwhelmingly points to one--that the distinctive penalty scheme that the legislature created for the liquor laws, which assigned a fixed fine to almost every offense, deprived the judge of almost all sentencing discretion and put the prosecutor in a position to manipulate sentences by manipulating charges. Similarly, in capital cases, the prosecutor had the power to spare defendants mandatory death by permitting them to plead guilty to a lesser charge. Prosecutors quickly exploited these narrow grants of sentencing authority and put in place a very modern practice of charge bargaining for pleas.

Explaining the source of prosecutorial power to plea bargain is the hard part--it is easy to see why prosecutors wanted to plea bargain. Prosecutors of the nineteenth century, like prosecutors today, plea bargained to ease their crushing workloads, made heavier in the nineteenth century both by their part-time status and utter lack of staff and by a caseload explosion perhaps set off by newly founded police forces and massive immigration. And of course they plea bargained to avoid the risk that wanton juries would spurn their painstakingly assembled cases. Given such clear incentives, the task of explaining the rise of prosecutorial plea bargaining in liquor and capital cases should have been complete at the end of Part I--for as soon as prosecutors had the power to plea bargain, they surely would have used it. Still, I will need to consider in Part II the arguments of those scholars of plea bargaining who have dismissed the importance of caseload pressure in explaining the rise of plea bargaining and have put forward other explanations for the early emergence of plea bargaining. And I will need to consider why judges were not partners in this early rise of plea bargaining. For if they had been, plea bargaining would not have been limited to liquor and murder cases--in which prosecutors had the power to plea bargain on their own--but could have extended across the criminal docket.

In Part III, I will step briefly out of the courthouse and into the State House. Massachusetts legislators reacted sourly when they discovered how prosecutors were using the power unwittingly bestowed on them by the liquor law's rigid penalty scheme. At mid-century the legislature eliminated this power and very nearly succeeded in snuffing out prosecutorial plea bargaining in liquor cases. The legislature did not, however, disturb the prosecutor's power to conduct charge bargaining in murder cases, for there had been relatively few such bargains by mid-century. The result was that during the third quarter of the nineteenth century, plea bargaining advanced more dramatically in murder cases than in any other category. Even in liquor cases, the legislature's efforts to eradicate prosecutorial charge bargaining failed. After losing formal power to manipulate sentences in liquor cases, prosecutors retreated to the more covert and informal tactic of placing these cases "on file." This procedural maneuver, often done in exchange for a defendant's guilty plea, allowed prosecutors to elude altogether the legislature's sentencing provisions. The primitive device of on-file plea bargaining evolved directly into what we know today as probation. By the end of the century, probation had become one of plea bargaining's most dependable foot soldiers.

Turning from the prosecutor's role in early plea bargaining to that of other actors, I will move on in Part IV to consider the part played by defendants. It is not hard to see why defendants, given the chance, would plead guilty for a measure of leniency, but it is far less clear why their behavior on this score might change over time. Middlesex court records disclose that during the first half of the nineteenth century, decades before plea bargaining began its dramatic ascent, there had been a long decline in the proportion of non-liquor cases that ended in a plea. I will argue that these early guilty pleas were not plea bargains made in exchange for leniency, but rather the hopeless gestures of unrepresented defendants who properly saw that they had little chance of winning if they went to trial on their own. The gradual increase in the number of defendants who chose trial during the first part of the century therefore may mean that more and more defendants had counsel. Then, in the third quarter of the century, a sudden assault on the power of defendants to take their cases to trial may have reversed this course and helped to speed plea bargaining's rise. Laws passed in Massachusetts and elsewhere that gave defendants the right to testify at trial had the probably unintended effect of discouraging defendants with criminal pasts from going to trial. Seasoned criminals knew that if they took the stand to claim their innocence, the prosecutor could impeach their testimony with their old convictions and thereby destroy any real chance of acquittal. Yet if they failed to testify, defendants believed, juries would convict them for their silence. Together with the growing practice of probation, defendant-testimony laws confronted every defendant with a good reason to plea bargain. Defendant-testimony laws helped to persuade accomplished criminals to plead guilty, while the promise of probation, which was available almost exclusively to first offenders who pled guilty, served as an incentive for everyone else.

The combined willingness of all prosecutors and many defendants to bargain for pleas was not, however, enough for the practice to thrive outside the narrow context of liquor laws and murder cases. The statutory penalty structure for most crimes gave Massachusetts judges such great discretion in sentencing that the prosecutor typically could not unilaterally guarantee a low enough sentence to win the defendant's plea. Plea bargaining's sweeping triumph during the last quarter of the nineteenth century suggests, therefore, that judges had entered plea bargaining's ranks. I will argue in Part V that a caseload explosion on the civil side of Massachusetts courthouses helped force this change of judicial heart on the criminal side, for in Massachusetts, as in most American jurisdictions, the same judges sat on both civil and criminal cases. The industrial boom of the last part of the nineteenth century--and especially the spread of railroads and street cars --spawned a whole new strain of personal injury litigation that, case for case, absorbed far more time than the contractual nonpayment cases that once had filled the civil dockets. The figures in Massachusetts are clear: As judges devoted a hugely increasing proportion of their time to the civil caseload, they devoted a shrinking proportion to the criminal caseload, and they resolved more and more criminal cases by guilty plea. Judges apparently discovered that they had more power to spur pleas in criminal cases than to coerce settlements in civil cases. After all, a criminal court judge could credibly promise a reward in exchange for a plea or threaten a penalty for going to trial, but in civil court, the jury--not the judge --generally set the loser's penalty.

By century's end, all three of the courtroom's major actors --prosecutor, defendant, and judge--had found reasons to favor the plea-bargaining system. For prosecutor and judge, who together held most of the power that mattered, the spread of plea bargaining did not merely deliver marvelously efficient relief from a suffocating workload. It also spared the prosecutor the risk of loss and the judge the risk of reversal, and thereby protected the professional reputations of each. In fact, by erasing the possibility of either factual or legal error in the proceedings, plea bargains protected the reputation and hence the legitimacy of the system as a whole. In Part VI, I will ask whether these variables of power and interest, which emerge from the records of Middlesex County and of the Supreme Judicial Court, can help explain the course of plea bargaining's rise in other jurisdictions. In England, New York, and California, the evidence suggests that they can.

I will move on in Part VII to explore the power that plea bargaining as an institution has amassed by serving the interests of power so well. The power of the various actors who stood to gain from plea bargaining became, in a sense, plea bargaining's power. This collective, systemic interest in plea bargaining encouraged the rise of those institutions of criminal procedure that helped plea bargaining and hindered those that hurt it. In the nineteenth century, plea bargaining fostered probation's rise and thereby created a hugely versatile plea-bargaining tool. In the late nineteenth and early twentieth centuries, plea bargaining helped stave off the indeterminate sentence, which had threatened to halt plea bargaining's progress. And in the twentieth century, plea bargaining played a surprisingly direct role in assisting the creation of public defenders. In turn, these organizations for defense of the poor assured that in a majority of criminal cases, the defense lawyer would share the prosecutor's and judge's interests in maximizing systemic efficiency--and hence in plea bargaining. These examples of plea bargaining's influence over other institutions of criminal procedure are merely case studies within a larger trend. In fact, it is hard to think of a single enduring development in criminal procedure in the last 150 years that has not aided plea bargaining's cause.

Finally, in Part VIII, I will examine how the power to plea bargain evolved in the late twentieth century. Before the advent of modern sentencing guidelines, both prosecutor and judge held some power to plea bargain without the other's cooperation. The result of their mutually independent bargaining strength was a certain balance of power, which to some degree protected defendants from abuses of power by either official. Today, however, sentencing guidelines have recast whole chunks of the criminal code in the mold of the old Massachusetts liquor laws. In the process, they have unsettled the balance of bargaining power by ensuring that the prosecutor, who always had the strongest interest in plea bargaining, now has the unilateral power to deal.

And so in time we will arrive again at the Massachusetts liquor laws, where we now begin.

I. LIQUOR LAWS, MURDER CASES, AND THE PROSECUTOR'S CHARGING POWER

A. Charge Bargaining in Liquor Cases

In 1807, Samuel Dana took office as the first Middlesex County attorney.(14) His main forum was the Court of Common Pleas, which occupied the middle tier of the Massachusetts criminal judicial enterprise. Dana did not launch his young office with honor. By 1809, he was before the legislature explaining why he had complained to three of the court's judges about their "unbecoming impetuosity, and snarling manner."(15) He had to explain as well why he had billed the court for services to crime victims that seemed to fall within his official duties(16)--and why he had represented a crime victim in a civil suit even as he prosecuted the case's criminal counterpart.(17) Against these well-founded charges, Dana mounted a pathetic and lawyerly defense.(18) By the end of 1809, he apparently was out of office, and the same "snarling" judges had appointed another in his place.(19) Dana won the last hand, though, for when the legislature reorganized the Court of Common Pleas in 1811, those judges lost their seats, and Dana was named the new court's chief.(20)

For our purposes, Dana's most important legacy lies not in his spat with the judges, but in another aspect of the December 1808 sitting of the Court of Common Pleas--the term of court that gave rise to the controversies just noted.(21) We find in the records of that term the case of Josiah Stevens of Tyngsborough, who faced prosecution under the state's 1787 liquor-license law requiring that alcohol retailers be licensed.(22) The Stevens prosecution was one of Dana's first liquor cases and only the third liquor case to come before the Court of Common Pleas in Middlesex County after that court had begun to hear criminal matters in 1804.(23) Dana had drawn up a four-count indictment against Stevens: Count one charged him with being a "common seller" of alcohol, counts two and three with making particular unlicensed sales, and count four with selling alcohol and permitting the buyer to drink on Stevens's premises. The court's clerk narrated the outcome:

   [T]he said Josiah [Stevens] says he will not contend with the Commonwealth.
   And Samuel Dana Esquire Atty. for the Commonwealth in this behalf says that
   in consequence of the defts. plea aforesaid he will not prosecute the first
   third and fourth counts against him any further.(24)

Dana and Stevens had struck a deal: In exchange for the defendant's plea of no contest, Dana dropped three of the four counts of the indictment. On the remaining count, Stevens paid a fine of $6.67 and $47.12 for the costs of prosecution.(25)

Dana prosecuted only one additional liquor case during the rest of his brief tenure as county attorney. In March 1809, he brought a four-count indictment against Nathan Corey, a husbandman of Stow: Count one charged Corey with being a common seller of alcohol, and counts two, three, and four with making particular unlicensed sales. The clerk's account of the result varies from that in Stevens only in that Dana lived up to his part of the bargain first:

   Samuel Dana Esquire Attorney for the Commonwealth in this behalf says that

   he will no further prosecute the said Nathan Corey upon the first second
   and fourth counts in said indictment and thereupon the said Nathan Corey by
   leave of Court says he will not contend with the Government as to the third
   count in said indictment.(26)

Corey paid a fine of seven dollars Together with $19.73 in costs.(27)

These are by no means the earliest guilty pleas I discovered in my search of the Middlesex court records. In fact, a surprisingly high percentage of cases in the earliest years of my study ended in pleas of guilty or no contest. In the Court of General Sessions of the Peace, which heard criminal cases before criminal jurisdiction passed to the Court of Common Pleas in 1804,(28) seventy-three percent of adjudicated cases in 1789-1790 ended in a plea,(29) and sixty-six percent of those in 1799-1800.(30) In 1809, when Nathan Corey stood before the Court of Common Pleas, fifty-eight percent of adjudicated cases ended by plea.(31) These were not merely liquor cases, but included both felony and misdemeanor cases from across the broad jurisdiction of the middle-tier courts. But Stevens and Corey are two of only three cases in these years that appear on the face of the records as clear plea bargains.(32) That is, these are the only cases in which the clerk's account discloses a concession made in exchange for the defendant's plea. Throughout this study, I will use the term "clear plea bargain" to refer to those cases in which the record makes it clear that the defendant won a concession by offering a plea. By contrast, I will use the expression "guilty plea" to refer to cases in which the defendant pled guilty but the record reveals no compensating concession. Sometimes we can look behind the silent record of a simple "plea" to discover the terms of a "bargain," but this is a matter we will take up in time.

Although Stevens and Corey are among the earliest clear plea bargains to emerge in my survey, they are of a remarkably sophisticated cut. We cannot know Dana's thoughts as he sought these four-count indictments, but today we would say that a prosecutor who brought several charges in the hope of gaining leverage in plea negotiations had "over-charged" the case. Dana may have had other motives for charging these cases as he did,(33) but given the four-count indictments, he was in a position to threaten the defendants with multiple penalties and then to reward their pleas of no contest by dropping three of the four counts. Note that Stevens and Corey chose not to plead plainly "guilty" to the remaining count of the indictment. Rather, they would "not contend with the Commonwealth," and by means of these pleas of nolo contendere, they spared themselves any admission of fault while giving the court the power to convict and sentence them.(34) To winnow the excess charges, Dana employed the nolle prosequi, or nol pros, which the Supreme Judicial Court only recently had declared to be an exclusively prosecutorial device.(35) Gadget for gadget--multiple-count indictments, nolo pleas, and nol prosses--these early plea bargains rival some of the best work of modern plea practitioners. And yet they emerged here, only a year into the tenure of the first county attorney, an almost instinctual urge of the prosecutorial soul.

Sophisticated as these pleas may have been, it is not immediately obvious why they worked. Why should Stevens and Corey have pled guilty, even to a single charge, if the presiding judges had the power to punish them severely for the one remaining violation? True, one's exposure is always greater having been convicted on four counts than on one, but many defendants might prefer to keep their right to seek an acquittal from a jury rather than to trade it for an uncertain reward. They might even flatter themselves with the hope that if they went to trial and lost, the judges still would not punish them severely. And conversely, why should Dana have dropped three charges and risked the possibility that the judges would, on the one remaining count, impose an overly lenient sentence? Had the judges been partners in these plea bargains and had they made their sentencing intentions clear, then any such reservation on Dana's or the defendants' part would have dissolved. But the record discloses no mason to think the judges took part in the deal.

In fact, neither Dana nor the defendants needed to worry that the judges might disappoint their expectations in the bargain, because the judges had no power to do so. The liquor law's penalty structure simply left them too little discretion. On a conviction of being a "common seller," as charged in count one of both indictments, the law prescribed a fine of twenty pounds exactly.(36) Twenty pounds was a good deal of money, amounting at the time to almost sixty-seven dollars,(37) or more than the total in fines and costs paid by either defendant. The common-seller count therefore gave Dana real strength in any plea negotiation, for if the defendant risked trial and lost on that count, the court would have had no choice but to impose the statutory fine--whereas once Dana nol prossed the count in exchange for a plea, the court had no power to impose it. Although the license law prescribed a specific fine for almost every offense,(38) it did grant judges some sentencing discretion in cases of single unlicensed sales by providing for a fine of between two and six pounds.(39) Despite the uncertainty injected into the outcome by this sentencing range, Dana and the defendants were able, if they chose, to agree in advance on the exact penalty the defendants would pay. For another statute gave Dana the power to determine the costs of prosecution, which the defendant had to pay.(40) Costs often were several times greater than the statutory fines, so Dana's discretion to set costs was more than sufficient to offset any uncertainty about the judges' fines.

By depriving judges of almost all sentencing discretion in liquor-law cases, the legislature had assured that prosecutors could--by over-charging, selectively nol prossing, and manipulating the amount of costs--dictate the defendant's sentence. That is, the legislature had empowered prosecutors to engage in charge bargaining without fear that the court's noncooperation might unsettle the terms of the bargain. In no other statute of consequence to the daily practice of the middle-tier courts had the legislature bestowed such power on the prosecutor. Excepting only those very serious crimes that carried mandatory life or death sentences, none of the typical common-law offenses called for a minimum sentence. The penalty for petty larceny was zero to one year or a fine of zero to $300; for grand larceny, it was zero to five years or zero to $600; for breaking a shop at night, it was zero to twenty years; for forging a bank bill or unarmed robbery, it was zero to life.(41)

In non-liquor cases, therefore, the prosecutor could not credibly threaten defendants with certain and steep punishment if they went to trial and lost, because the court's potential leniency had no bounds. True, the prosecutor had some power to limit the defendant's maximum exposure. At least after 1838, he could use a partial nol pros to reduce a charge of grand larceny to petty larceny and thereby guarantee defendants who pled guilty a maximum one-year sentence.(42) But the records in the earlier half of my study rarely disclose such behavior, and we may presume that simply reducing the maximum exposure, when the lowered maximum still involved substantial prison time, usually was not sufficient to induce a defendant to plead. It is therefore no surprise that of the forty-nine clear plea bargains disclosed in the record books of the middle-tier courts during the years I studied between 1789 and 1849, thirty-three--or fully two-thirds--took place in liquor cases.(43) Put differently, clear plea bargains took place in twenty-four percent of adjudicated liquor cases, but in fewer than three percent of all other adjudicated cases.(44)

As the figures in note 44 show, however, Dana's tactic in Stevens and Corey did not become a favorite of Middlesex prosecutors until sometime in the 1840s. Asahel Stearns, who succeeded Dana as county attorney and held the office for some two decades,(45) took a somewhat less aggressive tack in liquor cases. In 1824, for example, Stearns indicted twelve persons for violations of the liquor law. In three of these cases, he alleged multiple counts and exacted a plea bargain on the Stevens-Corey model. But in each of the remaining nine cases, he charged a single count of unlawful sale.(46) The defendants in all nine cases readily pled either guilty or no contest, apparently regarding their seven-dollar fines and modest costs to be preferable to an expensive trial. In 1829, the next year I studied, Stearns resolved eight out of nine liquor cases in this manner.

I do not consider these cases to be clear plea bargains, for the record discloses no concession made in exchange for the defendant's plea. They may be examples of implicit plea bargains, however, in which the defendant offered a plea expecting that some concession would be made.(47) After all, the law empowered Steams to set the costs of prosecution, and a trial no doubt would have inflated those costs. Steams tried only two liquor cases in the years I studied, only one of which ended in a conviction. In that case, from 1820, he assessed his defeated opponent costs of $39.85.(48) In the other liquor case that arose in the same term of the court, the defendant pled no contest, and Steams put costs at just $23.21.(49) And an assault-and-battery case Steams prosecuted in 1814 makes even plainer how useful a plea-bargaining tool the cost-setting power could be: After the first of two co-defendants, Archibald McIntire, pled guilty and the second, William Fletcher, went to trial and lost, the court fined each five dollars. Spurning the court's even-handedness, Stearns put McIntire's costs at $4.76 and Fletcher's at $39.53.(50)

There is some reason to think that those liquor defendants who pled guilty or no contest in 1824 and 1829 had in mind the amount of costs Steams might assess, for the records show a striking regularity in those amounts. Of the nine defendants from 1824 who pled either guilty or no contest and won no obvious concession, six were assessed costs of $13.77, two costs of $14.77, and one costs of $14.97. Of the eight defendants in this category from 1829, three were assessed costs of $11.49 and two costs of $18.14.(51) There is little reason to think that Stearns manipulated the amount of costs to induce defendants to plead. Rather, the most likely explanation for assessing different defendants the same, very specific figures is that Stearns pooled the costs of prosecuting liquor cases from a particular town and divided them among all the liquor defendants from that town who offered a plea.(52) This practice assured that defendants could know in advance the amount of costs they faced if they offered a plea. The record books disclose that in 1824 six of the nine defendants just mentioned had lawyers--and all but one of them the same lawyer--who no doubt could have helped the clients predict the consequences of their pleas.(53)

The power to set the costs of prosecution was perhaps a more potent plea-bargaining tool in liquor cases than in others. By custom, if not by law, prosecutors rarely assessed costs when a convict was sent to prison.(54) Because the liquor law did not call for imprisonment except on a second or third offense before the second half of the century,(55) costs were an almost invariable part of the disposition. And as a proportion of the total penalty, costs loomed large in liquor cases, in which the customary fine for making a single unlicensed sale was seven dollars, and costs were consistently more. At all events, Stearns was able to use cost-setting and, to a lesser extent, Dana's multiple-charge tactic as well as other stratagems to win pleas from twenty-four of the thirty-three liquor defendants he prosecuted in the years I studied. He went to trial only twice and nol prossed charges without winning pleas in seven cases. In contrast, of 224 non-liquor cases that Stearns prosecuted in these years, he resolved only eighty-one by plea (and only four by clear plea bargains), took sixty-three to trial, and abandoned prosecution by entering nol prosses in most of the rest.

If Asahel Huntington, who succeeded Stearns in about 1832,(56) proved to be even more resourceful in managing his liquor caseload, it was perhaps because he had to be. In 1834, the first year I studied of Huntington's tenure, he prosecuted ninety-five cases of all types, more than double the forty-one cases Steams handled in 1829. And that was just in Middlesex County. An 1832 statute had split the state into districts, so Huntington took on the new title of district attorney and assumed an enlarged realm that included Essex County, Middlesex's smaller but still substantial neighbor.

Within two years, Huntington had begun to formulate his signature contribution to plea-bargaining technology: the preprinted, multi-count liquor-indictment form. He deployed several primitive versions of this device in 1834. In the case of Samuel Elliot, he used a preprinted indictment form that alleged four counts of making single unlicensed alcohol sales. Huntington filled in Elliot's name, the relevant dates and places, and other information peculiar to Elliot's offense. His use of a preprinted form suggests that far from tailoring the charges to Elliot's specific acts, Huntington had embraced a routine practice of charging multiple counts of unlicensed sales. Facing this four-count indictment, Elliot chose to plead no contest on two counts in exchange for Huntington's nol pros of the other two.(57) In the case of Amos Adams, Huntington pasted together two of these four-count forms. Adams pled guilty to count one, and Huntington nol prossed counts two through eight.(58) In several other cases, Huntington used a preprinted form that alleged one count of being a common seller and one count of making a single sale, to which he often appended an additional, handwritten single-sale count.(59)

Huntington's multi-count liquor forms grew more sophisticated with time. Figure 1 reproduces one such form from 1843. In the first of five counts, it charges the defendants, Henry and Albert Sprague of Cambridge, with being common sellers of alcohol. Count two alleges a single sale, and counts three through five allege being an unlicensed retailer.(60) As revised in the state's 1836 code, the license law punished the first of these offenses by a fine of $100 and each of the last four by fines of twenty dollars exactly.(61) A defendant who dared risk trial on such an indictment could not hope for leniency if convicted, for the judge had no discretion in sentencing. This was the sad discovery of Daniel McCrillis, who in 1849 was convicted of four counts, fined $160, and assessed costs of $88.05.(62) Those defendants who chose instead to fight another day and pled to one or more of the twenty-dollar counts regularly secured nol prosses on at least the first and most expensive count and paid moderate costs of between twenty and forty dollars.(63) Huntington had complete; discretion to reward defendants in this way. As the Supreme Judicial Court wrote in 1838, "[I]t is perfectly clear that a nolle prosequi may be entered [before trial] at the pleasure of the prosecuting officer.... The Court has no right to interfere...."(64)

[Figure 1 ILLUSTRATION OMITTED]

However useful Huntington's multi-count forms and liberal nol prossing practices may have been in managing his liquor caseload, their virtues did not immediately appear to the legislators who discovered his doings sometime in 1843 or early 1844. A House investigating committee summoned Huntington to respond to two sets of charges. The first alleged simple graft--that Huntington had received fines and court costs from defendants and failed to pay the whole amount over to the state.(65) The second alleged "mal-administration ... [i]n taking less than might have been required on the discharge of indictments found and not tried."(66) These charges, the committee noted, "all relate to [Huntington's] official administration of the laws of the Commonwealth, against selling spirituous liquors without license."(67) It appears the committee was reacting in part to the September 1842 sitting of the Court of Common Pleas in Essex County, at which Huntington managed to dispose of all eighteen of his liquor cases without a trial.(68)

Shocked to find plea bargaining going on in the Commonwealth's courts, the committee demanded that Huntington appear and explain himself. He denied the allegations of graft, and the committee ultimately deemed those charges unsupported.(69) As for the charges of malpractice by plea bargaining, Huntington stoutly admitted that they "were true in fact," and "he claimed the right under the laws of the Commonwealth" to do exactly as the committee alleged.(70) Then he offered the legislators a plea-bargaining primer for use in liquor-license cases, which began with his acknowledgment of the practice of multiple charging:

[W]henever a number of penalties had been demanded in different counts of the same indictment, and the defendant came forward and proposed an adjustment, [Huntington's] usual course had been--

1st. To require the party to enter a plea of nolo contendere.

2d. To enter into an agreement to abstain from future sales of liquors without license.

3d. To pay at least one penalty to the Commonwealth, and all costs which had then accrued. And

4th. That the indictment should then stand continued as security that the defendant would fulfil his agreement, and to be further prosecuted in case the defendant still continued in his course of a wilful violation of the laws.(71)

I will return to the significance of steps two and four of this method in Part III when I take up the origins of prosecutorial probation. In steps one and three, Huntington merely adopted and institutionalized the plea-bargaining approach Dana had put to use in 1808.

Huntington contended that in light of his crushing workload, this procedure was "most conducive to the public interest."(72) The legislature had eliminated the office of attorney general in 1843,(73) leaving the various district attorneys saddled with "the entire direction and management of all prosecutions and suits in behalf of the Commonwealth."(74) And there was nothing sneaky about his plea-bargaining tactics: "It was very distinctly in evidence," the committee allowed, "that this course was taken openly and publicly ... and impartially applied to all; that it was known to the Courts, the Bar, the County Commissioners, and all other persons who had occasion to take any interest in the administration of this department of the law."(75) In the end, Huntington wholly won over the committee, whose report overflowed with his "extraordinary zeal and untiring industry"; his "intelligence, integrity, fidelity and ability"; his "arduous and exhausting" and "severe and long continued exertions"; and his devotion "even to the peril of his life"--for Huntington had been sick from overwork--"to the discharge of his official duties."(76) The committee was barely less florid in its praise of Huntington's method of plea bargaining in liquor cases, which it said "tend[ed] more than any other course in the class of cases to which it was applied, to attain the just end of all punishment, the prevention of the offence, the reformation of the offender."(77)

Despite the committee's vindication, announced in January 1845, Huntington departed from office later the same year because (it was said) he found his salary inadequate to his labors.(78) His successor perhaps took courage in the legislature's lavish endorsement of Huntington's tactics, for he dispatched his liquor cases in an even more brazenly public manner. Hence we find this notice concerning district attorney Albert Nelson in the Lowell Journal of April 2, 1847:

   LICENSE CASES. The Newburyport Advertiser says that Mr. Nelson, Attorney
   for the Essex and Middlesex district, states that at the present term, it
   is his intention when the parties complained of plead guilty, and enter
   into recognizance to observe the law, not to press for the fines which are
   incurred, but simply to exact the costs of Court. At the next term,
   however, he announces that no more settlements of cases will be made; but
   that full fines will be exacted in every case where the parties are
   convicted.(79)

I will take up in Part III how it was that Nelson could promise not to exact fines against defendants who pled guilty under a statute that mandated fines. For now, the significance of this mid-century announcement of a district attorney's fire sale in liquor-license cases is Nelson's manifest confidence in his own power to make such promises without concern about possible interference from the court.

Charles Russell Train, who followed Nelson into office in 1848,(80) proved to be the most enthusiastic of Middlesex's early plea-bargaining practitioners. Train quickly mastered the plea bargainer's art and deployed the Dana-Stevens-Corey model of charge bargaining and Huntington's preprinted, multi-count liquor-indictment forms with unprecedented regularity. In 1849, Train charged multiple counts in eighty-nine percent of his liquor indictments and managed to resolve twenty of fifty-six adjudicated liquor cases with clear plea bargains. All but one were built upon multi-count indictments, the district attorney's nol pros of selected counts, and the defendant's plea of guilty to the rest.(81)

B. Charge Bargaining in Murder Cases

Train was therefore the fifth in a chain of Middlesex prosecutors, beginning with Dana and running through Stearns, Huntington, and Nelson, to exploit an extraordinary power to craft deals in liquor cases. He appears to have been the first Middlesex prosecutor to make use of an analogous, if less precise, power to cut deals in capital cases. I mentioned earlier that the legislature abolished the office of attorney general in 1843.(82) For six years, before the legislature relented and reinstated the office,(83) the various county prosecutors took on the task of trying capital cases before the Supreme Judicial Court. During their six years before the high court, Middlesex prosecutors conducted three capital trials. In his only murder trial, Huntington won a manslaughter verdict and a seven-year prison sentence.(84) One of Nelson's two trials ended in a conviction and death sentence,(85) the other in a verdict of not guilty by reason of insanity.(86)

Train had a single capital case before the court--in October 1848--and he avoided trial with a plea-bargaining technique very similar to the one he used in liquor cases before the Court of Common Pleas. Barney Goulding stood charged with murdering his wife, Ellen, by beating her about the head.(87) He pled not guilty at his arraignment and claimed his right to a jury trial. Under a Massachusetts law that granted counsel to capital defendants, the court assigned two prominent members of the Middlesex bar to defend him.(88) The clerk recorded what happened next:

   And afterward in this same term the said Barney Goulding, otherwise called
   Barnett Goulding, retracts his [not guilty] plea above pleaded, and says he
   is guilty of manslaughter. And Charles R. Train, Esquire, attorney for the
   Commonwealth in this behalf, says, he will no further prosecute this
   indictment as to the malice aforethought, and the charge of murder.(89)

On his manslaughter conviction, the court sentenced Goulding to two years in the house of correction.(90)

Had Goulding insisted on trial and been convicted of murder, he would have faced the mandatory penalty of death. The legislature did not divide murder into degrees until 1858, and until then all murder remained capital. Train's generosity in nol prossing so much of the murder indictment as charged more than manslaughter shrank Goulding's possible sentence to a term of between zero and twenty years in prison. The judge, as the clerk reported, hewed to the bottom of that range.

Train's was not the first clear plea bargain in a capital case in Massachusetts, though it does appear to have been among the first. As early as 1804, the Supreme Judicial Court, sitting in its capacity as a trial court, accepted a guilty plea in a murder case. That plea was far from a bargain, however, as it merely secured the defendant's compulsory execution--perhaps explaining the Court's insistence that the defendant reconsider before it recorded his plea.(91) The earliest clear plea bargains I have found in capital cases took place in 1841.(92) Attorney General James Austin negotiated two such bargains before the Supreme Judicial Court that year. In one, a murder case, the bargain took much the same form as in Train's Goulding case.(93) In the second, charging rape, the defendant pled guilty to assault with intent to rape. Austin then nol prossed "the other and further part of the indictment"--that is, so much of the indictment as charged an actual rape.(94) Rape, like murder, carried mandatory death, while assault with intent to rape carried any term of years up to life in prison. In this case the judge sentenced the defendant to five years in prison. Rape soon ceased to be a capital offense--by 1852, murder was virtually the only crime still deemed capital(95)--and every other charge bargain I encountered in a capital case involved the crime of murder.

There was one other capital charge bargain during the 1840s--this one in a murder case prosecuted by the district attorney of the Southern District, John H. Clifford, in 1845.(96) Together, these four charge bargains by Train, Austin, and Clifford accounted for ten percent of all adjudicated capital cases in my survey of Supreme Judicial Court business in the 1840s.(97) In the 1850s--when Clifford, now attorney general, handled most murder prosecutions(98)--seventeen percent of adjudicated murder cases ended in charge bargains of this sort. Then in the 1860s, the trend turned sharply upward. Twenty-five of fifty-three murder cases--or forty-seven percent--ended in charge bargains. Although the rate of charge bargaining retreated somewhat to forty-three percent in the 1870s and thirty-five percent in the 1880s, the 1890s closed the century with a record sixty-one percent of murder cases ending in charge bargains.(99) In contrast, the rate of charge bargaining in all other non-liquor prosecutions never exceeded eleven percent of adjudicated cases in any one year,(100) and across all years stood at just three percent.(101)

At least three moments in this chronology of plea bargaining in murder cases seem to require some explanation: Why, first of all, did bargains in capital cases emerge in 1841 and not earlier? Why did they advance so dramatically in the 1860s? And why, after some slackening, did they advance so vigorously in the 1890s? The last question is best left until later, when we move into the last quarter of the century. I will take up the first two questions now.

At one level, it is surprising to learn that plea bargaining in capital cases emerged as early as 1841. It is true that scattered instances of plea bargaining, even in serious cases, crop up quite early in the historical record--I will consider these in more detail in Part VI. But Austin's 1841 charge bargains were not two scattered instances, but rather the beginning of a regular practice of charge bargaining in murder cases that, within three decades, accounted for nearly half of all murder cases and that persisted through the balance of the century. This early timetable challenges the views of some who argue that nineteenth-century courts resisted plea bargaining and discouraged its rise until at least the late nineteenth or even early twentieth century.(102) But in the context of my arguments here, 1841 seems rather late for the first such bargains to emerge. After all, charge bargains in liquor cases appeared as early as 1808, only one year after the appointment of the first county prosecutor, and they took place in some of the very first liquor cases heard by the Court of Common Pleas.(103) These early deals seem to have been the product of a nearly instantaneous reaction of two chemical ingredients--the existence of a public prosecutor and the prosecutor's power to bargain, conferred by the liquor law's rigid penalty scheme.

There are several masons to expect that charge bargains might have proved rarer in murder cases than in liquor cases. Perhaps the most obvious is that prosecutors might have felt that both justice and public opinion demanded a full measure of punishment in murder cases. Trading the mandatory death penalty assigned to murder for the uncertain zero-to-twenty-year prison term assigned to manslaughter might have offended basic prosecutorial instincts. It is also possible that the attorney general, who worked full time, felt less caseload pressure than did the part-time county prosecutors who handled liquor prosecutions--an issue I will take up more fully in Part II. But such arguments suggest only that charge bargains in capital cases should have been comparatively rare, whereas my research has turned up none before 1841.(104)

A possible solution to the mystery lies in the peculiar procedural form of charge bargains in capital cases. In liquor cases, as consideration for the defendant's plea to one of several counts of an indictment, the prosecutor would nol pros the remaining counts. Murder cases, in contrast, generally involved but a single offense, so there was no option to nol pros whole counts of the indictment. Moreover, as any one count charging murder was enough to assure the defendant's execution, nol prossing selected counts would create no incentive for a plea. Instead, the prosecutor's consideration for the defendant's plea took the form of a partial nol pros that effectively reduced the charge. In Train's 1848 murder case, for example, after the defendant pled guilty to manslaughter, Train said "he will no further prosecute this indictment as to the malice aforethought, and the charge of murder."(105) Similarly, in his 1841 rape case, Attorney General Austin responded to the defendant's plea of guilty of assault with intent to rape by "certif[ying] ... that he will prosecute on the other and further part of the indictment"--that is, so much of the indictment as charged rape--"no further."(106)

Only three years earlier, in 1838, Austin had asked the Supreme Judicial Court to ratify his power to nol pros part of a single-count indictment.(107) The court declared, "If the attorney general may enter a nolle prosequi as to the whole of an indictment, or of a count, so he may do it as to any distinct and substantive part of it."(108) Although the court maintained that earlier cases, "if they are law, are decisive of" the legality of partial nol prosses,(109) those earlier cases gave only shaky authority for the practice. In two of the three cases cited by the court, both from 1805, the Supreme Judicial Court itself had recommended that the prosecutor enter a partial nol pros to eliminate a possibly defective part of the indictment.(110) Capital charge bargains, in contrast, involved the attorney general's independent use of a partial nol pros. Moreover, these two cases did not consider the legality of the partial nol pros, as the defendants apparently contested only the sufficiency of the indictment.

The third case, from 1828, provided useful dictum on the question of partial nol prosses, but no law. The court wrote that in the usual case, entering a partial nol pros, at least after conviction, "could be of no prejudice to the defendant."(111) The court ruled, however, that entering a partial nol pros in the particular circumstances of that case was unfair to the defendant and therefore granted him a new trial.(112) Only in 1838, then, did the Supreme Judicial Court for the first time uphold the attorney general's unilateral power to nol pros part of an indictment.(113) It is true that a later treatise writer would cite the 1828 case as authority for the prosecutor's power to enter a partial nol pros,(114) but the Supreme Judicial Court itself, in later years, appears always to have relied on the 1838 case, though it sometimes cited one of the earlier cases as well.(115) After securing this affirmation of his powers in 1838, Austin put it to use in carrying out capital charge bargains as early as 1841. There is therefore no sharp disjunction between the early history of charge bargaining in liquor cases and its early history in capital cases. In both contexts, charge bargaining appeared almost as soon as a public prosecutor found he had the power to accomplish it.

The second puzzling moment in the chronology of capital charge bargains--the sudden surge of the 1860s--also finds at least partial explanation in the prosecutor's expanding bundle of charging powers. In 1858, the legislature separated the crime of murder into first and second degrees,(116) abruptly granting prosecutors more precise control over sentencing and more versatility in charge bargaining. Until 1858, plea bargains in murder cases could take only one form: the prosecutor's agreement to reduce the charge to manslaughter in exchange for the defendant's guilty plea. Unlike the nice calibration of sentences that prosecutors could manipulate in liquor cases through multiple charging and selective nol prossing, the penalty scheme in murder was a clumsy tool in their hands. By reducing the charge from murder to manslaughter, the prosecutor could spare the defendant the mandatory death penalty assigned to murder and guarantee him instead a prison term of from zero to twenty years. This measure of sentencing power is what permitted any charge bargaining at all in murder cases before 1858. The prosecutor could not dictate, however, where within this wide-open range the judge would set sentence. No doubt many conscientious prosecutors shunned such deals because the defendant's crime, although perhaps not warranting death, deserved far more than the two years given Barney Goulding for killing his wife.

The legislature's separation of murder into degrees helped to allay such qualms. After 1858, in exchange for the defendant's guilty plea, a prosecutor could agree to reduce a first-degree murder charge that carried mandatory death to a second-degree charge that carded mandatory life in prison.(117) The attorney general struck the first two such deals in 1863,(118) and six others followed before the end of the decade. Together they accounted for just under one-third of all plea bargains in murder cases in the 1860s. Although these second-degree deals cannot by themselves account for the striking increase in plea-bargained murder cases in the 1860s, they surely contributed to the rise. Later, I will return to this story of charge bargaining in capital cases to consider other reasons for the sharp rise in the 1860s. Guilty pleas to second-degree murder proved throughout the rest of the century to be a critical part of the prosecutor's arsenal, consistently outnumbering guilty pleas to manslaughter as a means to resolve murder cases.(119)

Despite the gulf that would seem to divide a liquor-law violation that carried a mandatory $100 fine from a murder that carried mandatory death, these two crimes share a common subplot in the story of plea bargaining's rise. In each case, the legislature's assignment of a mandatory minimum penalty--almost unique in the Massachusetts criminal universe--gave prosecutors the power to charge bargain. By using their nol pros power to reduce either the number of charges (in liquor cases) or the seriousness of the charge (in murder cases), prosecutors could promise defendants a clear and certain concession in exchange for their guilty pleas: Liquor defendants could save money, and accused murderers could save their lives. It is no wonder, then, that liquor and murder cases together accounted for the great majority of the clear plea bargains that took place in Massachusetts courts in the first two-thirds of the nineteenth century. The source of prosecutors' power to charge bargain in these cases is clear. It is time now to ask why they wanted to use it.

II. THE PROSECUTOR'S MOTIVES IN PLEA BARGAINING

A. Why Plea Bargain at All? The Role of Caseload Pressure

Asahel Huntington already has told us one answer: Prosecutors took up plea bargaining in part to escape the enormous burdens of their office. Huntington spoke from experience, for as Figure 2 shows, an unprecedented wave of cases began crashing upon him during his unlucky tenure. I noted earlier that the caseload jumped dramatically between the last years of Stearns's tenure and the beginning of Huntington's. After a lull of only forty-two cases in 1839, Huntington's caseload bounced back to 101 in 1843. In the next six years--as Huntington rounded out his years in office, Nelson came and went, and Train entered the scene--the caseload leapt to 443.(120) Although the trend turned upward again toward the end of the century, the Middlesex district attorney never again in our time frame encountered a caseload calamity of this magnitude. Across the Charles River in Boston, the Suffolk County prosecutor faced a similar shock to the system.(121)

[Figure 2 ILLUSTRATION OMITTED]

In part, no doubt, the caseload boom of the 1840s owed something to the sheer growth of crime that came with an exploding population.(122) Perhaps--though this is not clear--improved policing produced more arrests,(123) but in any event, more energetic enforcement does seem to have produced more liquor prosecutions.(124) The legislature caused a good deal of the caseload increase with its decision in 1832 to extend the jurisdiction of the Court of Common Pleas to all noncapital crimes, many of which the Supreme Judicial Court formerly had heard.(125) And the caseload grew because more defendants exercised their right to appeal from lower tribunals to the Court of Common Pleas(126)--perhaps in part a result of more defendants' having lawyers.(127)

Daunting as it surely was, the spiking caseload depicted in Figure 2 does not fully capture the pressures Huntington and his successors faced. The graph does not reflect the addition of Essex County to the district attorney's duties between 1832 and 1848,(128) or the absence of an attorney general during the legislature's unwise experiment with austerity between 1843 and 1849, which shifted responsibility for prosecution of the most serious cases onto district attorneys' shoulders.(129) All this Huntington and his successors handled with no apparent assistance, not even so much as a clerk to help draft their indictments.(130) Little wonder that on leaving office in October 1845, Huntington resorted to nol prossing 252 Middlesex cases apparently to avoid saddling Albert Nelson with his backlog.(131)

Even before the caseload boom of the Huntington years, Massachusetts prosecutors had strong reason to embrace the efficient promise of plea bargaining. Like most public prosecutors in nineteenth-century America, they worked part-time, drew (at best) part-time salaries, and therefore held more than one job. Hence Samuel Dana, who claimed he earned no salary at all as county attorney,(132) also served as president of the Massachusetts Senate during the first two years of his tenure.(133) Asahel Stearns not only served in Congress while he was district attorney, he also spent twelve years as the founding and sole full-time professor of the Harvard Law School.(134) Prosecutors with less exalted second jobs felt even more pressure to resolve their criminal cases quickly: Most supplemented their incomes with a civil law practice,(135) and because the Court of Common Pleas heard civil and criminal cases at the same session, these lawyers could do no civil business until their criminal work was done. Many district attorneys therefore rushed through their criminal cases to get on with the business of making money, as one explained to the legislature in 1844:

   The call for my attendance upon the grand jury, occasions a very frequent
   interference with my engagements in civil cases, and a consequent pecuniary
   loss to myself.... The practice to which [Samuel D. Parker, district
   attorney of Suffolk County,] alludes, of making extra exertions in order to
   dispatch the business before the grand-jury, is, I believe, common to all
   the prosecuting officers. I endeavor, if possible, to have the grand jury
   dismissed on the morning of the second day, and I generally succeed. To
   accomplish this, it is often necessary to continue the sessions of the
   grand jury until late in the evening, and the night is mostly spent in
   drawing indictments. I know of no combination of bodily and mental labor
   equal to that to which the prosecuting officers subject themselves on such
   occasions.(136)

The pressure to plea bargain was therefore part and parcel of part-time prosecuting: No matter how many criminal cases a district attorney had, he could make more money by handling them with dispatch.(137)

Still, the incentive to get through one's cases grew particularly intense in the 1840s as caseloads leapt higher. In 1843, by the same stroke with which the legislature wiped out the office of the attorney general, it salted prosecutorial wounds with a thirty-percent pay cut.(138) The result was a storm of protest from prosecutors and their supporters, who deluged the legislature with accounts of their overwork. One district attorney calculated that his new $700 salary worked out to eighty-eight cents a case.(139) Another detailed the burdens of following the court on circuit within his district:

   I reside in Greenfield. I have to travel twenty miles to attend the courts
   at Northampton, forty miles to go to Springfield, and something more than
   fifty miles by the most direct rout[e] to reach Lenox. There is no
   rail-road communication except in passing from Springfield to Pittsfield.
   As the public means of communication between Greenfield and Lenox are very
   inconvenient, and the season when the courts are held often renders it
   hazardous to travel with a private conveyance, I frequently go and return
   by the way of Springfield--which makes the distance travelled about one
   hundred miles.(140)

A third sounded themes from Dickens:

   [U]nremitted professional labors, night and day, with no vacation and no
   assistance, and perpetual confinement daily and all day, in badly
   ventilated court-rooms, will break down the best constitution.... If
   [relief] cannot be obtained, I must retire from an office too onerous to be
   borne by one unassisted individual.(141)

As for Huntington, a Middlesex County newspaper said his "inadequate and niggardly" salary was sad thanks for his "fidelity to his arduous duties,"(142) and a Middlesex legislator told his colleagues that if they "[p]ut the case of Attorney Huntington to the people of Middlesex county, ... nine-tenths of the people would say that he should have at least $1000 a year salary."(143) Admitting its errors, the legislature restored district attorneys' salaries in 1845(144) and in 1849 reinstated the office of attorney general.(145)

Perhaps from pure necessity, and perhaps from necessity tinged with pique at the legislature's offenses, prosecutors appear to have accelerated their plea-bargaining practices in the 1840s. That is when the pattern of multiple charging that Huntington built upon his preprinted forms first drew the legislature's attention and when Albert Nelson boldly put abroad the terms by which he would deal. It is when Charles Russell Train, who felt the full force of the caseload wave, began to plea bargain his liquor cases with unprecedented frequency. And it is when Train and two other pioneering prosecutors first struck plea bargains in capital cases.

I do not mean to argue that caseload pressure is sufficient for prosecutors to engage in plea bargaining. On the contrary, without the power to make meaningful concessions to defendants, prosecutors normally cannot induce them to plead. The nearly total lack of prosecutorial power to constrain the judge's sentencing discretion in non-liquor cases other than murder probably explains the almost complete absence of clear plea bargains in such cases during the first two-thirds of the nineteenth century. Nor is a big caseload a necessary condition of prosecutorial plea bargaining. Especially when prosecutors were part-time workers, as most American prosecutors were in this period, there was always some pecuniary incentive to reduce their criminal workload, however small it was. But as I believe the Middlesex experience of the 1840s helps to show, an increasing caseload obviously does increase pressure on prosecutors to plea bargain.

So I do not join the ranks of recent plea-bargaining scholars who reject "the myth of caseload pressure" told and retold by scholars of the generation before.(146) Two of the new generation of scholars, Mary Vogel and Theodore Ferdinand, have based their research in Boston courts in the nineteenth century, and their arguments against the influence of caseload pressure therefore seem directly contrary to my own analysis. I will consider their work in some depth, both on this narrow question of caseload pressure and, in a few pages, on the larger question of the impulses behind plea bargaining's rise.

Mary Vogel titles a section of her earliest work on the emergence of plea bargaining in Boston, "The Myth of Caseload Pressure."(147) Yet her evidence of caseload pressure and guilty-plea rates in fact supports a conclusion that the two are strongly and causally related.(148) Vogel's contrary belief that caseload was unimportant to the growth of plea bargaining arises not from her evidence, which shows caseload and guilty pleas rising roughly in tandem, but only from her way of charting these trends. I have reproduced in Figure 3 Vogel's diagram plotting the increase of guilty-plea rates and caseload between 1830 and 1920.(149) On the basis of this diagram, Vogel concludes that "the surge of guilty pleas, which heralded the rise of plea bargaining during the 1830s, 1840s, and 1850s, preceded rather than followed the marked increase in caseload seen after the 1840s"--suggesting that caseload pressure could not have brought on the rise of plea bargaining.(150) But the chronological ordering of the two trends is an unintended illusion of the graph's design. The graph plots both trends--caseload and guilty-plea rate--in terms of percentage increases over the year before. Both lines therefore start at zero in 1830 and climb promptly upward. Between 1830 and 1840, the guilty-plea line rises rather sharply (perhaps because the guilty-plea rate genuinely began near zero(151)), while the caseload line rises modestly. This incongruence, if we should call it that, shifts the line representing caseload to the right and makes it look chronologically behind the guilty-plea line. In fact, during the next period, 1840 to 1850, both lines rise sharply in nearly equal measure and then, between 1850 and 1860, both lines rise modestly in roughly equal measure.

[Figure 3 ILLUSTRATION OMITTED]

Between 1860 and 1870, the caseload line takes a dramatic leap upward, while the guilty-plea rate rises only modestly. But here we encounter another illusion of the graph's design: The graph compares caseload with the guilty-plea rate.(152) That is, guilty pleas are expressed as a percentage of all cases. Because caseload can (in theory) expand indefinitely whereas the guilty-plea rate cannot exceed 100%, the increase of the guilty-plea rate simply cannot keep up with an exploding caseload. It appears that between 1860 and 1870, caseload increased by a factor of two or three,(153) while the guilty-plea rate increased from about forty-five percent to sixty or sixty-five percent.(154) That is, both figures rose substantially, even if one rose more dramatically than the other. It is true that the two trends plotted on Vogel's diagram moved in opposite directions between 1870 and 1880 and between 1880 and 1890, but these instances were both exceptional--occurring only twice in the nine time periods shown--and modest, in the sense that neither trend was moving sharply.

Like Vogel, Theodore Ferdinand produces better evidence of a link between caseload pressure and the rise of plea bargaining than he seems prepared to admit. His study of the Boston Municipal Court, a jury forum nearly identical in form and function to the Court of Common Pleas,(155) shows that caseload held steady or even declined through the first third of the century. Caseload then turned sharply upward in 1836 and had more than tripled by 1844.(156) Meanwhile, Ferdinand says, plea bargaining arose in the late 1830s and early 1840s."(157)

Ferdinand then hesitates to link these two trends in a causal way: "Whether plea bargaining was a response to burgeoning caseloads," he says, "is debatable."(158) Yet the reason for his reticence seems wrong. Ferdinand concludes that plea bargaining in this era was "restricted to vice or regulatory" offenses(159)--two categories that either overlapped with the liquor law or had generally similar penalty schemes.(160) He hesitates to link rising caseloads with rising numbers of guilty pleas because in vice and regulatory offenses, "the sharpest increases [in caseload] were experienced after plea bargaining took hold."(161) But Ferdinand's concern to match the rise of plea bargaining in vice and regulatory offenses with the rise in caseload in those same offenses mistakes the mechanism that linked caseload pressure with plea bargaining. Caseload pressure generally increased prosecutors' incentives to plea bargain. Once prosecutors felt a general incentive to lighten their workload, they struck plea bargains whenever they had the power to do so--that is, whenever rigid penalty schemes permitted them to manipulate sentences by manipulating charges. There is no reason to expect that the category of cases in which prosecutors found that power--a category I label liquor-law violations and Ferdinand labels vice and regulatory offenses--should happen also to be the category that grew the fastest. But of course, my argument that prosecutors under pressure naturally chose to plea bargain in liquor cases assumes an understanding about why plea bargaining arose first in liquor cases. I will turn to that issue now.

B. Why Plea Bargain in Liquor Cases?

Ferdinand and I agree that plea bargaining in Massachusetts first took root in cases involving alcohol, but we part ways at the underlying question of prosecutorial motives. Why did prosecutors choose liquor (and murder) cases as the medium in which to mold the future? I note murder cases in parentheses because Ferdinand and Vogel did not extend their research into the Supreme Judicial Court and so did not discover early evidence of charge bargaining in murder cases. The discussion that follows therefore refers primarily to liquor cases, though lessons learned in murder cases inevitably intrude.

I will begin by setting out my own explanation of plea bargaining's early emergence in liquor cases. That the liquor law's rigid penalty scheme gave prosecutors the power to bargain in such cases is of course part of the answer, but a full answer requires as well an explanation of plea bargaining's initial failure to spread to those cases in which judges held the balance of sentencing power. Therefore, I will consider why judges did not join forces with prosecutors in promoting plea bargaining's cause.

In Section C, I will take up Ferdinand's and Vogel's alternative explanations of the early rise of plea bargaining. I have chosen to give such close attention to Ferdinand's and Vogel's work in part because theirs appear to be the only other full-scale empirical analyses of plea bargaining's rise and in part because both worked in Boston. But the greater importance of their work to my study lies elsewhere. In rejecting the importance of caseload pressure as a motivation in plea bargaining's rise, as Vogel does entirely and Ferdinand does in a more hesitant way, each has declared in differing degrees an intention to move beyond an analysis staked in the personal interests of courtroom actors and to embrace a more complex vision of the plea-bargaining dynamic. Often--perhaps usually--taking a broader focus improves the telling of history. I do not believe this is one of those times. The evidence overwhelmingly suggests that plea bargaining was the work of those who labored in the criminal courts and was the product of their personal interests and struggle for power. Though Ferdinand and particularly Vogel add richness and color to the story of plea bargaining's rise, and though I could with some ingenuity reconcile my telling with theirs, any gain in color would be a loss in accuracy.

My own explanation of plea bargaining's early dominance in liquor cases is therefore brief: Prosecutors plea bargained in liquor cases because they could. Among all of the most commonly prosecuted crimes, only in liquor cases did the statutory penalty scheme so tightly bind the judge's hands in sentencing that the prosecutor could, by manipulating charges, dictate the sentence. Not even in murder cases did the prosecutor have such complete sentencing power, for when a defendant pled guilty to manslaughter in the course of a prosecutorial charge bargain, the judge's sentencing discretion spanned a full twenty years.

The peculiar power that prosecutors held in liquor cases stands in sharp relief against the ineptitude they showed when attempting to charge bargain in non-liquor cases other than murder. In 1852, for reasons I will discuss shortly, Massachusetts prosecutors lost their power to charge bargain in liquor cases. The record books of the Court of Common Pleas for the next year show that Charles Russell Train handled 482 cases, but secured a clear plea bargain of the sort we have seen in liquor cases only once: In a two-count forgery indictment, district attorney Train nol prossed one count after the defendant pled guilty.(162) When Isaac Morse succeeded Train in 1855,(163) he apparently resolved to find new territory in which to establish a charge-bargaining regime. He settled on breaking-and-entering cases. Of thirty-nine indictments Morse sought in 1859(164) for either breaking and entering or theft from a building, he alleged separate counts for committing the crime by nighttime and by daytime in all but four of them. In contrast, in none of the fifty-two cases of this sort that Train had prosecuted in 1849 and 1853--the two years of his tenure that I studied--did he allege separate nighttime and daytime counts.(165)

Despite Morse's cleverness in positioning these cases for charge bargaining, only four of the thirty-nine ended in the sort of charge bargain that he apparently had in mind--with his nol prossing one of the two counts in exchange for the defendant's guilty plea to the other.(166) It is not surprising that the tactic was so weak. The maximum penalty for breaking and entering in the nighttime was twenty years. For breaking and entering in the daytime, it was five years.(167) But the judge's sentence in the thirty-five cases that ended in guilty findings(168) exceeded five years only twice(169) and never exceeded ten years. Because there was no minimum punishment assigned to either count and because judges were not much inclined to approach the maximum, Morse conferred little apparent benefit by nol prossing one count.(170) In fact, as the defendant could not logically be guilty of both a nighttime and a daytime break, the jury usually convicted the defendant on only one count. When the jury did return a general verdict of guilty, Morse often nol prossed one of the two counts in any event.(171) It is little wonder, then, that charge bargaining in breaking-and-entering cases worked only with unrepresented defendants. Although the defendant had counsel in thirteen of these thirty-nine cases, the four charge-bargained cases were not among them.(172) Lacking the kind of control over the sentencing decision that could make a charge bargain meaningful, Morse succeeded in deploying the tactic against only a few defendants--and they were the most poorly informed.

A difficult question lurks behind Morse's failure to broaden the frontiers of charge bargaining. I have argued that charge bargaining worked only in liquor and murder cases because only in those cases was the penalty scheme rigid enough to allow prosecutors on their own to guarantee sentencing concessions to those defendants who pled guilty. But why did prosecutors have to act on their own? The kind of sentencing power that rigid penalty schemes gave prosecutors mattered only if judges were not partners in plea bargaining. Had judges cooperated in plea bargaining, prosecutors could have bargained in almost any sort of case, even those with wide-open penalty structures.

Two bodies of evidence make it clear that through at least the first half--and probably the first three-quarters--of the nineteenth century, judges generally withheld their cooperation from plea bargaining. The first body of evidence concerns the liquor laws themselves. In 1852, when Massachusetts followed Maine in enacting statewide liquor prohibition, the legislature provided that in cases arising under the new law, "a nolle prosequi shall not be entered by the prosecuting officer, except with the concurrence of the court."(173) Apparently, the legislature had reconsidered its vindication of Asahel Huntington's charge-bargaining practices and reverted to its original, hostile posture. The impact on prosecutorial charge bargaining was immediate and devastating. In 1849, the last year I studied before the law change, Charles Russell Train resolved twenty of fifty-six adjudicated liquor cases with clear plea bargains. In all but one of these, he charged multiple counts; the defendant pled guilty on one or more counts; and Train nol prossed the balance. That year Train sought multiple counts in eighty-nine percent of his liquor indictments--evidence of a concerted plan to gain leverage in anticipation of a plea bargain.(174)

By 1853, having been stripped of his unilateral power to enter nol prosses in liquor cases, Train had abandoned his plan entirely. Not only did he not engage in a single plea bargain on the old model, it appears that he did not seek a single multi-count liquor indictment.(175) As a result, of twenty-four liquor cases adjudicated in 1853, twenty-one ended in trials. Only three ended in plea bargains, and all of these involved the somewhat covert plea-bargaining technique of placing the cases on file, which I will take up in Part III. By 1859, the volume of adjudicated liquor cases had grown, perhaps as authorities got used to the new law, but the pattern of disposition stayed much the same. Forty-three of the sixty-two adjudicated liquor cases(176) ended in trials; three ended in simple guilty pleas with no evidence of a concession;(177) and all sixteen plea bargains took the on-file form. In fact, virtually the only plea bargains that survived in liquor cases after 1852 were of the on-file variety. Only four of the 602 liquor cases in the years I studied between 1853 and 1910 ended in a plea bargain of the sort Train used so extensively in 1849.(178)

It would be hard to attribute the sudden death of multiple-count plea bargaining in liquor cases to any other aspect of the 1852 prohibition law than its elimination of the nol pros as a prosecutorial prerogative. The new law left in place the mandatory penalty scheme that had made the old license law so conducive to plea bargaining: Being a common seller, for example, still carried a mandatory $100 fine, and making a single sale carried a fixed fine of ten dollars.(179) If the 1852 law effectively killed charge bargaining in liquor cases, it must have been because judges disapproved of the practice. The law did not ban nol prosses outright, but banned them "except with the concurrence of the court." That charge bargaining in liquor cases died compels one conclusion: Judges withheld their concurrence.

The second body of evidence of judicial resistance to plea bargaining concerns the resolution of non-liquor cases in the first third of the century. The records of the Court of Common Pleas show that from 1789--the earliest year I studied--through 1829, between forty-eight percent and seventy-two percent of adjudicated non-liquor cases ended in pleas of guilty or no contest.(180) Despite these very high plea rates (which soon fell dramatically for reasons I will take up in Part IV), there is no evidence that plea bargaining was going on. That is, there is no evidence that these defendants won some understood concession for their pleas. Court records reflect a near-total absence of the sort of charge bargains I examined in Part I, in which the clerk's record made the nature of the concession clear. Outside the liquor-law context, only four such bargains appear in the fourteen years I studied between 1789 and 1829.(181) Because the penalty scheme in non-liquor cases left wide discretion in the judge's hands, the prosecutor could not dictate the sentence by manipulating charges--so any effective charge bargain required the judge's participation. The lack of charge bargaining in non-liquor cases, then, is some evidence that judges were not yet allies in prosecutors' plea-bargaining designs.

There is, moreover, no evidence of sentence bargaining in non-liquor cases during these early years. A numerical analysis of sentencing patterns discloses no sign of a well-understood practice of granting lighter sentences to those defendants who pled guilty. Tree, any such analysis is risky in this context. Each case has many unknown sentencing variables, including the defendant's background, the specific details of the crime, and the judge's habits and moods. More importantly, even if one had a large enough sample to smooth out such variations, a systematic bias still might slant the outcome: Those defendants who faced the heaviest sentences might have been more likely to go to trial, so even absent any form of plea bargaining, sentences might have been heavier after trial than after a plea. This risk of bias means that a pattern of higher sentences after trial would not necessarily indicate the presence of plea bargaining--but the absence of such a pattern probably would be evidence of the absence of plea bargaining.

The only non-liquor crime to appear during the early years of this study often enough to permit numerical analysis was simple theft.(182) Before 1809, the punishment of thieves generally involved whippings, restitution, and sale into servitude, often as alternatives to one another or to incarceration.(183) Rather than try to compare lashes and bondage with our modern punishments of prison time and fines, I have chosen to examine only those theft cases from 1809 or later, and I have excluded the few cases in those years that still authorized servitude. In the eight years I studied between 1809 and 1829, forty-three cases of simple theft ended in convictions and sentences of either time or fines. Fortunately, indictments in theft cases almost always attached an exact dollar figure to the goods stolen, so we have some way to judge whether the more serious cases tended to go to trial. As it happened, the average value of the goods stolen was very nearly the same--$14.02 for cases ending in a plea and $14.08 for those ending in a conviction after trial. And in fact, the likelihood that the defendant would go to prison was the same in both sets of cases--eighty-eight percent. But those accused thieves who pled guilty tended to suffer heavier sentences than those who went to trial and lost--on average drawing 2.5 months as opposed to only 1.8 months.

I believe this evidence, together with that presented over the last several pages, is enough to show that judges in the early nineteenth century generally were not partners in plea bargaining. Prosecutors therefore had to bargain where they could--that is, in liquor cases. In Part IV, I will ask what moved so many defendants in the first third of the century to plead guilty when they won no apparent concession in exchange. Now the important question is why no concession was forthcoming. Why didn't judges reward defendants for their guilty pleas? And why didn't they at least go along with prosecutors' attempts, after 1852, to enter nol prosses in liquor cases?

The first and most fundamental reason is that judges lacked prosecutors' incentives to plea bargain: They neither faced the same workload pressure nor found the same advantage in the certain (prosecutorial) victory of a plea. Their workload pressure was lower in part because they earned full-time salaries. In 1821 the legislature set salaries at $2100 for the chief justice of the: Court of Common Pleas and $1800 for each of the associate judges,(184) as compared to the $1000 that Asahel Huntington earned(185) until the legislature reduced his salary to $700 in 1843.(186) (As late as 1849, a salary of $2500 for the state's attorney general was considered "a good one."(187)) Although there seems to have been no law that barred judges from receiving outside income,(188) I have seen almost no evidence that any judge carried on a law practice or other trade.(189) It is true that four of the five judges of the Court of Common Pleas resigned their posts in indignation after the legislature reduced their salaries to $1700 in 1843.(190) But four new judges quickly took their spots,(191) and by 1867, associate judges of the Superior Court were drawing $4200,(192) while the Middlesex district attorney languished at $1200.(193)

Still, I do not suggest that huge caseload increases, especially those of the 1840s, did not burden judges. That the legislature expanded the Court of Common Pleas from four judges to seven between 1843 and 1851 makes it plain that all parties thought there was stress on the court.(194) But because judges drew full-time salaries and apparently sought no livelihood elsewhere, they lacked prosecutors' financial incentives to dispose of cases as quickly as possible. For them, a day spent trying criminal cases was not a day's civil income lost. Moreover, the judges profited from the sheer overwork of the district attorneys: A part-time unassisted prosecutor could charge and try only so many cases and therefore could put only so much pressure on a judge. Yet the Middlesex district attorney's salary did not begin to resemble a full-time salary until 1872, when the legislature raised it to $2000,(195) and I have seen no hint of an assistant district attorney in Middlesex before 1875.(196) When the district attorneys found themselves unable to manage their caseloads, they simply cleaned out their backlog by entering nol prosses en masse.(197) And if judges were therefore somewhat protected from the caseload pressures that plagued prosecutors, they felt perhaps not at all the prosecutor's second incentive to plea bargain--that it meant a clear victory.

Lacking prosecutors' clear incentives to plea bargain, judges may have found themselves more troubled than prosecutors by a principled objection to the practice. In the great majority of cases, the legislature had entrusted judges with very broad sentencing discretion entirely free from review. Once a judge chose a sentence of between zero and twenty years for manslaughter,(198) for example, no higher court could adjust the sentence, nor could any authority short of the governor release the defendant from it.(199) The notion of taking part in a plea bargain must have deeply offended any judge who felt the full weight of this burden. No judge could promise in good conscience to impose a particular sentence when ignorant of those facts most critical to the choice of sentence--the details of the crime and the defendant's character and criminal history. And the judge had no unbiased source of this information. It is easy to understand in this light a Michigan judge's affronted response to a defense lawyer's inquiry whether the judge would impose a certain sentence if the defendant pled guilty: "I informed him," reported the judge, "that it would be time enough for me to determine what the sentence should be after the respondent was convicted."(200) In fact, there seems to have been a widespread if not uniform nineteenth-century practice of holding a sentencing heating, perhaps involving witnesses, even after a defendant pled guilty.(201) Such a heating could enable judges to pass a conscientiously well-informed sentence, but could not become an aid in plea bargaining. As the hearing followed conviction, it could not assist the judge in determining what concession he should offer the defendant should she plead guilty.

If the sentencing power became in judges' minds a symbol of their great authority and esteem, then it would be no surprise if they bristled at prosecutors usurping any part in the course of striking a plea bargain. Here we stray from an objection founded on principle to one based on pride. I have seen no expression of judicial attitudes toward prosecutorial plea bargaining during the first half of the nineteenth century, when the practice first emerged. But around the turn of the century, at a time when one might have thought judges had grown reconciled to the phenomenon, they bitterly voiced their resentment. In 1897, the Rhode Island Supreme Court considered a nuisance case in which the prosecutor had entered a nol pros on condition of the defendant's payment of $100 plus costs. "We are aware," the court wrote,

   that the custom has obtained, to a considerable extent, for the
   attorney-general to compromise or settle this class of cases, as well as
   cases arising under the liquor law, upon the payment of a certain sum of
   money to the State by the defendant; but the practice is a vicious one, and
   meets with our entire disapproval. There is no law authorizing a sentence
   or any legal substitute therefor by consent of parties, without the
   imposition thereof by the court.(202)

In 1913, a former chief justice of Maine's Supreme Judicial Court put in still harsher terms his opinion of the way prosecutors deployed their nol pros powers:

   A reprehensible practice ... is said (and I fear with some truth) to have
   obtained in some counties for the prosecuting officer, even after
   conviction sometimes, to grant a nolle prosequi in consideration of the
   defendant's paying into the county treasury such sum of money as the
   officer may fix upon.... [This practice] is a plain, inexcusable usurpation
   of a power entrusted to the courts, the power of determining what the
   penalty should be within the limits fixed by the legislature.(203)

The author wondered at the effrontery of prosecutors who imagined "that they are better judges than the court of what is an adequate penalty for the offense."(204)

Plea bargaining in the early and mid-nineteenth century therefore encountered at least three obstacles to judicial acceptance: It served judges' needs less well than those of prosecutors; it met in some judges a principled aversion to discharging their awesome duty to sentence without full information; and it offended some judges' pride of power. As a result, prosecutors could plea bargain only where they could do so on their own, without judicial cooperation--and that meant only in liquor crimes and capital offenses. But as I have said, both Ferdinand and Vogel arrive at different explanations of the early rise of plea bargaining. It is time now to take up their evidence.

C. Competing Theories

1. The Role of Absent Victims and Savvy Defendants

Theodore Ferdinand's massive survey(205) of the court records of Boston, like my more modest study of Middlesex, produced solid evidence of a special relationship between liquor-license cases and plea bargaining in the first half of the nineteenth century. Just over half of the liquor-license cases in his sample ended in guilty pleas, as compared to only twenty-six percent of typical common-law crimes such as larceny and felony assault.(206) Moreover, the rate of guilty pleas in liquor cases increased greatly in the 1840S.(207) Although Ferdinand's statistics do not separate simple guilty pleas from clear plea bargains, as I use the term, he does report a vigorous practice of multiple charging and suggests anecdotally that the tactic was a favorite in liquor cases.(208) Hence, he tells us of Peter Bent Brigham, whose name survives on a Boston hospital, but who in humbler days in the 1840s came repeatedly to court to face a raft of liquor-law charges. Brigham pled guilty each time to one count and paid his modest fine.(209)

In other ways, Ferdinand's account of the early rise of plea bargaining differs from mine. He argues, for example, that Boston constables "invented" plea bargaining in the 1830s as a natural spinoff of the negotiations they conducted with criminals for information about other crimes or for the return of stolen goods.(210) The prosecutor, who "could readily appreciate [plea bargaining's] many advantages, and ... was in a position to implement it quickly," then expanded the practice that the constables had devised. The notion that Boston constables invented plea bargaining in the 1830s suggests that there was something so complicated about plea bargaining that constables had to invent it for prosecutors. Yet in Middlesex County, Samuel Dana hit upon a very elaborate form of plea bargaining in 1808;(211) Asahel Stearns used Dana's model in the 1820s;(212) and by 1834, around the time Ferdinand credits Boston constables with inventing the practice, Asahel Huntington had refined it with his preprinted, multi-count indictment forms.(213) As we will see in Part VI, plea bargaining had appeared in scattered other places long before its invention in Boston. That two parties might trade concessions in order to trade gains surely was not a new concept, and it took little inspiration to extend such familiar tactics to the criminal courts.

It did, however, take power. Unless the prosecutor could guarantee a sentencing concession in exchange for the defendant's plea, no bargain was likely to happen. Yet Ferdinand does not ask how it was that prosecutors were "in a position to implement" plea bargaining in liquor cases. In seeking to explain the early connection between plea bargaining and liquor cases,(214) he does not speak of the license law's rigid penalty scheme or the unusual power to craft charge bargains that it bestowed upon prosecutors. Instead he looks largely to the absence of a victim in liquor cases and to the presence of savvy, business-wise defendants. These factors have some appeal in setting liquor cases apart from most others, but perhaps less than at first appears.

Ferdinand sees two reasons why victimless crimes might have made fertile ground for plea bargaining. The less important, which he notes only in passing, is that there was no injured party to agitate for complete vindication of justice and to resist any proposed plea bargain.(215) Ferdinand does not lean hard on this argument, perhaps because he realizes that liquor crimes were not entirely victimless. Townspeople often made angry demands that prosecutors rid them of their corner tavern-keeper.(216) And although Ferdinand does not look at the work of the Supreme Judicial Court and does not write of plea bargaining in capital cases, he surely would not claim that murder cases lacked victims, though they were a source of vigorous plea bargaining in the middle of the century.

Ferdinand puts more weight on the argument that victimless crimes could end up in court only if the authorities actively went out to investigate. "[T]he investigating official," he notes, "was someone who could also begin negotiations."(217) Yet Ferdinand offers no evidence that liquor prosecutions in fact inevitably arose from official action, and there is some reason to think that official instigation was more the exception than the rule.(218) Nor does he ask how so many non-liquor defendants--for example, burglary suspects, who rarely pled guilty(219)--managed to find their way into court without the aid of an investigating official "who could also begin negotiations."(220)

As for savvy defendants, Ferdinand argues that those accused of liquor crimes "were often businessmen ... [who] recognized the waste in seriously contesting minor charges that carried little stigma" and who were therefore ready to negotiate.(221) Ferdinand surely is right that unless defendants saw a comparative advantage in pleas over trials, no plea bargain could take place. And he may be right that tradespeople were somewhat quicker than others to see the value of a good bargain. But I suspect that Ferdinand carries this argument too far when he suggests that it does much to explain why liquor laws came first in plea bargaining's chronicles. He does not, after all, consider how burglars and robbers, streetwise in so many ways, could have been dull to the benefits a plea bargain could promise. Nor does he consider whether defense counsel might have helped less savvy defendants to see the value of a good bargain.(222)

I will not press these points, however, because the lack of individual victims in liquor cases and liquor defendants' frequent status as tradespeople probably did accelerate, if not cause, the progress of plea bargaining in liquor cases. The real shortcoming of Ferdinand's theory of the evolution of plea bargaining is that its powers of explanation expire almost precisely at mid-century, when his study ends. Just two years into the second half of the century, the legislature withdrew the prosecutor's power to enter nol prosses in liquor cases and, as we have seen, killed prosecutorial charge bargaining. Whatever may have been the role of absent victims and savvy defendants in facilitating charge bargaining in liquor cases, these factors were not sufficient to keep the practice alive. In nineteenth-century liquor cases, the prosecutor's unilateral power to nol pros was apparently a necessary condition for charge bargaining to take place.

And despite Ferdinand's arguments about absent victims and savvy defendants, the prosecutor's ability to use his nol pros power to shape the defendant's sentence was also apparently a sufficient condition for charge bargaining to take place. The history of charge bargaining in murder …