The theme of the 2005 annual meetings of the Economic History Association has been War and Economic History: Causes, Costs and Consequences. In this essay I will address this theme by briefly examining the ways in which cliometricians have viewed one particular conflict—The American Civil War—over the past four decades. The first part of my essay deals with the attack, which began at the end of the 1950s, mounted by a group of “New Economic Historians” on the existing explanation of the war; the second part deals with my own adventures as I try to make sense of the economic and political factors that produced the conflict we call the Civil War.

When I entered the graduate program in economics at the University of Washington in 1959 the dominant paradigm dealing with the causes and consequences of the Civil War had been put forth by Charles and Mary Beard in their 1927 book, The Rise of American Civilization.¹

Their ideas were subsequently elaborated and refined by Louis Hacker in his book, The Triumph of American Capitalism, published in 1940.² Simply stated, the “Hacker-Beard Thesis” claimed that there was a rapid acceleration in the growth of GNP during and after the Civil War that was a direct result of changes brought about by the Northern victory in the war. Hacker and Beard called this acceleration a “Second American Revolution.” To support their interpretation of a revolutionary change, the Beards referred to “the flowing substance of things limned by statistical reports on finance, commerce, capital, industry, railways, and agriculture, by provisions of constitutional law, and by pages of statute books—prosaic muniments which show that the so-called civil war was in reality a Second American Revolution.”³

Despite this claim of statistical support for their thesis, neither Hacker nor the Beards actually presented any compelling data to support their assertion that a “revolution” had taken place during the war. By the late 1950s a growing body of statistical evidence had been collected by a group of “New Economic Historians” which confirmed that, as the Beards suggested, the American economy had indeed expanded rapidly in the three decades after 1870.⁴

But should this be considered a “revolution? Beard and Hacker saw a discontinuity in growth at the time of the war. However, the estimates of commodity output and GNP developed by Robert Gallman, which are displayed in Figure 1, show no such discontinuity. Gallman's figures provide the first estimates of income that extend back into the antebellum period. While we can not be sure what happened during the war decade, there is nothing in Gallman's data strongly to suggest a “revolutionary” change in the rate of growth between the antebellum and postbellum years on the order of magnitude suggested by Hacker and Beard.

In 1960, the New Economic History was still in its infancy. The implications of Gallman's data might have gone unnoticed for some time were it not for the efforts of historian Thomas Cochran, who published an article in the Mississippi Valley Historical Review, that challenged the conventional view of the Civil War and economic growth.⁵

“The superficial qualitative evidence is so persuasive,” he argued, “that apparently few writers have examined the available long-run statistical series before adding their endorsement to the conventional interpretation.”⁶ Cochran went on to point out that, if anything, Gallman's figures showed that the rate of growth in output may have actually slowed during the decade of the 1860s. Seldom has a major historical paradigm perished so quickly. Less than a decade after the appearance of Cochran's article, Stanley Engerman provided the coup de grace when he concluded after a careful review of the evidence that “the reservations regarding the conventional interpretation seem justified.”⁷ A new paradigm had been established: the Civil War was not a social or economic “revolution.”

While I reveled in this intellectual triumph of numbers over impressionistic interpretations, I can still recall being a bit uncomfortable over the ease with which a war that cost 625,000 lives and millions of dollars could be expunged from the explanation of economic growth between 1839 and 1899. The vast majority of my fellow New Economic Historians—who by this time were calling themselves “cliometricians” and had managed to shift the teaching of economic history from history departments to economics departments—had no such qualms. Wars, they concluded, were exogenous events that “just happened.” Therefore, the study of war was not in the purview of economic history. As Ross Robertson noted in his widely used textbook, A History of the American Economy: “Except for those with a particular interest in the economics of war, the four year period of conflict [1861–65] has had little attraction for economic historians.”⁸

When I transferred to the History Department at the University of California Riverside in the fall of 1984, my first teaching assignment included a course on the American Civil War. At this point I was confronted with a major problem. I had endorsed the cliometric view that the war was a meaningless event—at least in terms of economic growth. As an economic historian, I could blithely ignore the war by pointing out that the results of the fighting did not affect the outcome of their economic model. As a Historian, however, I would be forced to confront the hard reality that there actually had been a war! Moreover, by assuming away the economic effects of the war, the cliometricians had also eliminated some of the more significant causes of the war. In fact, I soon realized that I was left without any plausible “explanation” for why the conflict erupted in 1861. Searching for a way to explain the war to my students, I had to rethink my view of the “economics” of the Civil War. That quest would remain a central part of my research agenda for the next two decades.

At the outset of my quest, I returned to the arguments about the causes of the war put forward by the Beards a more than half century earlier. In dismissing the notion of a “Second Revolution” based on the data for GNP, the critics had largely ignored the Beards' argument that the war was an “Irrepressible Conflict” between competing economic systems. Ironically, the same data that critics used to undermine the claim that the war was responsible for the economic growth after 1860 could also be used to support the claim that sectional frictions which were evolving between the North and the South were based on economic changes. Let's take another look at Figure 1. The Beards implicitly argued that antebellum economic growth was not as robust as the Gallman figures show. Based on this view, they insisted that an economic and social “revolution” took place during the Civil War. The problem with this interpretation is that it ignores the extent to which an acceleration in economic growth was already well under way in the Northern states by the time of the Civil War.

The Rise of American Civilization was published in 1927. Sixty years later, a group of social and political historians were articulating a revised view of war's causes that incorporated many of the ideas espoused by the Beards. At the heart of this revisionism was what would become known as the “market revolution”: a complex set of economic, demographic, and social changes that swept through the Northern states in the quarter century before the Civil War.⁹

About the same time this scholarship emerged, Richard Sutch and I were examining the consequences of what we called a “life cycle transition” that took hold in the Northern states. We argued that this phenomenon created fundamentally different “visions” of political economy in the Northern and Southern states.¹⁰

Though they clearly understated the importance of antebellum growth in the North, the Beards were not unaware of what we now call the market revolution. A central point of their argument was that the North was steadily “winning” control of the federal political process. Southerners, as the Beards pointed out, “were after all fighting against the census returns, as the phrase of the day ran current.”¹¹

By the end of the 1850s, Northern political dominance had become a major threat in the eyes of Southern leaders. What the Beards and their followers failed to articulate is that, as Richard Sutch and I concluded at the end of our investigation of the war's causes:

All of this renewed the doubts that had gnawed at the back of my mind a quarter of a century earlier when the critique of the Beards had so easily succeeded in disposing of the Hacker-Beard Thesis. By the mid 1990s I was convinced that the Beards' critics had found the right answer to the wrong counterfactual question. They had shown that if there had been no war, the Northern vision of capitalism would have prevailed—just as it prevailed in the wake of the North's victory in the war. But that is exactly what the Beards themselves pointed out when they noted that, “[t]he main economic results of the Second American Revolution … would have been attained if there had been no armed conflict….¹³

It is also the result that Southerners feared in 1860. As James McPherson notes in his seminal work Battle Cry of Freedom, “[s]ecession was a pre-emptive counterrevolution to prevent the Black Republican revolution from engulfing the South.”¹⁴

A more interesting counterfactual question would be: What would have happened if the South had succeeded in creating an independent Confederacy? So it was that in the summer of 1998 my great adventure in counterfactual history began. I decided to write a book detailing a Counterfactual History if the South had won the Civil War! I titled it” The Confederate States of America: What Might Have Been.¹⁵

For cliometricians, counterfactual history is nothing more than an exercise in applying economic theory to history. You change some parameter in an economic model and discover what effect that change will have on subsequent events. Historians take a somewhat different approach. What most of them term “Alternative History” is largely an exercise in imagination. They identify some significant historical event and imagine “what if” that event had turned out differently. They then use their imagination to describe an alternative world that incorporates the effects of the counterfactual change. But in their hearts, historians know that counterfactual history is not “real” history. Because the counterfactual events they invent did not really happen, they need not take their counterfactual answers seriously. In the words of the British historian E. H. Carr, “History is a record of what people did, not what they failed to do.”¹⁶

In spite of the distain of scholars such as Carr, the world of “counterfactual” or “alternative” history includes a rather large group of authors who bring widely divergent approaches to their subjects. In its broadest scope, the counterfactual world includes science fiction devotees who tend to specialize in various schemes of time travel; novelists who place well-known historical figures into the fictional counterfactual plots of their novels; and all sorts of historians who present counterfactual vignettes to add depth to their historical narrative. If we focus more narrowly on what one might call the “academic” workplace, we find that many historians flirt with counterfactual thinking—but confine their efforts to occasional asides or footnotes. As Robert Cowley observed, “‘what if’ (or the counterfactual to use the word in vogue in academic circles) is the historians secret question.”¹⁷

One common feature among those writers who do take their counterfactual history seriously, is that their narrative tends to be highly constrained. Thus, for example, military historians love to reflect on possible counterfactual outcome of battles, but they seldom pursue the issue further. The larger issue of how that changed outcome on the battlefield affected subsequent history is touched on only briefly—if at all.¹⁸

Cliometricians eagerly embrace the idea of incorporating counterfactual questions into their economic models to “predict the past” under changed circumstances. However, they are typically reluctant to go beyond the narrow limits imposed by the assumptions of their model.¹⁹ The result is that scholars in both disciplines are not eager to tackle major counterfactual questions over a long period of time.

When I began work on my counterfactual history of the Civil War, I quickly realized that the crux of the problem confronting any writer of counterfactual history is that some basic element of the story must be imaginary. This means you can never prove or disprove any counterfactual scenario. You can only argue that this or that scenario is “better” in some way. As a reviewer of my manuscript noted: “Debating counterfactuals is like chasing dragons; you can't catch something that doesn't exist.” He went on to advise me against trying to develop any “methodology” for counterfactual history.

That seemed like good advice, but I needed some sort of overall framework within which to construct my counterfactual world. So, drawing on my longtime association with Richard Sutch—who happens to be an excellent chef—I made up a “recipe” for writing counterfactual history:

Roger's Recipe for Counterfactual History Pudding²⁰

Ingredients:

2 parts historical plausibility

1 part common sense

1 part imagination

Mix ingredients until they are blended into a smooth even texture. If the texture seems uneven or coarse, try a little more common sense. If the pudding seems gray and boring, add more imagination.

Carefully pour ingredients into a mold shaped in the form of a well-defined historical setting.

Allow to set until pudding has firmly jelled. Be careful not to remove pudding from historical setting.

Serve with a large dose of skepticism, and remember that there is no proof in this pudding!

Bon Appetit!

An obvious question to ask is: how does Cliometrics fit into all this? The answer is that cliometricians can provide the framework within which our imagination can be turned loose. When writing counterfactual history, it is easy to get so wrapped up in your imaginary world that you forget that there are many things in that world that will not have changed. To illustrate the point, consider Geoffrey Parker's analysis of what might have happened if the Spanish Armada had succeeded in defeating the English in 1588.²¹

The usual conclusion to such a counterfactual story would be that Spanish forces conquer England. But Parker carries the story a bit further by considering the fact that success of Spanish arms would not have changed the ineptitude of the Spanish king, Phillip II. “The conquest of England would have done nothing to improve the Hapsburg gene pool; it would merely have served to create more for Phillip III and his successors to lose.” What Parker refers to as “second order counterfactuals” in this case suggest that the effects of a Spanish conquest of England would be minimal.²² Parker's reminder that some things stay the same is particularly useful in understanding where cliometricians can contribute to counterfactual analysis. The “economic problem” is one of the more persistent constants of economic history. Economic models of history can provide very useful “second order counterfactuals.” To illustrate how all this works, let me quickly describe some of the challenges I faced in constructing a counterfactual analysis of a world where the South won the Civil War.

One of my first challenges was to construct a counterfactual scenario where the South wins the war. Military historians point out that war is a game of chance, which means there are few constraints that might limit the number of possible outcomes. Historians and novelists have a field day inventing ways for the South to win the Civil War.²³

Cliometricians, on the other hand, tend to find the issue of whether the South could have won the war unnecessary. They point out that the South must have won the war, otherwise there would be no reason to write a “history” of an independent Confederate States of America. Economists do not question the idea of a Southern victory—however improbable that outcome may have been. This is not surprising; these are the same scholars who were willing to accept a counterfactual world where there was no war in the first place.

However, I soon discovered that there is a rather large contingent of potential readers who would be interested in how the South wins. In fact, they are eagerly rooting for the South to win. For these people, the opening chapters are the highlight of the book, and they would just as soon the book ended with the signing of a peace treaty in the spring of 1865 in Toronto.²⁴

Modern day Confederates are not the only ones interested in the details of the war. For historians, demonstrating that there is a plausible way for the South to win is an essential part of an analysis showing how a Southern victory would have changed the world. If victory was implausible, then all the effects that might follow from that outcome are equally implausible. So it was important for the historians that I produce at least one plausible outcome that has the South win. That, as I noted, above, was not difficult. Fiddling with a few battles produced a stalemate that resulted in Lincoln losing the election of 1864 and the North agreed to a cessation of fighting.

This is the point where virtually every historian who has speculated about a Southern victory terminates the narrative. Indeed, the only significant work by a historian that addresses both the war and the postwar world following a Southern victory is McKinlay Kantor's If the South Had Won the Civil War.²⁵

Kantor's book was commissioned by Look magazine in 1960 as a part of the centennial celebration of the war. In keeping with the theme of sectional harmony at the time of the centennial celebration, his narrative eventually has the two countries cooperating in the two world wars, and being on the verge of reunion in 1960.

Kantor's book is an example of counterfactual history where imagination fails to consider the larger picture—the second order counterfactual. He totally ignores the economic realities facing an independent Confederacy. Whatever the outcome of the war, the CSA would still be an agricultural economy dependent on staple crops—the most important of which would be cotton. Once the effects of the “cotton famine” caused by the Union blockade wore off, the world cotton market would stagnate in the last part of the nineteenth century. So despite the high hopes of the Confederates, their economic future would be rather bleak. In fact, the economy of an independent CSA would look suspiciously like the economy of the postbellum defeated South that has been the object of so much scrutiny by cliometricians. Unfortunately, my “reality check” on the Southern economy raised a serious problem for early reviewers of my analysis. They thought the result rather dull. As Yogi Berra might say, “this is deja vue all over again!”

So where is the significance of the Civil War? To answer this question one must go beyond the economic situation and look at the impact of an independent CSA on the political and economic structure of the western hemisphere. Here the second order counterfactuals are far less constraining. In 1861 Lincoln said the struggle for the Union involved not just “the fate of these United States” but of “the whole family of man.” It was, he claimed, a struggle “not altogether for today” but “for a vast future” as well. A moment's reflection on the situation of a United States following a Confederate victory suggests why Lincoln's comment may have been on the mark. No longer an emerging colossus, the USA would be pinned between the CSA to the South and an unfriendly British Canada to the North. I will leave it to the analytical imagination of the reader where one goes from here.

This brings me to the final problem facing the writer of counterfactual history: How do you end a counterfactual narrative? After careful thought, I took the easy way out by conjecturing that there would be a world war in 1914 with the USA and CSA on opposite sides. Some readers have complained about this rather sudden exit, leading me to think I may have made a mistake. However, by this point in my counterfactual story every reader will have found imaginary “mistakes” in my story. Indeed, there were times—many times—when I worried if perhaps the entire enterprise was a mistake. I particularly recall one morning I woke up and felt very depressed. When my wife, Connie, asked why I felt so glum, I replied “Because I really don't want the South to win this war!” Yet deep down, I always believed the counterfactual project was not a mistake. One of my objectives was to get readers to take a fresh look at the significance of the American Civil War, and I think I did that. The world would have looked very different in 1914 had the South prevailed in 1861, and most of those differences would not have been for the better.

I will close with a word of advice to those who decide to challenge the perils of counterfactual history. Years ago, William Parker introduced my mentor, Douglass North, as President of this association. “Doug North,” he noted, “has never deigned to answer any of his critics because he is too busy leaping to his next great mistake.” Now I have not made nearly as many “mistakes” as Doug; nor have mine been nearly as large as his. But that is not from lack of effort on my part. Herbert Heaton, in the very first presidential address that I attended 40 years ago, said “If you conceive a bright new concept or hypothesis, don't be afraid to announce it…. [New ideas] have proved life giving in provoking more research than might have been inspired by sticking to the “old truths.”²⁶

I thought that was good advice at the time, and it is good advice today. So next time you have doubts about announcing your work, just remember Doug North and leap to your next “great mistake.”