How Much Did the Founders Quote the Bible?

[Much of this is a repost of my comment on Ed Brayton's blog.]

Ed Brayton recently wrote yet another post on the claim the Bible had a large political influence on the Founders of the United States, in response to a Worldnet Daily article by a Tom Flannery. Jon Rowe at Positive Liberty also posted a rebuttal of Flannery's column.

I'd seen this claim about the majority of Founding era citations being to the Bible before, of course, and at one time had seen the claim that the source was research by the American Political Science Association (APSA). A claim like that demonstrates, before any other analysis is done, the speaker's foolishness, as the APSA is just an organization of political scientists that hosts meetings and publishes journals--it is not itself a research organization.

But that led me to wonder what American Political Science Review (APSR) article was the source of the claim. I put about half an hour into figuring it out, didn't, then forgot about it. But a commentor on Brayton's blog, a helpful chap posting as "Somerville," gave the full citation, which is a 1984 article by Donald S. Lutz (Thank you, S!), and I immediately downloaded the paper from JStor and took a close look at it.

First, here's Flannery's claim, verbatim:

Indeed, of the 15,000 political writings of the men who crafted the Constitution, the source they quoted most frequently in expressing their political beliefs was the Bible. A whopping 34 percent of their political quotes came straight out of the Book they hailed as the inspired Word of God.

Like many others, Flannery doesn't have the decency to provide even a hint of a source. Perhaps he is no more aware of it than I was, and is just repeating words he heard from somewhere else. But there is no doubt that the APSR article by Lutz is the source, because both the "15,000 writings" and the "34% of quotes were from the Bible" are found in that article. So Flannery is right, no? No.

To begin, the actual topic of the article is not biblical influence on the Founders, but, as the title says with admirable clarity, "The Relative Influence of European Writers on Late Eighteenth-Century American Political Thought." I recognized the subject immediately, having had to endure a dreadful American Politics seminar in grad school where we were forced to endure Louis Hartz's The Liberal Tradition in America, as well as writings by Bernard Bailyln, J.G.A. Pocock, Joyce Appleby and other historians of the founding era, all of which hastened my shift of focii to political economy and evolutionary theory. In a nutshell, the debate is about whether the Founders were classical liberals, influenced primarily by Locke, or republicans, influenced more by Michael Harrington and others. And that is what Lutz is writing about. Unless we want to take the highly dubious step of calling the author (or authors) of the Bible "European writers," his primary interest is explicitly not the Bible.

But of course it's possible that in writing about X, the author dropped a juicy little reference to Y in the article, so let's explore Flannery's claim more closely. In a nutshell, his claim is that 34% of the references found in 15,000 political documents of the founding era were biblical references. But here's how Lutz describes his research methods.

Approximately ten years ago this author set out...to read comprehensively the political writings of Americans published between 1760 and 1805...Reviewing an estimated 15,000 items, and reading closely some 2,200 items with explicitly political content, we identified and rated those with the most significant and coherent theoretical content. Included were all books, pamphlets, newspaper articles, and monographs printed for public consumption...The resulting sample has 916 items...

So Flannery's claim begins with a serious inaccuracy--Lutz only reviewed 15,000 items, he didn't actually count the citations from 15,000 items. Then he read a sample of 2,200 of those items closely, and further reduced his sample. The sample he actually used comprised 916 document--quite a long way from 15,000. In fact Flannery is off by more than an order of magnitude, a pretty serious methodological error.

The claim that "34%" of citations were to the Bible is correct. But it is also incomplete. It was not 34% of citations from 15,000 documents, but 34% from 916 documents. Lutz does not give us any clues about the citations in the other 14,000 documents he left out of his sample, because he didn't count the citations in them. They could have had more biblical citations, the same, or fewer, but we don't have enough information to make any such extrapolations.

However it is clear what was in those 916 documents, and we can make reasonable extrapolations about the way the citations were distributed among them. Here, again, is Lutz:

From Table 1 we can see that the biblical tradition is most prominent among the citations. Anyone familiar with the literature will know that most of these citations came from sermons reprinted as pamphlets; hundreds of sermons were reprinted during the era, amounting to at least 10% of all pamphlets published. These reprinted sermons accounted for almost 3/4 of the biblical citations...

Before we dig into actual numbers, consider those percentages--only about 10% of the pamphlets published in that era were religious pamphlets, yet they contained roughly 75% of all the biblical citations, meaning around 90% of the pamphlets contained only 25% of the biblical citations. Evidently these sermons had "explicitly political content," as that was the primary standard by which Lutz did his first winnowing of the sample from 15,000 to 2,200. Neverthless they were sermons, which implies they were predominantly written by preachers, whereas most (not all, but most) of the Founders were not clergy. Consequently, the remaining 90% of the documents were likely written predominantly by non-clergy--and that vast majority of the documents contained a a clear minority of the biblical references, suggesting that most of the authors of the founding era did not use the Bible as a primary source of inspiration.

Fortunately, Lutz gives actual numbers of citations, so we can do a rough calculation of average biblical citations per sermon/non-sermon.

1. In 916 publications he counted 3,154 citations.

2. 34% of those citations were biblical, so there were 1,072 biblical citations (3,154 * .34).

3. There were approximately 92 sermons in the sample (916 * .1). (Note: Here I am making an assumption that 10% of Lutz sample were sermons, based on his statement that 10% of publications of the era were sermons. Lutz does not explicitly say so, but his statement about the percentage of publications that were sermons is placed within his discussion of the sample, so the extrapolation is reasonable. If his sample had been widly disproportionate, he would presumably have said so.)

4. Approximately 804 biblical citations came from the approximately 92 sermons (1,072 biblication citations * .75).

5. From that we can estimate the mean number of biblical citations per sermon at 8.7 (804 biblical citations/92 sermons).

5. Approximately 268 biblical citations came from non-sermon writings (1,072 - 804, or 1072 * .25)

6. The sample contained approximately 824 non-sermon publications (916 * .9)

7. Finally, we can estimate that the non-sermon writings had a mean of .3 biblical citations (268 biblical citations/916 non-sermon publications).

In summary, sermons contained more than 8 biblical citations each (on average), which is not too surprising for sermons, whether they contain political content or not. But the non-sermon publications contained less than 1 citation each, on average. In fact because the mean number of citations per non-sermon is less than .5, the liklihood of a non-sermon political publication from the founding era containing a biblical citation is not even random! The political writings were more likely than not to have no biblical references at all.

But there is yet more to the story. Keep in mind that Lutz was just counting citations, not analyzing them for substance. As he explained:

Another advantage is that a citation count need not distinguish between positive and negative citations;

So Lutz was not making any claim that the Founders used the Bible to support any particular political view, a point that becomes relevant when we look at the pattern of citations by Federalists and Anti-Federalist in 1787 and 1788, the years the Constitution was being debated (Lutz's Table 4). The Federalists--the founders of our actual political system, the suporters of the Constitution, so often said to have a biblical basis--never quoted the Bible! Lutz counts zero Federalist citations to the Bible. Conversely, 9% of the Anti-Federalists' citations were to the Bible, far less than the 38% of their references to enlightenment authors (and note that the enlightenment challenged the idea of God as the source of knowledge), but obviously far more than the Federalists. So the conclusion here, if we were to follow Flannery's method, is that God opposed the Constitution. A silly conclusion, of course, but the natural outgrowth of the "they cited it" school of historical analysis.

Also significant here is that the percentage of biblical references dropped dramatically in these two years (thanks to Ed Brayton for bringing that to my attention). The average for the 1780s, as it was for the whole period of the study (1760-1805) was 34%, but in 1787/88, it dropped to 0% for Federalists and 9% for Anti-Federalists (and for those two years, there was precious little political writing that was not one or the other, so there is no great residual of biblically based political writings that is being ignored). Here is Lutz's comment on the findings for these two years:

The Bible's prominence disappears, which is not surprising since the debate centered upon specific institutions about which the Bible had little to say. The Anti-Federalists do drag it in with respect to basic principles of government, but the Federalists' inclination to Enlightenment rationalism is most evident here in their failure to consider the Bible relevant.

Assuming my extrapolations from Lutz, and my math, are correct (and, at the least, I have tried to do a real analysis of the issue, rather than merely repeating platitudes), there clearly is no justification for arguing that the Founders, except for the handful of clergy, made substantial use of the Bible at all as a source of political inspiration.

To be fair, Lutz did think the number of biblical citations was of interest, saying:

It is relevant, nonetheless, to note the prominence of biblical sources for American political thought, since it was highly influential in our political tradition, and is not always given the attention it deserves.

Nonetheless, this rough analysis of the data shows that this influence was most prominent where we would expect it to be, among the clergy, and least prominent among the non-clergy political leaders, who predominantly cited the enlightenment rationalists.

****
Note: I have emailed Dr. Lutz asking him to comment on this issue. As it is summer, and as he has certainly never heard of me, I may or may not receive a timely reply. When, or if, I do, I will post it.

Full Citation: Lutz, Donald S. 1984. "The Relative Influence of European Writers on Late Eighteenth-Century American Political Thought," The American Political Science Review 78:189-197. One can see also Lutz further analysis in his book, The Origins of American Constitutionalism, which I have not read.