Binding Room of the Government Printing Office, early 20th century.

Which Congress is the Most Productive?

Michelle Whyman | Florida State University

Which country is more populous, China or the United States? Consider answering this question by counting the number of cities within each country. You would find that China has five times fewer cities than the United States.1 Knowing nothing more than the fact that China contains far fewer cities, you might be tempted to conclude that it has a smaller population. Of course, this measure (raw counts of cities) doesn’t account for the fact that Chinese cities are, on average, much larger and more densely populated than their U.S. counterparts. A granular count of individual people confirms that China has the largest population of any nation in the world (1.386 billion), roughly four times that of the United States (327 million). When population is measured at the gross level of geographic groups (cities) it appears that China is smaller. When measured at the more granular level of individual people, it becomes clear that China is bigger, much bigger.

Which Congress is the most productive? Conventional wisdom suggests that the modern Congress is unproductive, mired in a gridlock that limits the number of laws it enacts. This prevalent view relies on the assumption that a productive Congress is one that passes many laws or many consequential laws, and an unproductive Congress is one that passes relatively few. Counting cities to determine the size of a nation is much like counting laws to determine the productivity of a Congress. In the case of cities, their number tells you something about how people are geographically grouped, but little about the total size of the population. In the case of laws, their number tells you something about how legislators organize legal text, but little about the total volume of law they enact. I offer an alternative conception of legislative productivity based on the provisions within federal laws. Just as individual people are the most granular unit of measurement when assessing the total population of a nation, so are provisions the most granular unit of measurement when assessing the total volume of law Congress enacts.

Provisions are the smallest unit of law in the federal legal corpus that can express a compete policy idea. While most provisions contain a paragraph’s worth of text, they may be as short as an effective date, for example, or as long as a few paragraphs describing a new tax cut. Using provisions, rather than whole laws, to assess legislative productivity leads to very different inferences regarding Congress’ productivity. Measures of productivity based on counts of whole laws underestimate Congress’ productivity because, much like China’s more populous cities, modern laws contain far more provisions than earlier laws. Examining productivity at the granular level of provisions reveals that, contrary to contemporary expectations, the modern Congress is strikingly productive. Far from being gridlocked, Congress has enacted more substantive provisions in the last 40 years than any other period in its history.

At its core, measuring legislative productivity is about assessing Congress’s capacity for governing through policy change. The earliest and most widely adopted measure for assessing this capacity has been a count of the number of laws Congress enacts. Of course, laws are not equal in the degree to which they reshape governance. It was this insight that led David Mayhew to suggest that productivity may have less to do with the raw number of laws, and more to do with important or landmark legislation. In 1991, with the publication of his seminal book, Divided We Govern, Mayhew developed an innovative measure of legislative output and found that productivity was not lower under divided government than under unified party control.2 Mayhew attempted to differentiate between important laws with wide impact and trivial laws by constructing a measure that accounted for contemporaneous measures of importance at the time of the law’s passage and retroactive assessments of policy impact by policy experts. His measure consisted of a count of landmark pieces of legislation that were thought to be important at the time of passage by the political elite (Sweep 1), and considered impactful after enactment by policy experts who retroactively estimated each law’s significance (Sweep 2). The overall impact of Mayhew’s study was considerable, both in the degree to which it focused scholars on questions of productivity and the degree to which it locked in a conception of productivity that relies on a count of important laws passed, rather than the aggregate policy change that results from passage.

Subsequent research has built on and refined Mayhew’s classification of important laws. Scholars have added new categories of significance and even extended Mayhew’s method for identifying significant legislation back in time. Shortly after the publication of Mayhew’s study, observers noted that his measure captured the supply of important legislation produced by Congress, but failed to account for changes in the public’s demand for policies. In other words, Mayhew, and most productivity scholars who followed him, presented their preferred numerator without reference to a denominator. Sarah Binder responded by developing a proportional measure, incorporating demand for policies, as approximated by their coverage in the editorial pages of The New York Times.3 Her proportional measure led scholars to conclude that the modern Congress suffered from gridlock – the inability to enact legislation preferred by most Americans. With regard to less significant legislation, Dodd and Schraufnagel demonstrated that when gridlock seemingly prevents Congress from enacting significant legislation, Congress focuses on passing commemorative legislation of minimal substantive significance, something they define as “legislation that creates no rules or regulations, has no distributive import, carries with it the most minimal spending implications, generally is of parochial importance only, and will as a rule be of limited interest to major organized interests.”4

The perception that the modern Congress is deadlocked on important legislation has become indelible conventional wisdom. This thinking seems to be confirmed by both Mayhew’s dataset and a raw count of all laws enacted by Congress since 1945 (see Figure 1).


Between 1963 and 1977 (88th to 94th Congress), lawmakers passed roughly 15 landmark acts each Congress, after which productivity reached a local nadir between 1981 and 1982 (97th Congress). Between 1981 and 2014 (97th to 113th Congress), Congress passed fewer laws overall and fewer landmark laws, between 5 and 10 each Congress. The early 1980s mark the point at which legislative gridlock became seemingly entrenched. The relatively small number of laws Congress enacted in the decades that followed have done little to dissuade scholars that the modern Congress is anything other than gridlocked.

It’s likely that productivity scholars have relied on counts of whole laws, or incremental derivations thereof, to assess Congress’s capacity to govern, to the exclusion of more granular measures for a couple of reasons. First, laws are the unit of policymaking in a legislature. Congress uses bills – nascent laws – as vehicles to introduce, debate and vote on policy. Consequently, most studies of policymaking focus on this crucial unit of analysis. Because whole laws are the predominant unit of policymaking and the subject of votes, laws have emerged as the ubiquitous unit scholars adopt to trace policy change or the continuation of the status quo.

Second, congressional observers tend to assess policy change as “significant” only when it comes packaged as a highly visible new law. The number of significant or landmark laws a Congress enacts is more salient to scholars, members of the media and the public than the provisions within those laws. The number of provisions a Congress enacts measures legislative activity “under the radar,” below the threshold of general observation. Despite being less visible units of public policy, provisions are the fundament of law that determine what citizens are required, permitted or prohibited from doing. Citizens are no less bound by a provision packaged in a minor law than one packaged in one political scientists deem “important.” Nowhere is this more evident than in the U.S. Tax Code, which is an amalgamation of provisions from many landmark and nonlandmark statutes. The Internal Revenue Service is under no less of an obligation to enforce a provision if it was part of a “minor” law.

If measuring legislative productivity is about assessing the capacity of Congress to govern through policy change, then a compelling measure of that capacity must be the number of substantive provisions a Congress enacts across policy domains. The number of provisions a Congress enacts measures its ability to remold the legal order in its own image, for better or for worse. New provisions mean new requirements, benefits or burdens for all, or some subset of citizens.

Understanding policy change and, consequently, legislative productivity in terms of the accretion of provisions across policy domains is second nature to lawyers and bureaucrats. Lawyers’ duties to uphold the law and defend their clients’ rights depend upon their knowledge of the individual provisions comprising the field of law they practice. Something similar may be said of federal bureaucrats, whose job to implement the law depends upon their familiarity with provisions, but not necessarily the whole laws from which they emanate. Moreover, when Congress repeals extant law, it generally repeals individual provisions and not whole laws. Although whole laws are the salient and highly visible unit of policy change that legislators vote on, provisions are the unit that downstream political actors rely on to interpret, implement, enforce and modify the law.

Binding Room of the Government Printing Office, early 20th century

Binding Room of the Government Printing Office, early 20th century. Photo by Harris & Ewing.
Source: Library of Congress Prints and Photographs Division, public domain.

In addition to recognizing the importance of provisions as building blocks of public policy, as downstream political actors do, by embracing provision level policy change as a relevant measure of congressional productivity, I account for changes in the internal structure of federal statutes over time. Modern laws are longer and more complex. That is, they include a greater number of provisions and address a greater number of policy topics within a single piece of legislation. Because provisions are all roughly the same size and address a single topic, counting them approximates a volumetric estimate of Congress’s capacity for policy change. By contrast, counting whole laws reveals the number of successful final passage votes legislators took, but tells us little about the volume of law they enacted. As a result, I argue that any attempt to characterize Congress’s statutory output that relies on whole laws misses the true extent of Congress’s productivity, as many provisions and disparate topics are often embedded within the internal structure of a single law, sometimes called omnibus legislation.

To construct granular measures of Congress’s aggregate agenda, I turned to the United States Code, which records every provision Congress has enacted since 1789. The United States Code is the organized aggregation of all Acts of Congress and the foremost authority on the evolution of our legal corpus over time. Since 1974, the Office of Law Revision Counsel (OLRC) has been responsible for codifying and classifying the general and permanent laws of the United States into the extant structure of the Code. Whenever Congress passes and the president signs a new law, the text of the enrolled version of the bill is sent to the OLRC, whereupon counsel codifies (in the case of positive law titles) or classifies (in the case of non-positive law titles) the law’s various provisions into one or more of the 51 titles of the U.S. Code and makes a complete accounting of how its provisions affect existing laws.

To be included in the Code, provisions must affect all or some significant subset of citizens and be enacted with the intent that they be permanent additions to the federal legal corpus. Just as laws vary in their perceived importance and practical significance, so do provisions. By virtue of their inclusion in the United States Code, the provisions I include in this analysis automatically meet a minimum threshold of importance – Congress intended that they be permanent and generally applied. By relying on the Code to identify the aggregate policy output of Congress, I automatically exclude provisions of minimal substantive significance from the analysis, such as those naming post offices. Similarly, I exclude provisions from private laws, annual appropriations that are temporary in nature, and nonrecurring reports, while retaining permanent substantive provisions nested within appropriations bills, for example. In total, the federal government has enacted 268,935 provisions of federal law between 1789 and 2013.

Figure 2 contrasts two simple measures of legislative productivity, the conventional measure of the number of public laws and my new measure based on the number of provisions Congress enacted from the First to the 112th Congresses. Measuring the number of statutes enacted since 1789 suggests decreasing productivity in the modern era, whereas measuring the total number of provisions enacted suggests sustained and even increasing productivity over the last 40 years.


Which Congress is the most productive? As the opening of this essay suggests, the answer hinges on how you measure productivity. If you adopt whole laws as your unit of analysis, the 84thCongress (1955-1956) is the most productive, having enacted 638 unique public laws. If instead you adopt provisions as your unit of analysis, the 100thCongress (1987-1988) is the most productive, having enacted a whopping 13,265 substantive provisions across 395 laws. Additionally, by virtue of shifting your focus from whole laws to the provisions within them, the 84thCongress suddenly looks far less productive than it originally appeared, having enacted only 3,091 provisions.

The trend suggests that, with few exceptions, the U.S. Congress is passing more and more law as time progresses. Before the Civil War, lawmakers enacted an average of 110 provisions per Congress (First to 36th Congress, 1789-1861). After the onset of the war and during reconstruction (37th to 42nd Congress, 1862-1873), this number jumped to 890 provisions, before falling to an average of 450 between the 43rd and 72nd Congresses (1874-1933). In the shadow of both World Wars and the Great Depression, Congress kicked off a period of renewed productivity, passing an average of 2,300 provisions per Congress (73rd to 87th Congress, 1934-1962). This period overlaps with what scholars refer to as the “Textbook Congress,” continuing from the late-1940s to the mid-1960s. During the 1970s and 1980s, Congress adopted a series of legislative reforms aimed at decentralizing the power of full committee chairs and increasing the power and autonomy of subcommittees. In this period, Congress ramped up its provision level productivity to an average of 6,100 provisions per Congress (88th to 98th, 1963-1984).

Between the 1985 and 1998 (99th and 105th Congresses), lawmakers enacted more provisions than at any time in history, nearly 12,000 on average during each congress. Incidentally, it was at the end of this period that scholars began lamenting the prevalence of legislative gridlock. For example, the 102ndCongress (1991-1992) is cited by both Binder and Mayhew for its slim legislative record, despite having enacted more substantive provisions than all but three other congresses. In the decade that followed (1999-2010, 106th to 111th Congress), lawmakers passed marginally fewer provisions, around 9,300 on average. Productivity dropped during the 112th Congress to 5,535 provisions, the same as the 93rd Congress, regarded by Mayhew as one of the most productive since the Second World War. Inclusive of this decline at the end of the series, the modern Congress is far more productive, in terms of substantive provisions, than was the textbook Congress of the postwar period or the reform Congress of the 1970s and 1980s.

Sarah Binder remarked in her 2003 book that “casual observers of Washington Politics can usually tell the difference between a Congress that produces a lot and a Congress that does little.” Differentiating between a productive and unproductive Congress becomes more difficult when congressional leaders rely on larger and more complex pieces of legislation to pass their agenda. By relying on omnibus laws to “get things done” this productivity remains largely hidden from casual observers.

Adopting a finer unit-of-measurement radically changes the inferences one draws from the data. Conventional wisdom, based on counts of whole laws, says that the modern Congress is gridlocked, when Congress has enacted more substantive provisions since the early 1980s than in any other period in its history. The provisions dataset makes this formerly obscured increase in legislative productivity apparent. Coarse measures of whole laws may misstate the true level of Congress’s productivity and, consequently, gridlock in the American political system. Whole laws may be the foci of voting in a legislature, but individual provisions are the fundament of any legal corpus. For those interested in understanding how much law Congress passes, they are an invaluable new measure of productivity.



  1. World Population Review., July 25, 2019. 
  2. Mayhew, David R. 2005. Divided we Govern: Party control, Lawmaking, and Investigations, 1946-2002. New Haven, CT: Yale University Press.
  3. Binder, Sarah A. 2003. Stalemate: Causes and Consequences of Legislative Gridlock. Washington, D.C. Brookings Institution Press.
  4. Dodd, Lawrence C., and Scot Schraufnagel. 2009. “Reconsidering Party Polarization and Policy Productivity: A Curvilinear Perspective.” In Congress Reconsidered, ed. Lawrence Dodd and Bruce Oppenheimer. 9thed. CQ Press pp. 393-418. 

Comments are closed.