The role of theory and evidence in media regulation and law: a response to Baker and a defense of empirical legal studies.

• Author: Ho, Daniel E.
• Position: Article by C. Edwin Baker in this issue, p. 651

1. INTRODUCTION II. CLARIFYING THE MUDDLE A. Our Contribution B. Baker's Position III. WHERE WE AGREE IV. A VULNERABLE THEORY V. WHY VIEWPOINT DIVERSITY MATTERS A. The FCC and the Courts B. The Empirical Turn of the Law VI. THE DANGERS OF ANECDOTALISM A. Who Wants to Be A Millionaire? B. Conglomerate Support of Fascism C. The New York Times, Pharmaceuticals, the Washington Post, and Nixon D. The Rules of Inference VII. SYSTEMATIC MEASUREMENT A. The Importance of Empirical Evidence with Calamitous Risks B. The Enterprise of Measurement C. The Role of Editorials and News D. Information Versus Sample Size VIII. THE NORMATIVE AND THE PERSONAL IX. CONCLUSION: THE VIBRANT ROLE OF EMPIRICAL LEGAL STUDIES I. INTRODUCTION

   We are grateful for Professor C. Edwin Baker's lively response to our article (1) and for the opportunity to clarify the important issues he raises, chiefly about the relationship between theory and evidence in media regulation specifically and law generally. We appreciate in particular his praise of our article's "innovative statistical techniques," (2) given that our primary aim was to provide a new approach to empirically measure substantive viewpoint diversity. Our empirical measures and methods help to address long-standing questions about whether media consolidation leads to convergence in viewpoints (the "convergence hypothesis"). We are glad that Baker agrees that the article makes progress in the empirical understanding of the media and that it is "far superior methodologically to most empirical studies that [he] ha[s] seen." (3)

   At the same time, Baker "denies the policy relevance" (4) of our article. At heart, Baker asserts that empirical evidence about viewpoint diversity is "entirely irrelevant" (5) to media regulation. (6) In the place of empirical inquiry, we should conduct "value-based inquir[ies]" (7) (for example, in the form of the theory he espouses) and, he argues, "anecdotal tales of seriously objectionable past abuses by media moguls could be much more informative" (8) than statistical inquiry. As a result, Baker argues that the FCC should aim to maximize the number of media owners ("source diversity"), as opposed to "viewpoint diversity." (9)

   We write here to defend the role of empirical research in law, as well as the conclusions of our article. As we explain below, Baker's view suffers from deep ambiguity and internal inconsistency, ignores the empirical turn in communications law, and adopts the extreme position that normative theory should displace what he calls "law schools' recent romance with statistical empiricism [and] welfare economics," which Baker views as "malignant" to the "hermeneutic discipline[]" of law. (10)

   To clarify any confusion, Part I provides an overview of the contributions of our work and what we understand to be Baker's position. Part II discusses points of agreement with Baker. Although our aim is not to address Baker's unique normative theory, Part III briefly discusses some of its palpable vulnerabilities. Part IV demonstrates why viewpoint diversity matters from the perspective of the FCC and appellate case law, and discusses the irrefutable broader empirical turn in the law. We show that the categorical distinction, espoused by Baker, between "source diversity" and "viewpoint diversity" is nowhere recognized and is expressly rejected by current law. Part V shows the dangers of anecdotalism: Baker's own reliance on "anecdotal tales" (11) refutes the position that empirical evidence is "entirely irrelevant" (12) and systematically undermines his criticisms of our article. Part VI addresses issues of how to reliably measure viewpoint diversity and the specific methodological critiques Baker raises. Part VII deals with the misreading of our normative prescriptions and ad hominem claims.

   Most disconcerting is Baker's call for anecdotal tales and normative theory to take the place of--and preclude any role for--empirical scholarship (as well as positive and economic theory) in law. We therefore conclude in Part VIII with broader implications for what we instead view as fruitful, synergistic, and interdisciplinary interplay between theoretical and empirical inquiry in law. Our original article was entirely in that pluralistic spirit, invoking legal analysis and social science, qualitative and quantitative evidence, and drawing positive and normative inferences. Empirical inquiry has the potential to resolve parts of public policy problems that would be irresolvable on normative grounds alone. Its promise is not endless--a point we expressly highlight in our article--but it surely is not "entirely irrelevant."

2. CLARIFYING THE MUDDLE

   1. Our Contribution

      Our Stanford Law Review article (13) makes four contributions to the broad question of federal regulation of media ownership.

      First, we document a trend that the FCC and numerous commentators have noted, namely the sharp empirical turn that the law on structural media ownership regulations has taken over the past twenty years. (14) The Telecommunications Act of 1996 (1996 Act) requires that the FCC periodically "determine whether any of such [ownership] rules are necessary in the public interest as the result of competition." (15) Under arbitrary and capricious review (review as a matter of administrative--not constitutional--law), the courts have increasingly required the FCC to provide evidence of the convergence hypothesis, and the FCC responded in 2002 by commissioning an unprecedented number of empirical studies on the connection between ownership and viewpoint diversity.

      Second, our article surveys existing work and highlights serious limitations to extant measures of viewpoint diversity. (16) For example, editorial endorsements of Democratic presidential candidates result in little variation between media outlets in a two-party system. (17) Alternatively, conventional reading and coding of "bias" of news articles (content analysis) can be fraught with lack of transparency and replicability, (18) as is widely acknowledged (indeed by Baker himself (19)).

      Third, given limitations of existing empirical approaches, our major contribution is to provide a new, transparent, and replicable measure of editorial viewpoint diversity by capitalizing on rapid advances in statistical measurement methodology. (20) Much like the concept of "intelligence," the idea of a "viewpoint" is complex and cannot be readily observed in a direct fashion. Nonetheless, intelligence and viewpoints do have observable implications. In educational testing, the measurement of intelligence is commonly dealt with through standardized testing--most importantly, the administration of common questions to place students on a common scale. Our approach capitalizes on that insight to look for instances where newspapers opine on common issues, namely Supreme Court decisions. For over a year, with the help of a research team of fourteen Harvard and Stanford undergraduate and law students, we engaged in exhaustive data collection to collect every editorial written on Supreme Court decisions across twenty-five newspapers from 1988-2004, personally reading over 1,600 editorial positions in the process. Adapting finely-tuned statistical methods (to account for differences across Justices, outlets, questions, and time) allows us to scale the newspapers on a substantively meaningful dimension of the Supreme Court Justices and to examine the evolution of editorial viewpoints during mergers and acquisitions.

      Our results show stability in viewpoints for three conglomerate acquisitions, convergence for the case of the merger of the Atlanta Journal and the Atlanta Constitution to form the Atlanta Journal-Constitution, and divergence for the New York Times's acquisition of the Boston Globe. (21) In short, consolidation does not inexorably lead to convergence or divergence. Our article purposely adopts a broad view of empirical inquiry as encompassing both quantitative and qualitative research. Indeed, we devote an entire section of the article to an in-depth view of these two cases, drawing on a range of qualitative (e.g., detailed reading of editorials and phone calls with editors from every newspaper in our dataset) and quantitative (e.g., subscriber databases and census data) sources to complement our measures of viewpoint diversity and draw out complexities of these cases. (22)

      Fourth, we provide brief, but specific, policy implications based on the difficulty of empirical inquiry. Appellate interpretation of the 1996 Act holds that, unless the FCC finds that an ownership regulation continues to serve the public interest (typically involving evidence of convergence), the FCC shall repeal or modify it. Our results point to a deep tension in this statutory reading between empirical justification and deregulation--the call for empirical verification subject to a high evidentiary standard may be tantamount to wholesale deregulation. Instead, incremental modification of ownership regulations may better facilitate empirical evaluation while heeding appellate interpretation. We propose that the FCC collaborate with researchers to incorporate policy evaluation into regulation. (23) Lastly, our case studies also highlight the crucial roles of editorial policies, editorial board organizational structure, and critical distinctions between mergers and acquisitions. (24)

   2. Baker's Position

      Baker's response comes in two sections. In the first (Part II), he restates in a substantially similar form to previous writings (25) his normative theory of media consolidation: "The three major reasons to oppose media concentration in general, and mergers in particular, can be labeled: (i) the democratic distribution value; (ii) the democratic safeguard value; and (iii) the media quality value...." (26)

      The democratic distribution value posits that everyone should have an "equal voice" in the same fashion of the one person, one vote requirement familiar from...