Market Conditions and Public Affairs Programming: Implications for Digital Television Policy

Transcription

1 Press/Politics Napoli / Dig 6(2) i tal Tele Spring vi sion 2001 Pol icy Market Conditions and Public Affairs Programming: Implications for Digital Television Policy Philip M. Napoli This study inves ti gates whether market con di tions affect the pro vi sion of public affairs programming by tele vi sion broad cast ers. The study exam ined a random sample of 112 commercial broad cast sta tions in order to deter mine whether station char ac ter is tics, market size and demographics, and competitive con di tions affect the quantity of pub - lic affairs programming provided. The results sug gest that market con di tions have very lit tle effect on the quantity of pub lic affairs programming pro vided by individual broad - cast sta tions and that if reg u la tors wish to increase the amount of pub lic affairs pro - gramming in the digital broad cast ing realm, specific gov ern ment-man dated pro gram - ming require ments may be nec es sary. A prom i nent debate within the media pol icy arena revolves around whether the marketplace effectively pro motes the pro duc tion of certain forms of pro gram - ming. Media regulation in the United States has moved away from explicit pro - gram ming requirements for broad cast ers and toward a model in which a dereg - ulated, competitive marketplace is seen as the pri mary mechanism for ensuring sufficient availability of diverse forms of con tent (Fowler and Brenner 1982; Horwitz 1989; Napoli 1999a, 1999b, 2001). The ques tion remains, however, whether marketplace incentives are indeed sufficient for encour ag ing the pro - duc tion of cer tain forms of pro gram ming, specifically, pro gram ming with pre - dominantly educational, political, or pub lic affairs orientations (Baker 1997; Benton Foundation 1998; Brennan 1983). Proponents of the marketplace approach to media regulation argue that deregulated markets provide suf fi cient incentives for the fulfillment of a diver - sity of con tent interests, pro vided that there are numer ous com pet i tors par tic i - pating in the market (Fowler and Brenner 1982). Those who question the effi - cacy of marketplace incentives often argue that media indus tries and media Press/Politics 6(2): by the President and the Fellows of Har vard Col lege 15

2 16 Press/Pol i tics 6(2) Spring 2001 con tent possess dis tin guish ing attrib utes (Napoli 1997, 1999c) that necessitate a more proactive regulatory stance (Baker 1997). Perhaps the most impor tant of the many argu ments in this vein is that certain types of media con tent, such as educational or public affairs pro gram ming, exhibit positive exter nali ties that are not effectively captured by tra di tional eco nomic models. That is, the value of such pro gram ming extends well beyond the rev e nue it generates and the satis - faction con sum ers gain by con sum ing it. These positive externalities include enhanced cit i zen knowl edge and deci sion making, better-informed polit i cal par tic i pa tion, and a cit i zenry better capable of influencing gov ern ment to pur - sue its best interests. Accord ing to this perspective, the marketplace s inabil ity to capture these addi tional sources of value means that such pro gram ming is likely to be underproduced, relative to its true benefits to society (Brennan 1983). This situation has been described as a market failure that requires gov ern ment attention (Baker 1997). This debate recently reemerged within the con text of the Federal Communi - cations Com mis sion s (FCC) Notice of Inquiry in the Mat ter of Pub lic Interest Obligations of Broadcast Licensees. In the notice, the FCC asked, Are there suf - fi cient marketplace incentives to ensure the provision of pro gram ming respon - sive to community needs, obviating the need for addi tional requirements? (1999b:29). Pro viding pro gram ming sensitive to community needs and inter - ests has long been a significant component of broad cast ers public interest obli - gations (FCC 1999a). Such pro gram ming often has been represented by pub lic affairs pro gram ming, par tic u larly programming pro duced locally and/or addressing local interests and con cerns (FCC 1999b). In the past, broad cast ers operated under spe cific FCC-imposed programming requirements. The FCC previously required broad- cast ers to dedicate 5 percent of their air time to local pro gram ming and 5 percent to informational pro gram ming, defined as news and public affairs (FCC 1976). Clearly, such an approach reflects the assump tion that pro gram ming of this type will not be pro duced in sufficient quantities if market - place dynam ics are the sole determining factor. In the 1980s, these pro gram ming require ments were eliminated in an effort to rely on marketplace incentives to accom plish the pro gram ming objectives previously pursued via gov ern ment mandate (FCC 1984). Broad casters sub se - quently have been criticized for pro vid ing an insuf fi cient amount of public affairs programming, particularly locally produced pro gram ming (Benton Foundation 1998). The FCC is now ques tion ing its current reg u la tory stance, due largely to the arrival of dig i tal television. The FCC raised the ques tion regarding the effective - ness of marketplace incentives within the context of inquiring whether spe cific public interest pro gram ming obligations should be imposed on dig i tal tele vi sion broadcasters. The Notice of Inquiry in fact represented the FCC s first step

3 Napoli / Digital Tele vi sion Policy 17 toward the resolution of an ongoing debate about dig i tal tele vi sion s appro pri ate con tri bu tion to educational and polit i cal discourse. Dig i tal broad cast ing refers to the next generation of over-the-air tele vi sion ser vice. Dig i tal broad cast ing represents an advancement over tra di tional ana log broadcasting in that the same amount of spec trum space can be used to provide a signal with dramatically supe rior picture res o lu tion and sound (commonly referred to as high-definition tele vi sion). Dig i tal broad cast ing also allows for the trans mis sion of four to five stan dard def i ni tion chan nels within the space of a single high-definition channel (commonly referred to as multicasting or multiplexing) and for the pro vi sion of advanced data or interactive ser vices (see Advi - sory Com mit tee on the Pub lic Interest Obligations of Dig i tal Television Broad - casters 1998). The plan for the tran si tion to digital broadcasting, as out lined in the Telecom - munications Act of 1996, pro vides existing analog broad cast ers with a 6 MHz chan nel (in addi tion to their exist ing ana log channels) on which to develop dig i - tal broad cast ing services (see Telecommunications Act of 1996: sec tion 201). By the end of 2006, when the tran si tion to dig i tal tele vi sion is anticipated to be com - plete, broad cast ers will return their original analog spec trum to the FCC (Bal - anced Budget Act of 1997). 1 It is impor tant to note that, unlike most recent spec trum license assign ments, which involve FCC-administered auctions, the dig i tal television spec trum was granted to broad cast ers for free. Thus, the struc - ture of digital broad cast ing follows that of analog broad cast ing in that digital broad cast ers are licens ees of a scarce public resource (see Red Lion Broadcasting Co. v. Federal Communications Com mis sion 1969) and as such are expected to provide pro gram ming and services that serve the public interest, convenience, and neces sity (Communications Act of 1934). Consequently, in May 1997, the Clinton administration established the Advi - sory Com mit tee on the Pub lic Interest Obligations of Dig i tal Television Broad - casters (commonly referred to as the Gore Com mis sion due to then Vice Pres i - dent Al Gore s involve ment). The committee, composed of a combination of broadcast indus try executives, academics, and public interest group representa - tives, was charged with the task of recommending what public inter est obliga - tions, if any, should be imposed on dig i tal broad cast ers. In Decem ber 1998, fol - lowing a series of (often con ten tious) public meetings, the Gore Com mis sion delivered a report outlining a set of recommendations regarding how the pub lic interest standard should be applied in the dig i tal broad cast ing realm (Advi sory Com mit tee on the Pub lic Interest Obligations of Dig i tal Television Broad casters 1998). One important aspect of this report was the recommendation that com - mer cial broad cast ers should demonstrate a minimum commitment to pub lic affairs pro gram ming, par tic u larly pub lic affairs pro gram ming produced locally and/or addressing local interests and con cerns. Although the Gore Com mis sion had no author ity to imple ment any of the recommendations that emerged from

4 18 Press/Pol i tics 6(2) Spring 2001 its meet ings, the deliv ery of its report to the FCC pro vided the impe tus for the FCC s recent Notice of Inquiry, and the report s recommendation for spe cific public affairs pro gram ming requirements motivated the FCC s inquiry into whether there are sufficient marketplace incentives to ensure the provision of pro gram ming responsive to community needs. In address ing this issue, this study investigates whether marketplace con di - tions affect the provision of pub lic affairs pro gram ming by ana log tele vi sion broadcasters. The broad intention of this study is to provide insights into the ade - quacy of the marketplace in pro mot ing the production of socially beneficial con - tent. This examination of the rela tion ship between market conditions and pub lic affairs pro gram ming in the analog tele vi sion environment can also pro vide insights into television broadcasters pro gram ming practices that can then be applied to the spe cific issue of public interest pro gram ming obligations in the dig i tal realm. If the provision of pub lic affairs pro gram ming is responsive to mar - ket con di tions, then gov ern ment efforts to encourage its pro duc tion may be unnec es sary. If, however, the pro vi sion of public affairs pro gram ming is not responsive to market con di tions, then gov ern ment action may be nec es sary to ensure the adequate availability of such pro gram ming. Method To investigate the relationship between market con di tions and pub lic affairs programming, a quantitative analysis of the pro gram sched ules of a random sam - ple of 112 com mer cial television stations was con ducted. These 112 sta tions represent approximately 10 percent of the roughly 1,200 com mer cial tele vi sion stations licensed in the United States. Market conditions were operationalized in terms of the size and demo graphic com po si tion of the television audience within a sta tion s media market and in terms of the competitive con di tions in each sta - tion s media market. In addition, indi vid ual sta tion char ac ter is tics were taken into account in order to control for the possibility that sta tion char ac ter is tics affect pub lic affairs pro gram ming decisions. These audience, competition, and station variables were then regressed against stations pub lic affairs pro gram - ming hours to determine whether indi vid ual sta tion char ac ter is tics, market demo graphic fac tors, and competitive con di tions affect the quantity of pub lic affairs pro gram ming provided. The specific data gathered and variables created are described below. Data and Variables The sam ple of com mer cial television stations was obtained from the third edi - tion of the 1999 Investing in Television Mar ket Report, published four times a year by BIA Research. This report con tains up-to-date market-by-market information for all com mer cial and noncommercial tele vi sion stations operating in the

5 Napoli / Digital Tele vi sion Policy 19 United States. The nec es sary market and station data were also obtained from this report. The data set con structed includes information on the size (in terms of tele vi sion households), aver age annual house hold income, and minor ity pop u - lation of each station s market. Minor ity population was measured by add ing the per cent age Black, per cent age Asian, and per cent age Span ish-speak ing statistics provided in the Investing in Tele vi sion Mar ket Report (BIA Research 1999). This information was gathered in order to account for the pos si bil ity that the size and wealth of a sta tion s market affect the amount of public affairs pro gram - ming a sta tion provides (see FCC 1984: Appendix C) and for the pos si bil ity that minor ity populations fac tor into public affairs pro gram ming decisions. Larger audi ence bases may translate into a greater diversity of viewer interests (in terms of pro gram-type preferences) and, hence, more pub lic affairs pro gram ming. Wealthier markets may also be markets with higher aver age education levels, which may translate into greater viewer demand for public affairs pro gram ming. Finally, larger minor ity populations may translate into more public affairs pro - gramming, given that many pro grams labeled as public affairs pro grams are spe - cifically oriented toward minor ity audi ences and con cerns. Infor ma tion was also gathered on the competitive con di tions in each sta tion s market (e.g., cable penetration, num ber of public television stations, num ber of commercial tele vi sion stations). These measures were obtained to test whether the intensity of competition for television audiences affects the lev els of pub lic affairs pro gram ming that com mer cial broad cast ers provide. For instance, greater presence of cable or pub lic tele vi sion may discourage commercial broad - cast ers from air ing public affairs pro gram ming due to its availability via these alternative outlets, or it may encour age public affairs pro gram ming if broadcast - ers elect to compete with cable and pub lic tele vi sion for public affairs viewers. Greater num bers of com mer cial broad cast ers in the market may have sim i lar effects on the pro gram ming decisions of indi vid ual broadcasters. Finally, infor ma tion on individual station char ac ter is tics, such as estimated annual rev e nues, VHF or UHF broadcast fre quency, and network affiliation, was gathered in an effort to account for addi tional potential explanatory fac tors for variation in the quantity of pub lic affairs pro gram ming. For instance, network affiliates may be less inclined to air local pub lic affairs pro gram ming due to the quan tity of broadcast time they defer to the networks. Sta tion revenues may also factor into a sta tion s deci sion to pro duce public affairs pro gram ming, with wealthier stations per haps more likely to incur the expense of producing local public affairs pro gram ming (FCC 1984: Appendix C). In incorporating sta tion revenues as an independent variable, it was nec es sary to exclude from the sample those stations that did not report rev e nues in the Investing in Television Mar ket Report (BIA Research 1999). Only stations that reported revenues were included in the study, due to the fact that pre vi ous research suggests that sta tion rev e nues may be an impor tant fac tor in determining pro gram ming decisions (FCC 1984:

6 20 Press/Pol i tics 6(2) Spring 2001 Appen dix C). According to BIA Research (1999:6), almost 80 percent of sta - tions surveyed reported their revenues. This is a high level of participation that alleviates con cerns about poten tial nonresponse error affecting the results. To gather data on stations public affairs pro gram ming prac tices, the broad - cast sched ules for each station included in the sam ple were analyzed for the two-week period beginning on Jan u ary 17, 2000, and concluding on Jan u ary 30, This period appears reasonably representative of a typ i cal two-week broadcast period. 2 It represents the heart of the network broad cast ing sea son (which runs roughly from September through May). In addi tion, none of the four teen days studied falls into any of the four rat ings sweeps peri ods, in which pro gram ming strat e gies and practices typically deviate from the norm in an effort to boost ratings. Dur ing sweeps peri ods, it is more likely that public affairs pro gram ming will be pre empted (Moonves 1998). One possible bias to this data set is the selected time period s prox im ity to the presidential pri ma ries. This fac - tor may arti fi cially inflate the amount of public affairs pro gram ming presented. An examination of the data gathered, however, revealed very few pro grams devoted specifically to the pres i den tial campaign. Pro gram schedules for each sta tion included in the sample were obtained using ClickTV ( a national television sched ule database pro - vided by TV Data, one of the nation s leading pro vid ers of tele vi sion pro gram sched ule information (see ClickTV provides zip-code-based search ing of broadcast, cable, and satellite television schedules. The ClickTV database cov ers twenty-four hours a day and encom passes pro grams as short as fifteen minutes. The relevant station zip codes were entered in order to pro duce the corresponding pro gram sched ules for the two-week time period. These pro gram sched ules were then keyword searched, using the term pub - lic affairs, which is one of the pro gram-type designations used by ClickTV to iden tify pro grams and which was used to define the uni verse of potential pub lic affairs programs. Relying on this pro gram guide ser vice may have led to some public affairs pro grams being excluded. However, it was decided that rely ing on a third-party source was a far supe rior approach than relying on broad cast ers self-reports as to what con sti tutes a pub lic affairs pro gram as pre vi ous research has (FCC 1984) given the pos si bil ity for self-interested behavior on the part of broad cast ers (given that pub lic affairs pro gram ming is generally looked on positively by reg u la tors) and broad cast ers demonstrated his tory of misrepresenting their pro gram offer ings to pol icy mak ers (Kunkel 1998). In addi tion, it is impor tant to note that the public affairs pro gram-type designation is used not only independently but also in con junc tion with other pro gram-type designations (e.g., pub lic affairs/legal or public affairs/community). Thus, it is unlikely that a keyword search using the pub lic affairs ter mi nol ogy failed to produce scheduled public affairs pro grams. Indeed, pre lim i nary exploration of the ClickTV data base pro duced no instances in which related pro gram

7 Napoli / Digital Tele vi sion Policy 21 cat e go ries, such as community or legal, were used with out being linked with the public affairs cat e gory. Exploration of the data base also pro duced no instances in which pro grams clearly representative of the public affairs cat e gory were classified under a different pro gram type. The ClickTV list ings con tained the following information about the pro - grams: (1) time of broadcast, (2) sta tion call let ters/chan nel, (3) pro gram length (in minutes), and (4) brief descriptive information. This information was used to con firm whether a pro gram labeled as a pub lic affairs pro gram indeed mer ited inclu sion in the final data set. Using the pro gram titles and descriptions, each pro gram was assigned to one of three cat e go ries: (1) pub lic affairs pro gram, (2) non pub lic affairs pro gram, or (3) can not be determined by available data. The guid ing def i ni tion of public affairs pro gram ming came from the FCC, which has defined pub lic affairs pro gram ming as pro grams dealing with local, state, regional, national or international issues or prob lems, doc u men ta ries, mini-doc u men ta ries, pan els, roundtables and vignettes, and extended coverage (whether live or recorded) of public events or proceedings, such as local coun cil meet ings, con gres sio nal hear ings and the like. (1984:172) Representative public affairs pro grams include pro grams such as NBC s Meet the Press and CBS s Face the Nation. The FCC traditionally has differentiated pub lic affairs pro grams from news pro grams, which the FCC has defined as reports dealing with current local, national and international events, including weather and stock market reports, and com men tary, analysis, or sports news when they are an inte gral part of a news pro gram (FCC 1984:171-72). This dis tinc tion between public affairs pro gram ming and news was maintained by the Gore Commission as well, which noted that public affairs pro gram ming... is not defined as coverage of news itself (Advi sory Committee on the Pub lic Interest Obligations of Digital Television Broad casters 1998:48). Consequently, this dis - tinction has been main tained in this study, and all news pro grams were excluded from the data set. Using this cod ing scheme, a second coder coded a random sam ple of 10 per - cent of the data set. An intercoder reliability score of.92 was achieved using Scott s Pi (Riffe et al. 1998). For those pro grams that fell into the can not be deter mined cat e gory, the appro pri ate station was con tacted via tele phone or in order to make a final determination as to whether the pro gram was appropriately clas si fied as a public affairs pro gram. In each of these cases, pro - gram mers were asked to describe the con tent of the pro gram at issue. Based on these descriptions, both cod ers then decided whether the pub lic affairs pro gram designation was appropriate. Although locally produced public affairs pro grams often have been the focus of communications pol icy mak ers, this study also

8 22 Press/Pol i tics 6(2) Spring 2001 approached public affairs pro grams more broadly, given that, in many instances, local pro gram mers import pub lic affairs pro gram ming from out side their mar - ket in an effort to appeal to particular audi ence seg ments within their commu - nity (e.g., import ing for eign-lan guage public affairs pro grams or senior citizen focused public affairs pro grams). As pol icy mak ers have noted on occasion, local ism need not be expressed purely in terms of geography (see Duggan 1991:11). Localism can also be expressed in terms of shared cul tural val ues or interests (Napoli 2000). Moreover, many public affairs pro grams are national network pro grams (e.g., Meet the Press, Nightline) or are nationally syndicated pro grams (e.g., Amer ica s Black Forum). Consequently, the anal y ses that follow examine both locally produced pub lic affairs pro gram ming and pub lic affairs pro gram ming in its entirety (local and nonlocal public affairs pro gram ming combined). To dis tin guish between local and nonlocal pub lic affairs pro gram ming, pro - gram titles and descrip tions were used to assign each pub lic affairs pro gram (pro grams iden ti fied in the pre vi ous coding stage as non public affairs pro grams were not included in this stage) to one of three cat e go ries: (1) local public affairs pro gram (pro grams pro duced locally and deal ing with local issues), (2) nonlocal public affairs pro gram (pro grams pro duced outside the sta tion s market area), or (3) can not be determined by available data. Intercoder reliability for this cod ing scheme was.95, using Scott s Pi. The television stations were con tacted via tele - phone or to clar ify the many instances in which it was unclear from a pro - gram s descrip tion as to whether or not the pro gram was a local public affairs pro gram (i.e., those pro grams assigned to cat e gory 3). Each sta tion s aggregate hours of local pub lic affairs pro gram ming and total (local + nonlocal) pub lic affairs pro gram ming for the two-week time period serve as the depend ent vari - ables for this study. All of the independent and dependent variables created for this study are sum - ma rized in Table 1. Given the lack of pre vi ous research on this subject, no spe - cific hypoth e ses have been for mu lated regard ing the rela tion ships between the inde pend ent and depend ent variables. Results Overall, this sample of 112 stations included stations from 83 of the 211 tele - vision markets in the United States. Eighty-four of these stations (75 percent of the sam ple) are affiliates of one of the Big Four broadcast networks (ABC, NBC, CBS, FOX). Twenty-four stations (21.4 per cent of the sample) are affiliated with one of the three smaller networks (WB, UPN, PAX). The remain ing four sta - tions (3.5 percent of the sam ple) are not affiliated with any of these networks. The VHF-UHF split is 50.9 per cent UHF and 49.1 percent VHF.

9 Table 1 Independent and de pend ent variable labels and de scrip tions Vari able Napoli / Digital Tele vi sion Policy 23 Descrip tion Independent variables Station characteristics REVENUE Com mer cial broad cast sta tion revenues (000). VHFUHF Whether a station is VHF or UHF (0 = UHF, 1 = VHF). BIG4 Whether a station is a Big Four (ABC, CBS, FOX, NBC) OTHERNET affiliate (0 = no, 1 = yes). Whether a station is an affiliate of one of the smaller networks (PAX, UPN, WB) (0 = no, 1 = yes). Audience demo graph ics TVHH Number of tele vi sion house holds in a station s market (000). HHI Aver age household income in a sta tion s mar ket (000). MINORITY Percentage minor ity population in market (% Black + % Asian + % Spanish-speaking). Competitive conditions PUBLICTV CABLE STATIONS Number of public television stations in a sta tion s market. Percentage of households in a sta tion s market sub scrib ing to cable television. Number of commercial tele vi sion stations in a sta tion s market. Dependent variables LOCALPA TOTALPA Hours of local pub lic affairs pro gram ming broadcast by a sta tion during the two-week time period. Com bined hours of local and nonlocal pub lic affairs pro gram ming broad cast by a sta tion dur ing the two-week time period. Table 2 Public af fairs pro gram ming practices of national commercial broad cast sta tion sample for two-week time pe riod (N = 112) Percent of Hours per Public Affairs Pro gramming Type Hours Broadcast Time Station Local Total (local + nonlocal) As Table 2 indicates, the 112 stations sam pled aired a total of hours of local public affairs pro gram ming during the two weeks studied. These hours represent 0.3 percent of the total broadcast hours stud ied (14 days 24 hours 112 stations) and an aver age of 1.06 hours per sta tion. The sampled sta - tions aired a total of hours of all forms of pub lic affairs pro gram ming

10 24 Press/Pol i tics 6(2) Spring 2001 Table 3 Sum mary of simultaneous re gres sion analysis for variables predicting hours of lo cal public af fairs pro gram ming (N = 112) Vari able B Standard Error of B Beta REVENUE VHFUHF BIG OTHERNET TVHH HHI MINORITY PUBLICTV CABLE STATIONS * Con stant Adjusted R 2 =.03, p >.05 Note: See Table 1 for an explanation of the variables. *p <.05. (local + nonlocal). These hours represent 1.09 percent of the total broadcast hours studied and an aver age of 3.66 hours per sta tion. It is difficult to make an objective determination as to what com prises an ade - quate or appro pri ate amount of pub lic affairs pro gram ming. Indeed, it is beyond the scope of this article to deter mine whether the amount of time commercial broad cast ers are devoting to public affairs pro gram ming is suf fi cient. However, the data clearly indicate that com mer cial broad cast ers devote a very small per - cent age of their overall broad cast time to public affairs programming, particu - larly locally pro duced public affairs pro gram ming. Local Pub lic Af fairs Pro gramming Table 3 pres ents the results of a regres sion analysis with local public affairs hours as the depend ent variable. It should be noted that tol er ance statistics and cor re la tion coef fi cients indi cated no sig nif i cant multicollinearity prob lems among the independent variables, nor were there any sig nif i cant indications of non lin ear rela tion ships between any of the independent and dependent vari - ables. Consequently, no variables have been com bined or omitted, nor have any linear trans for ma tions been imposed on the data set for this regres sion or for the regres sion that follows. As the table indicates, the adjusted R 2 for this model is.03 (p >.05). Among the inde pend ent variables, only the total number of com mer cial tele vi sion sta - tions ( STATIONS ) in the mar ket was sig nif i cant at the.05 level (beta =.37, p <.05), although the low R 2 indicates that this relationship is so weak as to be of

11 Napoli / Digital Tele vi sion Policy 25 Table 4 Sum mary of simultaneous re gres sion analysis for variables predicting hours of to tal public af fairs pro gram ming (N = 112) Vari able B Standard Error of B Beta REVENUE VHFUHF BIG OTHERNET TVHH HHI MINORITY PUBLICTV CABLE STATIONS ** Con stant Adjusted R 2 =.23, p <.05 Note: See Table 1 for an explanation of the variables. **p <.01. no practical significance. The remain ing competitive con di tions indi ca tors (cable penetration and the num ber of pub lic tele vi sion stations in the market) exhibited very weak rela tion ships with the dependent variable. Neither of these variables was significant, nor were any of the market size/demo graphic or sta - tion char ac ter is tic variables sig nif i cant at the.05 level. These results sug gest that individual stations do not respond to increasingly competitive market conditions by pro duc ing more local public affairs pro gram - ming. Nor, for that mat ter, do they respond by reducing the amount of local pub - lic affairs pro gram ming they provide. Instead, pub lic affairs pro gram ming appears to be unaf fected by competitive conditions. The results also suggest that local pub lic affairs pro gram ming is not a function of the size or demo graphic char ac ter is tics of the poten tial audience, nor is it a func tion of the basic attrib utes of the broadcast sta tion. Thus, the pro vi sion of local public affairs pro gram ming appears highly resis tant to variations in marketplace conditions. Total Public Af fairs Programming A slightly different picture emerges, however, when public affairs pro gram - ming is defined more broadly specifically, in terms of both local and nonlocal public affairs pro gram ming. Table 4 pres ents the results of a regres sion analysis with total (local + nonlocal) public affairs pro gram hours as the depend ent vari - able. As the table indicates, the adjusted R 2 for this model is.23, which is sig nif i - cant at the.05 level (p <.05). The total num ber of com mer cial television sta - tions ( STATIONS ) is sig nif i cant at the.01 level (beta =.46, p <.01). No other inde pend ent variables are sig nif i cant at the.05 level.

12 26 Press/Pol i tics 6(2) Spring 2001 The sig nif i cant positive coef fi cient for the num ber of com mer cial tele vi sion stations in the market (beta =.46, p <.01) suggests that higher num bers of com - peting com mer cial tele vi sion stations will com pel com mer cial television broad - cast ers to increase the amount of pub lic affairs pro gram ming they provide. Thus, when public affairs pro gram ming is defined more broadly (to include local and nonlocal public affairs programs), increased competition from other com mer - cial tele vi sion stations does appear to have a modest positive effect on the amount of pub lic affairs pro gram ming that com mer cial broad cast ers choose to air. How - ever, the fact that more than 75 per cent of the variance in pub lic affairs pro gram - ming remained unexplained by the model suggests that pub lic affairs pro gram - ming decisions are quite resistant to marketplace influences. Conclusion Overall, these results provide sup port for the conclusion that market incen - tives may not be suf fi cient to pro mote the provision of pub lic affairs pro gram - ming, par tic u larly local public affairs pro gram ming. The availability of pub lic affairs pro gram ming is not significantly related to any of a variety of audience and station characteristics. Broad casters commitment to pub lic affairs pro gram - ming does not appear to be affected by the composition of a station s potential audience, suggesting that pub lic affairs pro gram ming is not perceived by broad - cast ers as an effective means of appeal ing to particular audi ence seg ments (in terms of either income or ethnicity). Nor do broad cast ers pub lic affairs pro - gram ming decisions appear to be affected by a station s finan cial resources, broadcast fre quency, or network affiliation. The fact that station revenues were not a factor is particularly important, given that wealth ier stations are often assumed to be more able and more willing to provide pub lic ser vice pro gram - ming (FCC 1984: Appendix A). The results presented here raise questions about this assump tion. Competitive con di tions within a sta tion s market also bear very lit tle rela tion - ship to the amount of pub lic affairs pro gram ming a sta tion provides. The rela - tion ship between the num ber of competing com mer cial television stations and public affairs pro gram ming was found to be so weak within the con text of local public affairs pro gram ming as to be of no practical significance. Competition has previously been found to positively affect the provision of news pro gram ming (Busterna 1980; Lacy and Riffe 1994). Thus, the results presented here suggest that local pub lic affairs pro gram ming, unlike news pro gram ming, is not per - ceived by com mer cial broad cast ers as an effective means of combating competi - tion within the media marketplace. A modest relationship was found between the num ber of com mer cial television stations and all forms (local + nonlocal) of public affairs pro gram ming. It is possible that this relationship between competi - tive con di tions and public affairs pro gram ming is stron ger within the con text of

13 Napoli / Digital Tele vi sion Policy 27 all forms of pub lic affairs pro gram ming than within the con text of local pub lic affairs pro gram ming because stations are likely to respond to competitive pres - sures to provide public affairs pro gram ming (weak as they may be) by air ing cheaper nationally syndicated fare rather than by incur ring the expense of pro - duc ing their own pro gram ming. Or, per haps, broad cast ers simply perceive a stronger demand for public affairs pro gram ming dealing with national-level issues than with local-level issues. These are, however, merely speculations that can not be con firmed with the available data. Previous research, which studied, in the aggregate, a broader range of pro - gram types (news, local pro gram ming, and all forms of public affairs), found much stron ger rela tion ships between market conditions, sta tion char ac ter is tics, and the amount of pro gram ming provided (FCC 1984: Appen dix A) than were found in this study, in which only pub lic affairs pro gram ming was stud ied. These contrasting results sug gest that pub lic affairs programming, in par tic u lar, may be resistant to variation in sta tion and market conditions. As pol icy mak ers consider whether to impose spe cific public inter est pro - gram ming requirements on dig i tal broadcasters, the results presented here sug - gest that, at least in terms of pub lic affairs pro gram ming, it is unlikely that mar - ket incentives will pro mote the production of such pro gram ming. If pol icy mak ers desire a level of pub lic affairs pro gram ming in dig i tal broad cast ing that exceeds the lev els currently available in the analog environment, then the insti - tu tion of spe cific public affairs pro gram ming obligations may be nec es sary. Such actions may be com pelled further by the like li hood that public affairs pro gram - ming is not being produced in quan ti ties that are representative of its full value to society, given the positive externalities associated with such pro gram ming (Baker 1997; Brennan 1983). Possible regulatory actions might include requir - ing broad cast ers to devote part or all of the broad cast time on one of their multi - plexed chan nels to educational and public affairs pro gram ming or estab lish ing minimum quantitative guide lines as to the per cent age of total pro gram hours devoted to public affairs pro gram ming. Of course, any deci sion to impose such regulations must also be weighed against the finan cial and pro gram ming burdens such obligations might place on dig i tal broad cast ers as they attempt to develop viable business mod els for this emerg ing media technology for which consumer demand is uncer tain, at best. In addi tion, pol icy mak ers deci sion making on this issue should also take into con - sideration the supply and demand dynam ics of pub lic affairs pro gram ming. That is, to what extent does an increase in the supply of available public affairs pro - gram ming lead to increased viewership of such pro gram ming? Is it possible that audi ence demand for public affairs pro gram ming is being sat is fied effectively by currently available quan ti ties? Answering such questions requires additional data (namely, ratings data) beyond what was gathered for this analysis. However, investigations of such questions are vital to informed deci sion making on this issue.

14 28 Press/Pol i tics 6(2) Spring 2001 Notes 1. It should be noted that there is much skepticism in many circles as to whether the tran si tion to dig i tal tele vi sion can be achieved in this relatively short time period (see Advi sory Com - mit tee on the Public Interest Obligations of Dig i tal Television Broadcasters 1998). 2. It would have been preferable to com pile a two-week sample by randomly select ing days from through out the broadcast year. However, given that this research was con ducted in response to the Federal Communications Com mis sion s Notice of Inquiry (NOI), and that there was only a three-and-a-half-month window from the announcement of the NOI to the dead line for fil ing comments, time limitations required the use of a less systematic sampling procedure. References Advisory Committee on the Public Inter est Obligations of Digital Television Broad casters Charting the Digital Broadcasting Future. Wash ing ton, DC: Gov ern ment Printing Office. Baker, C. Edwin Giving the Audience What It Wants. Ohio State Law Journal 58(2): Bal anced Bud get Act of 1997, Pub. L. No , Sec tion 3003, 111 Stat. 251 (1997). Benton Foun da tion What s Local about Local Broadcasting. Available: org/tele vi sion/whatslocal.html. BIA Research Investing in Tele vi sion Market Report. 3rd. Edition. Chantilly, VA: BIA Research. Brennan, Tim o thy J Economic Effi ciency and Broadcast Con tent Reg u la tion. Federal Communications Law Jour nal 35(2): Busterna, John C Ownership, CATV and Expenditures for Local Television News. Jour - nalism Quarterly 57: Communications Act of 1934, Pub. L. No. 416, 48 Stat (1934). Duggan, Ervin S Dig i tal Audio Broadcasting: Getting It Finished and Getting It Right. Remarks before the Annenberg Wash ing ton Program, April 11. Federal Communications Com mis sion Amend ment to Section of the Commission s Rules: Delegations of Authority to the Chief, Broad cast Bureau, 59 FCC 2d 491. Federal Communications Commission Revision of Programming and Commercialization Pol - icies, Ascer tain ment Requirements, and Program Log Require ments for Commercial Tele vi sion Stations, 1984 FCC LEXIS Federal Communications Commission. 1999a. The Public and Broad cast ing. Report prepared by the Mass Media Bureau. Washington, DC: Fed eral Communications Com mis sion. Federal Communications Com mis sion. 1999b. Public Inter est Obligations of TV Broad cast Licensees, 1999 FCC LEXIS Fowler, Mark S., and Brenner, Daniel L A Marketplace Approach to Broad cast Regula - tion. Texas Law Review, 60:1-51. Horwitz, Rob ert B The Irony of Regulatory Reform: The Deregulation of American Telecommuni - cations. New York: Oxford Uni ver sity Press. Kunkel, Dale Pol icy Battles over Defining Children s Educational Television. Annals of the American Academy of Political and Social Sci ences 557: Lacy, Stephen, and Riffe, Dan iel The Impact of Competition and Group Own er ship on Radio News. Journalism Quarterly 71(3):

15 Moonves, L State ment before the Open Meet ing of the Advi sory Com mit tee on the Pub - lic Interest Obligations of Dig i tal Television Broadcasters, Washington, DC, April 14. Avail - able: Napoli, Philip M A Principal-Agent Approach to the Study of Media Organizations: Toward a The ory of the Media Firm. Political Com mu ni ca tion 14(2): Napoli, Philip M. 1999a. Decons truct ing the Diver sity Principle. Journal of Communication 49(4):7-34. Napoli, Philip M. 1999b. The Marketplace of Ideas Metaphor in Communications Regulation. Journal of Communication 49(4): Napoli, Philip M. 1999c. The Unique Nature of Communications Regulation: Evidence and Implications for Communications Pol icy Analysis. Jour nal of Broad cast ing & Elec tronic Media 43(4): Napoli, Philip M The Local ism Prin ci ple under Stress. Info 2(6): Napoli, Philip M Foundations of Communications Pol icy: Prin ci ples and Pro cess in the Regulation of Electronic Media. Cresskill, NJ: Hampton Press. Red Lion Broadcasting Co. v. Fed eral Communications Com mis sion, 395 U.S. 367 (1969). Riffe, Dan iel, Ste phen Lacy, and Fred er ick G. Fico Analyzing Media Messages: Using Quanti - tative Con tent Analysis in Research. Mahwah, NJ: Lawrence Erlbaum Associates. Telecommunications Act of 1996, Pub. L. No , 110 Stat. 56 (codified at 47 U.S.C. 151 et. Seq.) (Feb. 8, 1996). Note This project was conducted with funding from the Benton Foundation Biographical Note Napoli / Digital Tele vi sion Policy 29 Philip M. Napoli is an assis tant professor in the Graduate School of Business Admin is tra tion at Fordham University in New York. He is the author of Foundations of Communications Pol icy: Prin ci - ples and Pro cess in the Reg u la tion of Elec tronic Media. Address: Graduate School of Busi ness Administration, Fordham Uni ver sity, 113 West 60th Street, New York, NY 10023; phone: ; pnapoli@fordham.edu. Paper sub mit ted May 26, 2000; accepted for publication Octo ber 23, 2000.