378.10111 N867U cop. 2 ■■hwimi^ James D. Nowlan, et al. The University "of Xllinois Trustees: "Invisible" "" Statewide Candidates? (1984) flLTHCIS HISTORICAL SURYH UNIVERSITY OF ILLINOIS LIBRARY ATURBANA-CHAMPAIGN ;li.-./T:rti-INSfTltUTE..OF GOVERNMENT AND PUBLIC AFFAIRS s^ ? V-t r.. •.^, Occasional Papers in Illinois Politics Number 5 THE UNIVERSITY OF ILLINOIS TRUSTEES: "INVISIBLE" STATEWIDE CANDIDATES? UNIVERSITY OF ILLINOIS UNI/Z-CIT/ OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN STACKS THE UNIVERSITY OF ILLINOIS TRUSTEES: "INVISIBLE" STATEWIDE CANDIDATES? Oames D. Nowlan Christopher 0. Ross Mildred A. Schwartz September 1984 THE UNIVERSITY OF ILLINOIS TRUSTEES: "INVISIBLE" STATEWIDE CANDIDATES? By James D. Nowlan, Christopher 0. Ross and Mildred A. Schwartz* The University of Illinois is one of the few public universi- ties with an elected board of trustees (National Commission 1980). Every two years since 1889 each political party has run three candidates for three openings; those elected serve stag- gered six-year terms. Trustee candidates are held to be "invisi- ble" because the responsibilities of office and the candidates who run for them are generally unknown. The prevailing impression is that "invisible" candidates create "depersonalized" races (MacRae and Meldrum 1960, 673), in which votes are cast "almost entirely on the basis of party loyalty" (Monroe 1982, 159) and thus are "the truest indicators of party power" (Colby and Green 1983, 18). Because of these generally held views, political analysis and candidates have consistently used the trustee results to measure underlying * James Nowlan is director of graduate programs in public administration, Department of Political Science, University of Illinois at Urbana-Champaign . Christopher Ross is an assistant professor and Mildred Schwartz a full professor, both in the Department of Sociology, University of Illinois at Chicago. The authors wish to acknowledge the assistance of Antje Kolodziej for her help in collecting data for this project, Shirley Burnette for cheerfully keeping the word processor humming throughout all the changes we made in the manuscript, and Anna Merritt for editing and re-editing it all. partisan strength for such purposes as legislative redistricting and the allocation of campaign funds. This conventional view received additional credibility and significance in 1981 when it was cited by a federal court panel in support of the congres- sional redistricting plan selected by the court: An "approximation of statewide political strength" in Illinois can be made, according to the experts, by looking at the results of the University of Illinois Board of Trustees elections. These results indicate that Illinois is a "swing" state, that it is split roughly fifty-fifty between the major politi- cal parties. Projections based upon these same results further indicate that congres- sional elections based upon the Otto maps will most closely result in an Illinois delegation mirroring this partisan division. Thus, if Gaffney is applicable to federal congressional redistricting matters, the Otto plans are preferrable as they best advance the goal of "political fairness" (U.S. District Court, Northern District of Illinois, 1981, 22). The elections of 1978 and 1980 focused further attention on the trustee vote. Both produced a party split among the three trustees elected, a result not found since 1930. Since trustees had been listed on the ballot in fixed order, and the Republican candidate at the top of the list was the only trustee of his party to be elected, it was surmised that he had benefited from this auspicious position. As a consequence, it was decided in 1980 to rotate, by congressional district, the listed order of trustees. The outcome was another split, but this time two to one for the Republicans. Are trustee candidates then totally invisible? Do trustees deserve to be treated uncritically, as the soundest measure of partisan support, unaffected by other factors? If not, what affects their stability as such measures? Finally, what do the trustee elections in the two aberrant years of 1978 and 1980 tell us more generally about the political process in Illinois? To answer these questions and to test some of the underlying assumptions about the relative visibility of trustee candidates, we looked at two categories of issues relating to the trustees, especially those running in the important years 1978 and 1980. First, we studied differences among trustee candidates, using a variety of factors both over time and during any one election. Second, we compared trustee candidates with candidates for other offices. ■2- DIFFERENCES AMONG TRUSTEE CANDIDATES Certainly the most significant difference among the candi- dates is how each fared in his or her election. Others factors that we looked at are ballot position, the friends and neighbors effect, how they approached the campaign itself, the impact of newspaper endorsements, and the personal characteristics of the individuals themselves. Election Results An obvious conclusion from the 1978 and 1980 elections is that each party's trustees were not alike in the support they received. For one thing, incumbency was no protection; two of the three incumbent trustees were defeated in 1978 and 1980. More importantly, as Table 1 shows, there was as little as 0.3 percent difference between the second and third ranking Republi- cans in 1978 and as much as 11.7 percent difference between the first and third ranking Democrat in 1980. That kind of variation is not a new phenomenon. In 1976, for example, the percent difference between the winning Democratic candidates who ran first and third was 11.3 percent. On the losing Republican side, the top and bottom candidates were separated by a difference of 9.7 percent. Within each party, the three candidates won support from at least slightly different constituencies. This had not previously been given much attention. MacRae and Meldrum tell us, in a footnote, only that they have "adjusted to reduce effects of local popularity of one trustee candidate in a county" (1960, 673). Monroe states that "the similarity of vote totals for all three candidates from the same party indicates that there must be almost no attempt to distinguish individuals" (1980, 159). In the court hearing on redistr ict ing cited earlier, the primary document used was based on the second ranking candidate for trustee in each party, as though this took care of any relevant variation. Ballot Position One way to account for the differences among trustees is to attribute them to position on the ballot (Montgomery, 1983). It was just this concern with the advantage to first position that led to the decision to rotate each party's candidates on the 1980 ballot. We surmise the effects of ballot order on voters by comparing smaller geographic units within the state. Counties are suitable for this, but because they are highly variable in size, aggregate results may otherwise obscure positional advan- tage. We overcome this by transforming votes in each county into candidate rankings. Table 2 reveals that such advantage was present only for candidates in first position in 1978. -3- Table 1 Trustee Election Results: 1978 and 1980 Republican: Votes Democratic: Votes 1978 Hahn (W) 1,324,857 Stone (W) 1,409,086 Rader 1,285,296 Donoghue (W) 1,348,011 Livingston 1,281,284 Webb 1,302,741 % Difference (Hi-Lo)=3.3 % Difference (Hi-Lo)=7.5 1980 Madden (W) Day (W) Gougler % Difference 2,009,945 1,914,231 1,786,141 (Hi-Lo)=11.1 'Shepherd (W) Lenz Velasquez 1,995,637 1,840,099 1,761,259 ?o Difference (Hi-Lo)=11.7 Source: State Board of Elections, State of Illinois Official Vote, 1978 and 1980. -4- Table 2 Relation Between Ballot Position and Electoral Support, 1978 Democratic Ballot Position 1 Stone 2 Donoghue 3 Webb Republican Ballot Position 1 Hahn* 2 Livingston* 3 Rader Rankings within Counties First Second Third Note: Numbers refer to counties in which each candidate ranked first, second, or third in total votes. N=102 * Hahn and Livingston have one tie for first place. 78 23 1 2 19 81 22 60 20 69 16 16 17 41 43 15 44 43 -5- On the Republican side, although Rader was listed third on the ballot, she in fact was second in overall votes. But in terms of rank order within counties, her pattern was almost identical to that of the less successful Livingston. On the Democratic side, Donoghue was both second in position and in overall votes. Yet converting votes to rank orders shows that he was most frequently in third place. The contrast between within-county order of preference and overall outcome was largely due to his strong performance in Cook County, where almost half of all the votes in the state are cast. The rotation used in 1980 to overcome positional advantage was a limited one. Candidates were each moved twice, for a total of three rotations out of a possible eight. Moreover, the order of the two parties was kept constant, so that the same candidates of the two parties were always paired. This introduces some possible bias, since Velasquez, who did most poorly among the Democratic trustees, was always matched with the front-running Republican, Madden. Similarly, the best performing Democrat, Shepherd, always had the poorest performing Republican, Gougler, as her partner. This limited rotation was done on the basis of congressional districts, not counties. Because of the overlap in the case of Kane, Lake, and Montgomery counties, and the extension of Cook County over thirteen congressional districts, we eliminated these counties from our consideration. Adams and Bureau counties were retained in the analysis, although a small portion of each is contained in congressional districts with a different trustee order. The uneven fit between counties and congressional districts also meant that each pattern of trustees was not equally represented in our comparison. Table 3 shows the relation between ballot position and ranking by voters within counties taking rotation into account. For example, on the Republican side, the candidate listed first was ranked first in thirty-nine counties, second in twenty, and third in thirty-nine. We see that rotation coincided with a reduction in the degree to which ballot position determined ranking. Position remains important, but it is clearly not the whole story. Friends and Neighbors Some the differences among trustees are due to a "friends and neighbors" effect, as is clearly demonstrated by the relatively better showing of candidates in their own county of residence (see Table 4). In fact, in both 1978 and 1980, each candidate enjoyed some advantage in his or her local area. This differen- tiated them from the trustee leading the state ticket, or when that was not an appropriate comparison, the next most successful Table 3 Relation Between Ballot Position and Electoral Support, 1980 Rankings within Counties First Second Third Democratic Ballot Position 1 2 3 51 10 37 12 40 46 35 48 15 Republican Ballot Position 1 2 3 39 49 10 20 34 44 39 44 44 Note: Position and rank were calculated for subsets of counties according to rotation. Cook, Lake, Kane, and Montgomery counties were omitted. N=98 Table 4 Trustee as Local Favorite, 1978 and 1980 Candidate Home County Ord Pre er of ference Proportionate Advantage Over Leader* In Home In State County 1978 Dam. Stone Moultrie S>W>D + .52 + .04 Donoghue Cook D>S>W <+.01 -.09 Webb Richland S>W>D -.03 -.08 Rep. Hahn Sangamon H>R>L + .19 +.03 Livingston Cook H>L>R -.01 -.05 Rader Union R>L>H +.63 -.03 1980 Dem. Shepherd Cook S>V>L + .11 +.06 Lenz McLean L>S>V + .14 -.08 Velasquez Cook S>V>L -.06 -.19 Rep. Madden Macon M>D>G +.40 +.05 Day Boone D>M>G +.10 -.05 Gougler Cook M>D>G -.10 -.12 * Leader is the top trustee vote-getter. In his/her own county, he/she is compared to the second place trustee for the same party in the state. Home county effects have been removed from the results reported for the state, candidate in the state, by as much as 63 percent for Rader, 52 percent for Stone, and 40 percent for Madden. In all instances, and even for relatively weak trustees, local candidates did better locally than they did in the state as a whole. The appeal of a local candidates, when it is strong, is not confined to the ability to gain more support than the other trustee candidates. In 1978, for example, Rader obtained more votes than any other Republican running in her small county of Union with the exception of one immensely popular state repre- sentative. Both she and that state representative were able to win this support in what is otherwise a Democratic county. Stone, running in a largely Republican county, outdrew his party's candidates for other offices in seven out of nine races. He trailed only the secretary of state, the Democratic ticket leader in the state, and one state representative. In Cook County the Democratic trustee candidate Donoghue led his party's ticket in comparison to five other offices where the Republicans had an advantage statewide, but he did less well compared to other Democrats running for the remaining offices. In 1980, with fewer offices to compare, there is also less evidence of the drawing power of local candidates. The clearest example is Day, who, in strongly Republican Boone County, outdrew her party's candidates for president, U.S. senator, state senator, and one state representative, but not the other state representative or congressional candidate. Another possible manifestation of the friends and neighbors effect shows up beyond the boundaries of candidates' home counties, in adjacent ones. Rader led her party's trustees in eight contiguous counties in addition to her own, Donoghue in two additional, Day in one, and Lenz in two. In addition, although Webb did not lead in his own county of Richland, he was his party's leading trustee candidate in eleven contiguous counties. The friends and neighbors effect found for 1978 and 1980 has also been present in other years. From 1970 to 1976, out of twenty-four possible comparisons, some advantage was experienced by twenty trustees in their home counties. In 1982 this advantage was enjoyed by four out of six trustees. Campaigning Anderson (1978) has suggested that trustee candidates are relatively unknown because they do not campaign. This is only partly true. Some trustee nominees are rocking-chair candidates who eschew personal campaigning; others campaign extensively, although none has yet spent more than about $10,000 in his or her own behalf in a single election. Forsyth estimates he expended $8,000 in 1982; Hahn, about $2,000 in his last campaign. As Table 5 shows, the place of money in trustee campaigns appears minor when compared to other statewide races. Trustee William D. Forsyth, Democrat of Springfield, ex- plained his views on campaigning for the trustee position by saying: Yes, I do campaign. I feel it is my responsi- bility. On the other hand, I have had running mates who felt strongly they should not do so, that it was an honor bestowed....! campaign because the elections are often close, and you can make a difference. For example, in 1976 only 2,500 votes separated winner Howard (a Democrat) from loser Swain (a Republican). That is only one vote difference per every four precincts. Trustee Ralph Hahn, Springfield Republican, also campaigns actively. He notes: Hahn distributed more than 100,000 imprinted postcards in 1978, personally handing out many of them at Chicago train stations, Illinois football games, and political party meetings. In his opinion, "I think this must have been worth 10,000 or so votes" ( Alumni News , 1978). Those candidates who do campaign, typically have postcards or simple brochures in hand. Forsyth had 30,000 printed in 1982, 15,000 listed all Democratic candidates, and 15,000 were without partisan identification of any kind. "I do this because in many ways the office is nonpartisan and people will vote for me regardless of party." For one election, Forsyth, who had been elected trustee three times, sent letters to approximately 6,000 elected Democratic precinct committeemen. In 1982 he placed advertisements in about eight downstate daily newspapers the Sunday before the election, Hahn estimates he spent ten to twenty days campaigning each of the three times he has been a candidate: I try to get to some of the party events, to show party people I'm out there trying. They -10- Table 5 Campaign Expenditures in General Elections 1978 U. S. Senate Percy Seith R D $2,000,000* 1,300,000* Governor Thompson Bakalis R D 1,936,053 1,064,937 Attorney General Scott Troy R D 449,630 119,000 Secretary of State Sharp Dixon R D 51,642 1,344,435 Comptroller Castle Burris R D 182,343 195,965 Treasurer Skelton Cosentino R D 128,206 184,790 U. of I. Trustees 1980 Hahn Livingston Rader Stone Donoghue Webb R R R D D D 5,983 2,315 2,058 ** ♦* X-* U.S. Senate O'Neal Dixon R D 1,270,650 2,007,409 U. of I. Trustees Madden Day Gougler Velasquez Lenz Shepherd R R R D D D ** 2,999 1,330 ** * Estimates based on Federal Election Commission Reports. ** Candidates are not required to file reports unless they accept or expend more than $1,000 in any twelve-month period. Source: Illinois Board of Elections, Springfield, Illinois -11- appreciate this, and it can be helpful later when you need to talk to legislators about a U. of I. problem. For one campaign, Forsyth put 10,000 miles on his car, "toured Chicago six to eight times," plus Rockford, the St. Louis area, and many downstate cities. Forsyth notes that, "As major candi- dates spend most of their post-Labor Day time in the metro Chicago area, I try to fill in at party meetings downstate, in Petersburg, Mt. Vernon, or wherever." There is no agreement on how much effect campaigning has, even in the best-studied presidential campaigns, although the implication is that campaigning must be important (Asher 1980, 316-318). Difficulty in separating the campaign effects from other conditions and factors (e.g., Kessel 1980, 167-170) makes it impossible to evaluate the efforts of trustee candidates like Hahn and Forsyth. At the least, however, we can conclude that some candidates are not satisfied to be treated as indistinguish- able members of an "invisible" group. Endorsements If differentiation from other trustees is a goal, obtaining newspaper endorsements may be worth a candidate's efforts. Trustees Forsyth and Hahn feel strongly that such endorsements are the single most important variable they can affect person- ally. In 1978 Hahn took a three-day trip to visit newspaper editors in Chicago, the suburbs, and Rockford. According to Hahn, "Every newspaper that endorsed supported me, to my knowledge, and the people who interviewed me were generally quite knowledgeable about the issues." In 1982 Forsyth talked with editors at each of almost fifty newspapers with over 20,000 in circulation. "The problem is that only about eight newspapers will endorse," said Forsyth. Publicity is not absent, however, and nearly all newspapers will carry short background profiles of trustee candidates in their pre-election coverage. The effects of endorsements on voting behavior remain unclear (Scarrow and Borman, 1979, 388-394; Hooper, 1969, 302-305). Unlike other campaign influences, whose effect can occur in ways that are not directly perceived by voters (e.g., Iyengar, Peters, and Kinder, 1982, 848-858), endorsements are deliberate efforts to shape choices and they succeed to the extent that they are recognized as such. Moreover, and probably more important, the significance of endorsements lies in their impact on the candi- dates themselves, financial contributors, and party activists, since they validate the claim of candidates to be treated seriously. -12- Our efforts to discern the impact of newspaper endorsements on votes for trustees were limited to five newspapers in 1980 and four in 1982. All relevant endorsements in 1978 were for Republicans, so there is no variation to examine. In eleven of the twelve cases examined in 1980 the endorsed candidate fared better than his or her running mates in the relevant county than was true statewide. For 1982 this was so in six of ten in- stances. Comparing Trustees All the variations among trustee candidates that we have looked at so far are sufficient to warrant that they not be treated as mere carbon copies of one another. Variation is partly the result of ballot position, but it also comes from the ability of individual candidates to draw support in their own and neighboring counties. While campaigning may be relatively modest, the distribution of votes for trustees in some counties suggests that the efforts of the candidates themselves, aided by newspaper endorsements and the help of some constituents (e.g., concerned University of Illinois faculty) may even buck a trend against their party. This may have worked in 1980, when there was a noticeable concentration of votes for Democratic candidate Shepherd in some counties. She pulled ahead of the second ranking Democrat by 23 percent in Champaign County, 17 percent in DuPage, 18 percent in Lake, 24 percent in McHenry, and 19 percent in Peoria — all normally Republican areas. In 1978 this kind of discrepancy occurred in Champaign where Democratic candidate Stone, apparently aided by his recognition as a former state representative, exceeded the vote for his next running mate by 74 percent. The strong performance by candidates Rader and Shepherd and the weak one by Velasquez may suggest that women candidates generally do well, while those of other than Northern European origin do poorly. Our particular findings cannot be so easily generalized, however, since Velasquez was the only candidate we could find not from the dominant ethnic groups, while, at least during the 1960s, women did less well than their male counter- parts. RELATION TO OTHER OFFICES The extent to which votes for trustees can be viewed as an indication of anything more than a preference for one trustee candidate over another, lies in the relationship between the votes cast for them and the votes cast for other offices. To determine the significance of any such relationship we looked at three issues: 1) How many people vote a straight party line, a -13- split ticket, or not at all, and how have these numbers changed over time? 2) Do parties play a different role in trustee campaigns than in campaigns for other offices? and 3) How do the outcomes of the elections for trustees compare with those for other offices? Straight, Split, or Pass If partisanship is reflected in straight party voting, then the trustees may be treated as one of several interchangeable party offices. Straight party voting, in turn, can be aided or inhibited by the form of ballot that is being used and how the voting machines are arranged (Walker, 1966, 443-463; Sorauf, 1980, 235). Illinois law gives some latitude to the 111 local election units (102 counties plus 9 municipal commissions) in selecting either paper ballots or types of voting machines. But regardless of the type of ballot used, the state requires that there be a place at the top--or the beginning of a general election ballot — where a voter may make a single mark (or stylus punch or lever pull) to cast a straight ballot for all candidates of a single party. In this respect, Illinois would appear to aid straight party voting. Since the early 1970s, however, when the state moved from the paper ballot and the large voting machines that display a pull-down lever for each party and each candidate, to small computer-based punch card devices, ticket splitting may have been given some encouragement. The paper ballot and the large voting machines display all candidates of a party in a single column or row; the columns or rows are headed by a circle or lever by which one may cast a straight party ballot with a single X-mark or lever pull. The punch card devices present a kind of small loose-leaf notebook to the voter, with a computer punch card immediately underneath. The first page of the notebook provides the straight party option. The pages that follow list candidates by office blocks, that is, all general election candidates for president are listed in a block, then all candidates for U.S. senate in a block, and so on down to the local offices. Each candidate name in each block is identified separately with his or her party affiliation. In other words, the punch card devices may en- courage ticket splitting, or the utilization of cues other than, or in addition to, that of political party, because candidates are grouped by office block rather than by party list. By 1974, when the State Board of Elections was established, most election units were utilizing the punch card devices. Cook County adopted a punch card device in 1976, and Chicago did so in 1982. As of 1983 only five small counties still used paper ballots (Fayette, Hamilton, Saline, Washington, and Warren), while only Kankakee -14- County and the city of East St. Louis still used the large voting machines. Some influence might also be expected from appeal exerted by the top of the ballot, in a manner suggested by MacRae and Meldrom: "The trustee vote may reflect a partial association of straight-ticket voting with local traditional party identifica- tion, even though the straight ticket is also affected by the personality at the head of the ticket" (1960, 677). In general, the impact of the top-of-ticket candidate is also confirmed by Campbell and Miller (1957, 293-312), from whose research we can estimate that anywhere from one to 10 percent of their sample of straight party voters may have cast straight ballots because of the appeal of the top of the ticket. One way to determine the effect of the top of the ticket on straight ticket voting is by averaging the three trustee votes from 1942 to 1982 (excusable in this instance because of the need to compare) and comparing the trend line with that of the top of the ticket. Figure 1, based on the percent of the two-party vote going to the Republicans, indicates that voting for trustees moved in a relatively narrow range, within the 45th to 55th percentile, for all but two elections. Voting for the top of the ticket and for trustees generally stayed close together until 1970, with the exception of the presidential years of 1956 and 1964, when unusual races contributed to above-average ticket splitting. Since then the discrepancy between offices has widened, suggesting, and confirming other evidence collected from ballots, that straight ticket voting has declined. In Republi- can-oriented Champaign County, straight ticket voting declined from 48 percent of ballots cast in 1968 to 21 percent in 1976, and rose to 30 percent in 1982. In another central Illinois county. Democratic-oriented Macon, straight voting declined from 35 percent of all ballots in 1970 to 18 percent in 1978, and back up to 33 percent in 1982. What is also apparent is that the presence of a popular candidate at the top of the ticket now actually encourages ticket splitting. Whatever the pressure for a consistent partisan vote may be, it is unclear how the trustees contribute to it. In fact, if the party indicator at the beginning of the ballot is not marked, there is some danger that voters will simply skip the trustees. Voter fatigue has been estimated as the cause of voting drop-off of from 20 to 30 percent (Sorauf, 1980, 235; Walker, 1966, 448-463). Use of punch card devices with pages of candidates, now adopted in most election districts in the state, would appear to encourage such voter fatigue. For Illinois, Colby and Green (1983, 19) state that by 1980, 20 percent of those who took a ballot did not vote for trustees. -IS- •a "O c a O c « 1 <» 3 0)_- c ^ o o 0) CO o ^p O H- 5 o I- a <^ o O F 00 o 00 00 / / / / / /- y 1 1 1 1 r \ \ \ \ \ \ to ■-\ __ '"^ ^- X „. CM ■* ->. .^\ N- O y /' 00 to t/ "*■ ^*w^ ■* "*• CO CVJ to o \ \ \ \ I \ \ \ \ \ \ \ (O 00 lO (O .^'^ -;> . ^^^ X ,/ ^ CM i o \ \ 00 X' '^ (O V, ;-, /^ CM 35 i X> < 3? D 1 £> 1 /) < ti 1 35 3 1 5^ < o < a < "0 < (0