INTERPARTY COMPETITION IN , 1954-68: A HISTORICAL AND POLITICAL PERSPECTIVE

By ROBERT L. HEALY*

HE dynamics of Pennsylvania politics have been examined T from numerous vantage points. A survey of the literature would incorporate historical descriptions of significant state political events,. biographical studies of leading political per- sonalities, 2 journalistic accounts of machine power politics,3 studies of political party organization,4 and treatises concerning the Gen- eral Assembly-its participants, its operation, and its relationship with other governing institutions. 5 One aspect of Pennsylvania politics has held a continuing interest for political scientists-the competition for votes between the two major political parties. Scholarly analysis has usually focused on county voting patterns; these studies had their genesis in the 1940's with the publication of a pamphlet series that researched imterparty county competition from 1922-1944.' These studies and *The author lives in , Pennsylvania. ' See Sylvester K. Stevens, Pennsylvania History in Outline (Harrisburg, 1960). Also, Norman B. Wilkinson, Bibliography of Pennsylvania (Harris- burg, 1957), Pennsylvania Manual, Vol. 97, 63. 2For a summary of state bosses, see E. F. Cooke and G. E. Janosik, Guide To Pennsylvania Politics (New York, 1957), 6-19. Also, Frank Hawkins, "Lawrence of Pittsburgh: Boss of the Mellon Patch," Harper's Magazine, LV (August, 1955). ' Beverly Smith, "How to House Clean a State," Saturday Evening Post, April 14, 1951, 38-39. 'Frank J. Sorauf, Party and Representation (New York, 1963). 'William Keefe, "Parties, Partisanship, and Public Policy in the Penn- sylvania Legislature," American Political Science Review, XLVIII (June, 1954), 452. Kenneth T. Palmer, "The Pennsylvania General Assembly," Pennsylvania Assembly on State Legislatures in American Politics (Pitts- burgh, 1968). Malcolm Jewell, "Party Voting in American State Legis- latures," American Political Science Review, XLVIX (September, 1955), 773-779. Thomas Dye, "A Comparison of Constituency Influence in the Upper and Lower Chambers of a State Legislature," Western Political Quarterly, XIV (June, 1961), 473-480. 'H. F. Alderfer and Robert Sigmond, Presidential Elections by Pennsyl- vania Counties, I920-I940 (State College, Pennsylvania, 1941). See also, H. F. Alderfer and Fannette H. Luhr, Gubernatorial Elections in Pennsyl- vania, 1922-1942 (State College, Pennsylvania, 1946). 352 INTERPARTY COMPETITION IN PENNSYLVANIA 353 a subsequent article published in PENNSYLVANIA HISTORY' at- tempted to answer such questions as: In which counties are the parties most competitive? In which counties are the parties least competitive? In which counties are the Democrats strong? In which counties are the Republicans strong? This article is an updating of the previous works through the 1968 elections in order to give a current perspective on the status of interparty competition in Pennsylvania. In any analytical scheme, the validity of the findings rests on key operational decisions. Decisions crucial to this study are (1) the time span to examine, (2) selection of electoral offices around which to (3) construct indices of competition and (4) establish categories for county competiveness classification. For two reasons, the interval from 1954 to 1968 was selected for analysis: First, earlier studies of competition in Pennsylvania counties had covered the years 1922 to 1944 and 1944 to 1958. A study covering the years 1954 to 1968 was a logical extension of the other works. Second, 1954 seemed to 'have been a turning point in the political climate of the state. In that year, and for the first time since 1934, Democrats captured a major political office-the governorship. Two years later, Democrats managed to wrestle a senatorial post away from the Republicans; then, in 1958, Democrats retained control of the governor's chair. A second decision is that of the contests to be studied. The earlier studies of county electoral behavior have employed in various combinations presidential, senatorial and gubernatorial contests; yet, there is a question whether or not all of these con- tests would yield an accurate representation of political competition in the counties. The assumption is that different races and offices affect the electorate in different manners. Examining only the presidential contest, for example, might tend to skew the findings and accent oddities associated with high stimulus presidential elections. This study chose to examine all in-state, statewide races be- tween 1954 and 1968. Elections used were senatorial, four gubernatorial, four races for Secretary of Internal Affairs, four races for State Treasurer and State Auditor General.

'Edward F. Cooke, "Patterns of Voting in Pennsylvania Counties, 1944- 58," Pennsylvania History, XXVII (January, 1960), 69-87. 354 PENNSYLVANIA HISTORY

The third decision is designing operational indices. Two types of indices are required: dominance and competition. To construct a measure of party dominance, the following county-row electoral offices were examined: Commissioner, Treasurer, District Attorney, Sheriff, and Coroner. These offices were selected because they were the only ones, during the time period under study, that were elected in every county. As a quantitative measure of party dominance, this paper employs an index of percent offices won; a political party is classified dominant, senmi-dominant, or non- dominant in a county according to the following scale:

Dominant - Won over 80o of offices listed Semi-Dominant - Won from 60 to 79.9% of offices listed Non-Dominant -- Won from 50 to 59.9%o of offices listed

The Party Dominance Index (PDI) has two functions. It com- plements the Party Competition Index (explained below) to in- crease the descriptive power of the study; and, it will be used to break mathematical ties in the construction of the Party Com- petition Index (PCI). In the literature, three main variables have been used to con- struct Party Competition Indices (PCI) : alternation of parties in office, percentage of elections won, and percentage of votes won.' This article takes the position that any measure of com- petition calculated from percent elections won or percent terms in control or a rank ordering of election results is primarily a function of votes cast. This study will calculate the Party Com- petition Index from percentage of votes won over all in state- wide races.9 The last decision is that of category specification. The following categories were chosen because the percentages represent the most natural breaks among the clustering of total mean pluralities:

S Austin Ranney and Wilmoore Kendall, "The American Party Systems," American Political Science Review, XLVIII (June, 1954), 477-485; Joseph Schlesinger, "A Two Dimensional Scheme for Classifying States According to the Degree of Interparty Competition," American Political Science Re- view, XLIX (December, 1955), 1120-1128; Robert T. Golembiewski, "A Taxionomic Approach to State Political Party Strength," Western Political Quarterly, XI (December, 1958), 494-513; Richard Hofferbert, "Classifica- tion of American Party Systems," Journal of Politics, XXVI (August, 1964), 550-567. 'David G. Pfeiffer, "Measurement of Interplay Competition and System Stability," American Political Science Review, LXI (June, 1967), 457. INTERPARTY COMPETITION IN PENNSYLVANIA 355

Plurality Percent Categories for Winning Party Strong Two-Party County 0 to 4.9 Weak Two-Party County -- 5 to 7.9 Weak One-Party County 8 to 11.9 Strong One-Party County 12 to 19.9 Exceptionally Strong One-Party County 20 and over Thus a county is classified as a strong two-party county-com- petition between parties is at a peak-when a plurality of two-party vote is less than five percent for the winning party. Yet, in terms of partisan status, these categories are not too meaningful; they do not, for example, tell whether or not a county is Democratic or Republican. Hence, a modification of the above index was devised: Plurality Percent Categories for Winning Party One-Party Democratic County Over 20 Modified One-Party Democratic County 8 to 19.9 Weak One-Party Democratic County ------5 to 7.9 Two-Party County ------0 to 4.9 Weak One-Party Republican County -- 5 to 7.9 Modified One-Party Republican County - 8 to 19.9 One-Party Republican County --- - Over 20 A county is classified as One-Party Democratic if the plurality between the two parties is twenty percent or over and if the winning party is Democratic. Ties in the rankings of counties on the PCI index were broken *on the basis of the dominant party (PDI) in the county. If, for example, after all the statewide races had been scaled, Clarion and Clearfield were to tie as to PCI, then the tie would be broken by looking at the party relationships within the county. If Clear- field were a dominant Democratic county and Clarion a semi- dominant Republican county, the tie would be broken by awarding Clarion the more competitive position. Counties were then classi- fied according to natural breakings in the pluralities.

* * * * * Historically, electoral politics in Pennsylvania has been non- competitive. Pennsylvania has long served as a loci of Republican- ism; the Democratic party has been benignly dormant. In the last thirty to forty years though, statewide interparty electoral 356 PENNSYLVANIA HISTORY competition has markedly increased. Yet, even during the Roose- velt years Pennsylvania retained much of its Republican orienta- tion, electing only one Democratic governor between 1932 and 1954. During this period, the Democratic party began to secure its hold on the industrial, blue collar working counties of Allegheny, Fayette, Greene, Washington, Westmoreland, and Lackawanna. Before the Democratic party presented a serious challenge to Republican superiority, though, it was necessary to obtain control of populous . With the election of Joseph Clark as Mayor of that city, the Democratic party enlarged its leverage in the state politics.10 Cooke has called the time period of 1944 to 1958 the "Transient" period of Pennsylvania politics.", As evidenced in Table One, it would be quite appropriate to call the period from 1954 to 1968 as the "competitive period":

TABLE ONE

1954-1968, NUMBER OF OFFICES WON BY MAJOR PARTY

Senator Governor Statewide Republicans 3 2 3 Democrats 2 2 9

Total Possible Offices 5 4 12

In 1954, the Democratic party captured the governorship for the first time since 1934. George Leader won election through a combination of factors, one of which was a split within Repub- lican ranks. In 1956, Eisenhower running for a second term carried the state; but the budding urban support for the Demo- cratic party offset much of his coattail strength and Mayor Jo- seph Clark of Philadelphia, a Democrat, won the U. S. Senate seat over two-term Senator and former Governor Edward Martin. In the 1958 final electoral tally, Governor Leader was defeated by Republican for the Senate seat while Mayor David Lawrence of Pittsburgh, a Democrat, won the Governor's chair. In 1962, the Republicans won the governorship with beating Richardson Dilworth of Philadelphia. How- "°John Bryant, "Political Stability and Change in Pennsylvania," unpub- lished seminar paper, , 1968. "Cooke, "Patterns of Party Competition, 1944-5&," 81. INTERPARTY COMPETITION IN PENNSYLVANIA 357

ever, even in the face of a Scranton gubernatorial victory, Clark again demonstrated his appeal to the urban areas and won re- election to his Senate seat. The 1964 Presidential race was a landslide for Lyndon Johnson; he carried all but three of Pennsylvania counties,-an unheard of feat for a Democrat. Once more though, the electoral picture was selective-incumbent Scott beat Democrat Genevieve Blatt for the U. S. Senate. In 1966, the Republicans retained their hold on the capitol with Raymond Shafer winning over Philadelphian . Senator Clark lost his seat in 1968, but the remainder of the Democratic ticket succeeded with pluralities of over 250,000 votes. This brief historical picture, crude and oversimplified as it is, does give an indication of the competitive nature of the statewide offices. What about the legislative offices? Perhaps the legislative divisions give a more accurate indicator of the statewide and county competition.' 2 (Until 1966, the reapportionment was limited; each county had at least one representative allotted to it.) The Senate, reflecting its predominantly rural character was firmly in the hands of the Republican party while the House, mirroring the patterns of statewide races and reflecting a Demo- cratic buoyancy in the urban areas, tended to be more evenly divided between the two parties.

TABLE TWO

POLITICAL DIVISIONS OF ASSEMBLY, 1954-1968

'53-54 '55-56 '57-58 '59-60 '61-62 '63-64 '65-66 '67-68 '69-7o R D R D R D R D R D R D R D R D R D Senate 32 18 26 24 27 23 28 22 25 25 26 22 27 23 27 23 27 23 House 110 98 98 112 126 83 102 108 101 109 108 98 94 116 106 97 98 104 Session Majority: 54-68 Senate House Republicans 8 4 Democrats 1 5

Total 9 9

'2 Al tables and calculations based on records printed in Pennsylvania Manuals. 358 PENNSYLVANIA HISTORY

From Table Two and focusing upon the House, the vacillating nature of legislative politics is strikingly evident. This might well mean the legislative seats are intertwined with and respon- sive to the coattail effects of statewide races. Evidence for example, the contrast between the 1953-54 House and the 1955- 56 House-elected primarily as a result of the 1954 Democratic gubernatorial victory. Evidence also the 1957-58 House as it reflected a switch-coming on the coattails of the Eisenhower victory. The reverse is true in 1959-60 following Lawrence's victory. With the Scranton Republican win in 1962, the House again changed hands, this time going from a seven seat Demo- cratic majority to a ten seat Republican majority. In 1965-66, the highly urbanized Democratic counties showed some of their legislative numerical muscle-over 55 of the representatives in the Democratic majority of 1965-66 House were from Philadel- phia and Pittsburgh. In 1966, a Republican won the gubernatorial contest, and the Democratic fortunes in the House likewise dipped. Supposedly the Governor's race provided the margin of a few hundred votes-the Democratic State Committee estimated it lost the House by only 477-it took to win precarious control of the House. A final indicator of Pennsylvania's political atmosphere in the period of 1954-68 is that of voter registration:

TABLE THREE PERCENT DISTRIBUTION OF VOTER REGISTRATION

Year Democratic Republican 1953 40.0 59.6 1954 40.5 59.0 1955 45.2 53.4 1956 46.1 52.6 1957 47.3 51.4 1958 48.7 50.0 1959 49.3 49.3 1960 50.7 47.9 1961 51.0 47.6 1962 50.1 48.6 1963 50.3 48.2 1964 50.3 48.2 1965 50.3 48.3 1966 50.1 49.9 1967 49.8 50.2 'IiI INTERPARTY COMPETITION IN PENNSYLVANIA 359

More than any single indicator, these registration figures docu- ment the thesis that the period of 1954-68 was a period of Democratic reentry into competitive politics. Beginning in 1954 the resurgence of the Democratic party in the House, Senate, and especially the Governor's chair could not but help the picture of party registration within Pennsylvania. * * * * * Table Four shows Pennsylvania's counties ranked according to the degree of interparty competition (PCI). These rankings are a composite over all the races considered. The table indicates a two-party plurality difference among the counties, ranging from 1.4 percent to 48.7 percent. In itself, the overall listing is not really crucial; what is im- portant are the subgroupings that can be constructed from the table. Table Five classifies the counties according to the scheme developed on page four. Of those counties utnder the heading of Two Party, ten are subclassified as strong two party and seven are classified weak two party. One interesting factor about these strong and weak two-party counties is the lack of an apparent reason that would explain why these particular counties have high levels of interparty competition. Often in the literature, assertions are made that the more urban an area, the higher degree of competitiveness; conversely, the more rural the area, the lower the degree of competitiveness."3 The preceding state- ment is an over-simplification but it does allow the point to be made that of these seventeen competitive counties, the urbanity levels are mnixed. Erie County, for example, is 77.7 percent urban and is classified "strong two party"; Fulton County is 0.5 percent urban and is also classified as "strong two party." Another interesting point can be seen by comparing Table Five with Table Six. Notice that one-party dominated counties may still be competitive counties for statewide races. Majorities in the county row offices do not automatically lead to (1) strong political organization or (2) assured majorities for the election of statewide candidates. Of those competitive counties classified under "strong two party and weak two party," Allegheny, Clin- ton, Columbia, Mercer, Schuylkill, Northunlberland are classified

"1 See for example, Heinz Eulau, "Ecological Basis of Party Systems," Mlidwest Journal of Politics (1957), 125-135. 360 PENNSYLVANIA HISTORY as dominated by one major party as far as local county politics is concerned. What these two illustrations suggest is that to assess adequately the interparty competition *of a county, one must examine a number of variables. One is likely to find that competitiveness is a function of political leadership, the structure and viability *of the political organizations within the county, the long-standing historical traditions that continue to leave their mark upon polit- ical outcomes, significant events, and personalities of statewide candidates. More central to the themes running throughout this paper is Table Seven-partisan classification of the counties. The Table indicates heavily skewed distribution in favor of the Republicans -numberwise but not necessarily populationwise. Twenty-two counties are classified as "one-party Republican"; five are classi- fied as "weak one-party Republican," and only four Republican plurality leaning counties are considered as "two party." In the Democratic columns, only three are classified as "one-party Democratic"; six are designated "modified one-party Demo- cratic," two are considered "weak one-party Democratic," while six are listed as two-party counties. In general the more competitive counties are skewed in terms of plurality toward the Democratic party while the least com- petitive counties are skewed towards the Republican party. Quantitatively, seventeen of the counties can be roughly called Democratic plurality counties while fifty counties can be roughly called Republican plurality counties. Of the top ten competitive counties, six tend to be Democratic while four are Republican. Democratic counties range from 1.4 percent to 23.3 percent plurality over the Republican party; Republican counties range from 1.8 to 48.7 plurality over the Democratic party. INTERPARTY COMPETITION IN PENNSYLVANIA 361

TABLE FOUR

PENNSYLVANIA COUNTIES RANK: MOST TO LEAST COMPETITIVE, 1954-1968

Rank County Plurality PercentBetween Parties

1 Erie 1.4 2 Fulton 1.8 3 Lawrence 2.5 4 Carbon 2.6 5 Mercer 3.2 6 Columbia 3.5. 7 Elk 3.7 8 York 3.9 9 Berks 4.7 10 Lehigh 4.9 11 Monroe 5.2 12 Luzerne 5.3 13 Northumberland 5.5 14 Clinton 5.8 15 Allegheny 6.7 16 Schuylkill 7.3 17 Armstrong 7.7 18 Sullivan 8.0 19 Clearfield 8.1 20 Bucks 8-.2 21 Juniata 8.8 22 Montour 9.5 23 Northampton 10.4 24 Somerset 10.5 25 Cambria 10.6 26 Lackawanna 10.7 27 Lycoming 12.1 28 Indiana 12.3 29 Franklin 12.6 30 Beaver 12.9 31 Adams 13.0 32 Mifflin 14.3 33 Westmoreland 15.3 34 Delaware 15.6 35 Crawford 15.7 36 Clarion 16.0 37 Bedford 16.8 38 Jefferson 18.1 39 Washington 18.6 40 Forest 18.9 41 Centre 19.6 362 PENNSYLVANIA HISTORY

TABLE FOUR (Continued)

PENNSYLVANIA COUNTIES RANK: MOST TO LEAST COMPETITIVE, 1954-1968

Rank County PluralityPercent Between Parties

42 Cumberland 19.9 43 Fayette 20.0 44 Butler 20.1 45 Blair 20.5 46 Cameron 21.0 47 Dauphin 21.2 48 Perry 21.4 49 Potter 22.2 50 Philadelphia 22.5 5.1 Huntington 23.1 52 Greene 23.3 53 Lebanon 23.8 54 Warren 24.1 55 Chester 24.2 56 Montgomery 24.6 57 Susquehanna 26.4 58 McKean 27.1 59 Venango 27.8 60 Pike 30.0 61 Bradford 31.1 62 Lancaster 32.7 63 Tioga 36.0 64 Wayne 36.5 65 Snyder 41.9 66 Union 43.3 67 Wyoming 48.7 INTERPARTY COMPETITION IN PENNSYLVANIA 363

TABLE FIVE

PENNSYLVANIA COUNTIES Two-PARTY PARTY COMPETITION CLASSIFICATION, 1954-1968

Strong Two Party Weak Two Party

Erie Monroe Fulton Luzerne Lawrence Northumberland Carbon Clinton Mercer Allegheny Columbia Schuylkill Elk Armstrong York Berks Lehigh One-Party Counties

Weak Strong Exceptionally One Party One Party Strong One Party Sullivan Lycoming Fayette Montgomery Clearfield Indiana Butler Susquehanna Bucks Franklin Blair McKean Juniata Beaver Cameron Venango Montour Adams Dauphin Pike Northampton Mifflin Perry Bradford Somerset Westmoreland Potter Lancaster Cambria Delaware Philadelphia Tioga Lackawanna Crawford Huntington Wayne Clarion Greene Snyder Bedford Lebanon Union Jefferson Warren Wyoming Washington Chester Forest Centre Cumberland 364 PENNSYLVANIA HISTORY

TABLE SIX

PARTY DOMINANCE IN PENNSYLVANIA COUNTIES; PERCENT OF COUNTY OFFICES WON 1BYMAJOR PARTY, 1954-1968

Percentage County Party of Offices

Adams R 50- 60 Allegheny D 80-100 Armstrong R 60- 80 Beaver D 60- 80 Bedford R 60- 80 Berks D 60- 80 Blair R 80-100 Bradford R 80-100 Bucks R 60- 80 Butler R 80-100 Cambria D 80-100 Cameron R 60- 80 Carbon R 60- 80 Centre R 50- 60 Clarion R 60- 80 Chester R 80-100 Clearfield D 50- 60 Clinton R 80-100 Columbia D 80-100 Crawford R 80-100 Cumberland R 80-100 Dauphin R 80-100 Delaware R 80-100 Elk D 80-100 Erie D 60- 80 Fayette D 80-100 Forest R 80-100 Franklin R 60- 80 Fulton D 60- 80 Greene D 80-100 Huntington R 50- 60 Indiana R 60- 80 Jefferson R 80-100 Juniata R 50- 60 Lackawanna D 80-100 Lancaster R 80-100 Lawrence R 50- 60 Lebanon R 80-100 Lehigh R 60- 80 Luzerne R 60- 80 INTERPARTY COMPETITION IN PENNSYLVANIA 365

TABLE SIX (Continued)

PARTY DOMINANCE IN PENNSYLVANIA COUNTIES; PERCENT OF COUNTY OFFICES WON BY MAJOR PARTY, 1954-1968

Percentage County Party of Offices

Lycoming R 60- 80 McKean R 80-100 Mercer R 80-100 Mifflin R 50- 60 Monroe D 60- 80 Montgomery R 80-100 Montour R 80-100 Northampton D 80-100 Northumberland R 80-100 Perry R 60- 80 Philadelphia D 80-100 Pike R 60- 80 Potter R 80-100 Schuylkill R 80-100 Snyder R 80-100 Somerset R 60- 80 Sullivan R 60- 80 Susquehanna R 80-100 Tioga R 80-100 Union R 80-100 Venango R 80-100 Warren R 80-100 Washington D 80-100 Wayne R 80-100 Westmoreland D 80-100 Wyoming R 80-100 York D 50- 60 366 PENNSYLVANIA HISTORY

TABLE SEVEN

PARTISAN CLASSIFICATION OF PENNSYLVANIA COUNTIES, 1954-1968

One-Party Modified Weak Democratic One-Party Demiocratic One-Party Democratic

Greene Washington Allegheny Philadelphia Westmoreland Luzerne Fayette Beaver Lackawanna Cambria Northamiaon Two Party Lehigh Mercer Berks Carbon York Lawrence Columbia Fulton Elk Erie

One-Party modified Weak Repnblican One-Party Republican One-Party Republican

Butler Sullivan Monroe Blair Clearfield Northumberland Cameron Bucks Clinton Dauphin Juniata Schuylkill Perry Montour Armstrong Potter Somerset Huntington Lycoming Lebanon Indiana Warren Franklin Chester Mifflin 2Montgomery Delaware Susquehanna Crawford McKean Clarion ATeiango Bedford Pike Jefferson Bradford Forest Lancaster Centre Tioga Cumberland Wayne Snyder Union Wyoming INTERPARTY COMPETITION IN PENNSYLVANIA 367

One question that might be asked is how much change in county electoral alignments has there been over time. Which counties, for example, were most Republican (consequently least competitive) in the years of the earlier studies? Has there been any change in the competitive rankings of Pennsylvania's counties ? Alderfer and Cooke focused on these same general questions. Alderfer calculated rankings for the ten most Republican and ten most Democratic counties (in other words, the ten least com- petitive counties of both parties). Cooke duplicated part of Alderfer's study, only using different base years. Both of these rankings along with the findings of the present study are sum- marized in tables eight and ten.

TABLE EIGHT

TEN MOST REPUBLICAN COUNTIES, 1922-1968

Alderfer Cookc 1922-1944 I944-1958 1968

Tioga Tioga Wyoming XIcKean Snyder Union Union Wayne Snyder Snyder Union Wayne Bradford Venango Tioga Wayne Bradford Lancaster Montgomery Wyoming Bradford Lancaster McKean Pike Venango Warren Venango Wyoming Pike McKean

Examining the Republican party first (Table Eight), the most dramatic changes occurred from 1944 to 1958. In these years, Pike and Warren counties were added to the list while Lan- caster and Montgomery counties were dropped. McKean makes a most significant change, becoming more competitive relative to the other ten. In the period 1954-68, Lancaster County is again added to the list and Warren County is dropped; of these ten counties, McKean moves to the most competitive position on the table. Wyoming continues its movement to the least com- petitive position of all Republican counties. 368 PENNSYLVANIA HISTORY

It may be of more utility to describe these particular voting patterns through the employment of an index. For this purpose an Index of Changeability 'has been calculated:

IC = Mm or Index of Changeability = Moves made by a Mt County Total Moves possible

Using the base years of 1922-1944, each of the Ten Most Re- publican and Ten Most Democratic counties could move twenty total possible positions. Thus Tioga County could move ten places -from first to bottom of the list. It could also move another ten places, from last to first on the 1954-66 scale. The number of moves it did make was five. Its Index of Changeability is then calculated to be 5/20 or .25. A plus or minus sign only denotes direction-plus means that the county became more competitive whereas a minus means the county became less competitive. The greatest index of change a county could have is unity. The closer to zero, the less change in electoral posture over the 46 years under consideration.

TABLE NINE

REPUBLICAN INDEX OF CHANGEABILITY: TEN MOST REPUBLICAN COUNTIES, 1922-1968

I922-1944 1944-I958 I954-I968 Net Index Tioga 0 5 5 .25 Snyder -2 1 -1 -.05 McKean 8 2 10 .5 Union 1 -2 - 1 -.05 Bradford 1 1 2 .1 Wayne -3 1 -2 -.1 Montgomery off scale off scale off scale off scale Lancaster off scale off scale off scale off scale Venango -4 4 0 0 Wyoming -3 -7 -10 -. 5

Compiling the IC, Montgomery County, going off the scale from 1944 to 1968, was the most changeable-moving to posi- INTERPARTY COMPETITION IN PENNSYLVANIA 369 tion fifty-six (See Table Twelve). Of the top ten counties, McKean demonstrated the most consistent changeability-be- coming more competitive relative to the other counties (but still remaining, in an absolute sense, quite uncompetitive). Totally, the IC for all ten counties was .115; the pattern has become increasingly competitive but the size of the IC suggests that overall, a rather stable voting response existed from 1922 to 1968. Table Ten (below) indicates the ten most Democratic counties:

TABLE TEN

TEN MOST DEMOCRATIC COUNTIES, 1922-1968

I922-1944 1944-I958 i954-i968 Lackawanna Greene Greene Elk Fayette Philadelphia Washington Washington Fayette Allegheny Cambria Washington BerksV Westmoreland Westmoreland Greene Lackawanna Beaver Columbia Northampton Lackawanna Fayette Philadelphia Cambria Westmoreland Beaver Northampton Philadelphia Allegheny Allegheny

Examining the table, one first notes the reversed condition of Allegheny and Philadelphia from that of 1922-44. The Repub- lican party during those years controlled the wards of Phila- delphia, and it was only through splits within the organization that one could effectively speak of "Philadelphia going Demo- cratic." By 1954-68, Philadelphia, through successive Democratic control of the Mayor's Office and a shrewd political captain such as Congressman William Greene, was to solidify its position as a cornerstone in Democratic politics. (With recent elections in Philadelphia, such as that of Arlen Specter and Thomas Gola as District Attorney and Comptroller respectively there may be some truth to the observation that being Republican and being in Philadelphia is once again a viable combination.) Allegheny County, on the other hand, has suffered a relapse in Democratic viability at the hands of both infighting and a resurgence of the 370 PENNSYLVANIA HISTORY surrounding suburbia-challenging the hold of the city on the outcome of election. Table Eleven indicates that the Ten Most Democratic counties with the three exceptions noted tended to remain relatively stable in competition-the total IC over the entire spectrum was only .195.

TABLE ELEVEN

DEMOCRATIC INDEX OF CHANGEABILITY: TEN MOST DEMOCRATIC COUNTIES, 1922-1968

1922-1944 1944-I958 I954-i968 Net Index Lackawanna 6 1 7 .35 Elk off scale off scale off scale off scale Washington 0 1 1 .05 Allegheny 6 0 6 .3 Berks off scale off scale off scale off scale Greene -6 0 -6 -.3 Columbia off scale off scale off scale off scale Fayette -6 1 - 5 -.25 Westmoreland -4 0 - 4 -.2 Philadelphia -2 -8 -10 -.5

There are, however, significant differences between the Demo- cratic and Republican "Ten Most" scales. The first is that of population. On the Republican scale the only county of sizable population is that of Lancaster while on the Democratic scale, the urban areas are highly represented. The second distinction is that of plurality ranges. The Republican range is from 27.1 percent plurality to 48.7 percent plurality-a difference of 21.6 percent. On the other hand, the Democratic "Ten Most Com- petitive Counties" though large in population do not fair so well in the plurality category-a range of 6.7 to 23.3 and a difference of 16.6 percent. This is important in two respects. First, the Democratic counties are in reality much more com- petitive than their Republican counterparts. Second, the competi- tive differential and the fact that this differential occurs in counties of large population has implications for campaign strategy of Democratic candidates. Table Twelve presents the Index of Changeability and the competitive rank of all Pennsylvania counties in two time periods, INTERPARTY COMPETITION IN PENNSYLVANIA 371

1922-44 and 1954-68. The table is arranged alphabetically to facilitate comparison. If a county's Index of Changeability has a minus sign before it that sign denotes only direction of change. The closer a county's IC approaches unity the more or less competitive, relative to all other counties, it has become.

TABLE TWELVE

INDEX OF CHANGEABILITY: PARTY COMPETITION CHANGES FROM 1922 TO 1968

Competitive Rank Coinpetitive Rank Index of County I922-I942 1954-I968 Changeability

Adams 41 31 .149 Allegheny 3 15 -.179 Armstrong 38 17 .313 Beaver 15 30 -. 223 Bedford 7 37 -. 447 Berks 43 9 .507 Blair 8 45 -.552 Bradford 63 61 .029 Bucks 20 20 .0 Butler 24 44 -.298 Cambria 48 25 .343 Cameron 52 46 .089 Carbon 49 4 .671 Centre 9 41 -.477 Clarion 35 36 -.014 Chester 26 54 -.119 Clearfield 17 19 -. 029 Clinton 23 14 .134 Columbia 56 6 .746 Crawford 22 35 -.194 Cumberland 13 42 -.432 Dauphin 34 47 -.194 Delaware 45 34 .164 Elk 11 7 .059 Erie 12 1 .164 Fayette 46 43 .044 Forest 42 40 .029 Franklin 4 29 -. 373 Fulton 54 2 .776 Greene 57 52 .074 Huntington 47 51 -.059 Indiana 40 28 .179 372 PENNSYLVANIA HISTORY

TABLE TWELVE (Continued)

INDEX OF CHANGEABILITY: PARTY COMPETITION CHANGES FROM 1922 TO 1968

Competitive Rank Competitive Rank Index of Coœ1ty 1922-I942 1954-I968 Changeability

Jefferson 29 38 -. 134 Juniata 30 21 .134 Lackawanna 37 26 .164 Lancaster 27 62 -. 522 Lawrence 39 3 .537 Lebanon 28 53 -. 373 Lehigh 33 10 .343 Luzerne 2 12 -. 149 Lycoming 1 27 -. 388 McKean 66 58 .119 Mercer 18 5 .194 Miffin 6 32 -. 388 Monroe 51 11 .597 Montgomery 61 56 .074 Montour 53 22 .462 Northampton 55 23 .477 Northumberland 25 13 .179 Perry 10 48 -. 567 Philadelphia 16 50 -. 507 Pike 44 60 -. 238 Potter 14 49 -.522 Schuylkill 31 16 .223 Snyder 64 65 -.014 Somerset 32 24 .119 Sullivan 5 18 -.194 Susquehanna 60 57 .044 Tioga 67 63 .059 Union 65 66 -.014 Venango 59 59 .0 Warren 36 54 -. 268 Washington 21 39 -.268 Wayne 62 64 -.029 Westmoreland 19 33 -. 208 Wyoming 58 67 -.134 York 50 8 .628 INTERPARTY COMPETITION IN PENNSYLVANIA 373

TABLE THIRTEEN

TEN MOST COMPETITIVE COUNTIES, 1922-1968

1922-I944 1954-i968

1. Lycoming 1. Erie 2. Luzerne 2. Fulton 3. Allegheny 3. Lawrence 4. Franklin 4. Carbon 5. Sullivan 5. Mercer 6. Mifflin 6. Columbia 7. Bedford 7. Elk 8. Blair 8. York 9. Centre 9. Berks 10. Perry 10. Lehigh

TABLE FOURTEEN

ELEVEN COUNTIES THAT CHANGED COMPETITIVE POSITION THE LEAST FROM 1922-1944 TO 1954-1968

County IC

Bucks .0 Venango .0 Clarion -.014 Snyder -. 014 Union .014 Bradford .029 Wayne .029 Clearfield -. 029 Forest .029 Fayette .044 Susquehanna .044 374 PENNSYLVANIA HISTORY

TABLE FIFTEEN

TEN COUNTIES THAT EVIDENCED THE MOST CHANGE- BECOMING MORE COMPETITIVE, 1922-1944 TO 1954-1968

County IC

Fulton .776 Columbia .746 Carbon .671 York .628 Monroe .597 Lawrence .537 Berks .507 Northampton .471 Montour .462 Cambria .343

TABLE SIXTEEN

TEN COUNTIES THAT EVIDENCED THE MOST CHANGE- BECOMING LESS COMPETITIVE, 1922-1944 TO 1954-1968

County IC

Perry -.567 Blair -. 552 Potter -. 522 Lancaster -. 507 Philadelphia -. 507 Centre -. 477 Bedford -. 447 Cumberland -. 432 Lycoming -.388 Mifflin -. 388

Bucks and Venango counties changed the least during the period under study. The IC for these two is 0.0. This does not mean that the counties remained stagnant or that the same party always won every election. The IC merely indicates that, over the years, both counties remained in the same positional rank- ing. Internally within the counties, there may well have been an increased vote getting ability of the parties; but relative to the other Pennsylvania counties, these two were the most stable in competitive posture. INTERPARTY COMPETITION IN PENNSYLVANIA 375

Counties manifesting the largest change-to more competitive posture-were Fulton and Columbia. The IC is rather high for both counties, indicating a considerable positional change (a relative more competitive change also) during the years 1922-68. This does not necessarily mean that Fulton County is the most competitive county; it does mean that during the period analyzed it did change the most-going from a rather noncompetitive posi- tion to an increasingly competitive position. Perry and Blair counties were the reverse of Fulton and Columbia. Perry County as did Blair went from a competitive ranking to a decreasingly competitive ranking. One possible route for an explanation of the changing nature of Pennsylvania politics is V. 0. Key, Jr.'s, Theory of "critical elections" in which

the decisive results of the voting reveal a sharp altera- tion of pre-existing cleavages within the electorate. Moreover, and perhaps this is truly the differentiating characteristic of this sort of election, the realignment made manifest in the voting in such elections seems to persist for several succeeding elections.':

Nationally, Key has termed those of 1896 and 1932 as critical elections. In a single state such as Pennsylvania the demarcations are not quite as clear; but surely if any elections would qualify as "critical" it would likely be the 1934 and 1954 gubernatorial elections. In 1934, the Democrats elected a governor for the first time since 1890. Thirty-six of the counties went Democratic while the vote separating the two parties was reduced consider- ably from the years before. Thus, in 1926, Republicans polled 1.1 million votes while Democrats polled 349,000 votes. In 1930, Republicans polled 1.0 million votes; Democrats, 643,00. In 1934, Reputblicans polled 1.41 million and Democrats 1.47 million. Undoubtedly, the forces at work in the nation as a whole did influence the vote pattern in Pennsylvania. Though Democrats did not win another election until 1954, their vote total remained " V. 0. Key, Jr., "Theory of Critical Elections," Journal of Politics, XVII (February, 1955), 3-18. See also Duncan MacRae, Jr., and James A. Meldrum, "Critical Elections in Illinois: 1888-1958," American Political Science Review, LXIV (September, 1960), 669-683. Charles Sellers, "The Equilibrium Cycle in Two Party Politics," Public Opinion Quarterly, XXIX (Spring, 1965), 16-38. 376 PENNSYLVANIA HISTORY considerably closer to the Republican total than the low of 1926. Thus, up to the 1954 election, pluralities of only two hundred to three hundred thousand votes separated the parties. In 1954, George Leader captured the Governor's chair. The trend this election set in motion 'has been well documented on previous pages. Suffice to note that these two elections 1932 and 1954 deserve special attention in any explanation of political change. The previous analysis has raised more questions than it has answered; and, no attempt will be made to deal with all these questions. One aspect of the analysis is interesting though-that which deals with the strategic-electoral implications of competition and competitive change. The significance of the difference between the Democratic and Republican ten least competitive counties is more apparent when viewed from the perspective of electing politicians to office. Cooke writes:

The importance of marginal counties to the Repub- lican party becomes increasingly clear when a com- parison is made with the list of the top ten Democratic counties, the result indicating that over half of the marginal counties are also ranked high on the list of the most Democratic counties. Thus, if the Republicans can cut into Democratic vote in these counties, they stand to increase their own pluralities but also tend to reduce Democratic changes for statewide victory since that party is dependent upon a large urban vote."'

Thus, by cutting into the six Democratic leaning "two-party counties"-York, Elk, Berks, Carbon, Lawrence, and Erie-and by cutting into the Democratic "weak one-party counties" of Allegheny and Luzerne, theoretically, the Republican party can lose the large urban centers and still win statewide races. A similar strategy was employed by Senator Hugh Scott in his 1964 victory when he won reelection despite a landslide vote for President Johnson. Scott carried all the one-party Republican, all the modified one-party Republican, and all the weak one-party Republican counties. While losing all but one of the two-party counties. Scott effectively neutralized a coattail vote by thoroughly cutting pluralities of the Democratic candidate.

"Cooke, "Patterns of Party Competition, 1944-58," 84. INTERPARTY COMPETITION IN PENNSYLVANIA 377

TABLE SEVENTEEN

1964 SENATORIAL VOTE, TWO-PARTY COUNTIES

Scott Blatt Johnson County (Plurality) (Plurality) (Plurality) Lehigh 1,000 28,000 Berks 8,000 35,000 York 1,000 25,000 Elk 2,000 6,000 Mercer 1,000 14,000 Carbon 2,000 8,000 Lawrence 600 15,000 Erie 800 45,000 Fulton 300 400 Weak Two-Party Counties (Dem.) Allegheny 16,000 240,000 Luzerne 25,000 65,000

The extent to which Scott reduced Democratic pluralities is rather striking. In Allegheny County, Scott actually managed to win while Johnson swept the County with over 240,000 plurality. Whether this strategy is a tribute to Scott's political astuteness or a political quirk pushed along by a divisive 1964 Democratic Senatorial primary is not the question. What is evident is that it did happen; and for Democratic candidates framing statewide races, the plurality-population problem is crucial. Let us touch hypothetically for a moment on the effects of interparty competition as it helps shape a Democratic candidate's statewide strategy. Normally, such a strategy would be to count in his corner the urban centers of Philadelphia; the hard coal regions of Lackawanna, Northampton, and Luzerne; and the Southwestern portion of Pennsylvania. In the two-party areas he would normally concentrate his campaign on the urban areas within the county. At the same time though, any realistic strategy will include making inroads in the weak and modified one-party, mostly rural Republican areas. Yet, it is with this very strategy that the effects of party competition are operating. In these rural counties, the competitive factors have often reduced the Democratic party to an ineffective skeleton. Democratic Party Chairmen call meeting after meeting. 378 PENNSYLVANIA HISTORY usually failing each time to obtain a committee quorum. Many of the crucial committee positions in the townships and boroughs are not filled, often, election after election goes by when the grass roots offices-those races which might build party organiza- tion and vote producing strength-are uncontested. In sum, the quality of leadership of the Democratic party in these rural counties is usually not conducive to obtaining the kinds of marginals a Democratic candidate needs to augment or offset any reduction in his core vote. The other side of the dilemma posed is that the core city areas are normally subject to much pressure from the Republican party. First, a large Republican suburban vote can neutralize a Democratic city majority. Second, the level of competition within the city itself is one in which there is seldom a complete abdica- tion of the minority party competitive role. These outlines are hypothetical and this paper makes no com- plete effort to examine them. Yet, the inference suggest that an important key to successful Democratic statewide races lies not entirely with the large urban areas; much of the key may be obtaining quality leadership in the rural areas to make respect- able the Democratic voting lever. A second dimension to the hypothetical outline above-and one that dispells somewhat the bleakness sketched-is that of correlations between party dominance and party competition. Table Eighteen presents the product moment correlations be- tween these two variables over the entire range of counties and races.

TABLE EIGHTEEN

CORRELATIONS: PARTY COMPETITION AND PARTY DOMINANCE

Dem. Rep. Dem. Rep. Dem. Rep. Sen. Sen. Gov. Gov. St. Wide St. Wide Rep. Dominance .715 .510 -.751 .751 - .512 .532 Dem. Dominance .778 -.490 .806 .806 .653 -. 688

The striking aspect of the table is that there is more association between those counties dominated in the courthouse by Demo- crats and the vote Democratic than there is between those INTERPARTY COMPETITION IN PENNSYLVANIA 379 counties dominated in the courthouse by Republicans and the vote Republican. The Democratic Senatorial candidates, for ex- ample, correlated .778 with Democratic dominance while the Republican Senatorial aspirants correlated only .510 with counties of Republican dominance. These correlations indicate that a Democratic strategy to in- crease resources and obtain quality leadership in Republican dominated (Republicans control the county row offices) counties has potential electoral payoffs. County courthouse domination by Republicans does not translate as readily into effective statewide vote production as does organization and leadership stemming from those counties in which the Democrats dominate. It is as the table indicates easier for Democrats to generate votes in Republican dominated counties than it is for Republican candi- dates to secure votes in Democratic dominated counties. This paper has classified the counties into various categories and has examined these categories from a variety of perspectives such as changes in interparty competition, historical, and electoral strategy. The analysis has indicated that the period from 1954 to 1968 can appropriately be termed "the competitive period" of Pennsylvania politics. During these years, both of the major parties won and lost an almost equal number of elective offices. In addition, registration figures indicate a near fifty-fifty split in party adherents. Numerically, the Republican party has achieved pluralities in fifty of the sixty-seven counties; Democrats, seventeen counties. The preponderance of Democratic pluralities in counties classified as "two party" and "weak two party" increases the election time importance of these counties. Republicans can concentrate their attention on these counties while maintaining with relative ease their rural stronghold. Democrats must receive a large plurality from these marginal counties in order to compete in statewide elections. All is not totally gloomy for the Democrats though, for measures indicate that Democratic candidates have shown a greater ability to attract support in Republican counties than Republican candidates can attract in Democratic counties. The most competitive counties during the period studied were Erie, Fulton, Lawrence, Carbon, Mercer, Columbia, Elk, York, Berks, and Lehigh. Of these, six-Erie, Lawrence, Carbon, Elk, 380 PENNSYLVANIA HISTORY

York, and Berks-evidenced Democratic leanings; the remaining four leaned to the Republican party. Over an extended period of time, 1922-68, only two counties maintained a stable relationship as to interparty competitiveness. Overall, there was considerable change in the competitive com- plexion of Pennsylvania's politics; one possible explanation for this change are the forces set in motion via "critical elections" within the Commonwealth. To venture a relatively safe prognosis-Pennsylvania's politics, along with many -of her similarly sister industrial, urbanized, evenly matched political party states, will continue the competitive trend begun with the 1934 and accelerated by the 1954 guber- natorial elections.