Community, Class and Voting Behavior

Taken from “ Democracy and Clientelism in Jamaica” by Carl Stone pp 139-157 

The pattern of party voting in Jamaica represents a deceptively com­plex, stable, and predictable trend of electoral choices. The factors that influence these trends are many, but the more important ones can be iso­lated and disentangled to establish the precise lines of causal influences that shape the Jamaican voter.

Earlier studies1 of voting in Jamaica have established certain patterns of class and demographic influences as well as regional parish traditions as factors affecting voting behavior. These studies have focused mainly on single elections or party sentiments at single points in time. As a conse­quence, they lack an adequate treatment of how these class, regional, and demographic factors affect the movement of the vote over the cycles of electoral change that have taken place since the 1940s. 

In addition, the precise relationship and overlap between class, urban-rural patterns and local or regional party voting traditions in Jamaica have yet to be established empirically. This is especially important given the fact that there is a close association between class and urbanization and because regional parish voting trends are also influenced by class and demographic factors. 

One recent study2 by the author attempted a new point of departure by focusing on community voting patterns. The emphasis is justified partly on the grounds that there are strong local level agencies of political socializa­tion which function to maintain stability in the level of party voting. These include community and neighborhood pressures and community influen­tials; the high levels of illiteracy in the society which filter information and ideas through “word of mouth” or face to face communication networks that are controlled by local opinion leaders; the strength of local peer group influences on individual behavior patterns in an essentially small scale and closely knit society; and the high level of unemployment and own-account trading activities, both of which result in a large proportion of the community being at home or in the neighborhood most of the time. In addition to factors that help to channel local influences on the political and party predispositions of the family or individual voter, two other important factors have to be taken into account. The first concerns the extent to which many local communities have an explicit sense of political territory and a complementing identity with one or the other of the two political parties. The second factor which reinforces the first is that vio­lence and intimidation are liberally used in the more politicized urban communities to limit the organizational penetration by opposing party activists and to coerce the more independent voters into supporting the dominant local party. The open use of violence in party activities has increased considerably in recent years due to the involvement of large sec­tions of the violent “Iumpen” class in mercenary activities on behalf of the two parties. In addition, some very committed activists in many urban poor areas also belong to the parasitic sector that lives by crime and violence. 

As a consequence of these stabilizing factors on the vote, approximately fifty percent of the communities in the twelve more rural parishes have strong voting tendencies towards one of the political parties and another thirty percent show similar but weaker party preferences. The remaining twenty percent of communities show no decisive or clear pattern favoring one of the two parties. In the Kingston and St. Andrew metropolitan area, party voting trends have been less stable than in the rural parishes, but a similar trend of concentration of communities with strong party voting tendencies in either a J.L.P. or a P.N.P. direction also exists. 

The twelve more rural parishes and the two parishes making up the met­ropolitan area of the island can be divided into three main types of com­munities. These include, first, urban concentrations with at least a population size of 5,000 residents. Second, there are the “main road” com­munities that are located close to the main highways that connect the major population, production and commercial centers of the island. Third, there are the more remote rural and mainly hillside districts. 

The remote districts (much of which tend to fall in the traditional inte­rior zones of hillside peasant farming) have a large concentration of small farmers and agricultural workers as their main class component, a minor­ity of manual nonagricultural labor and very few nonmanual workers. In all these communities, the electorate reflects a numerical predominance of small-scale agriculturals as a defined set of class interests embracing the interconnected groupings of small peasant farmers, the rural poor, and landless wage labor. The following data in Table 7.1 (summarized from voters’ lists for a sample3 of the more remote communities in three par­ishes) illustrate the pattern of small-scale agricultural strength in the local electorate. In these areas there is a predominance of small-scale agricultu­ral class interests.

  These remote communities can therefore be accurately typed as small peasant areas, but with the clear implication that they represent the most isolated,4 rural, poor, and economically backward communities in the island. These communities are consequently defined as isolated peasant communities. 

The second category, “main road areas,” is a much more heterogeneous grouping of communities. Essentially they lie within easy reach of the main urban centers and therefore represent intermediary communities between “town” and “country.” They serve as important links of communication between the more rural and more urban regions of the island. Unlike the isolated peasant communities, nonagricultural labor tends to be numeri­cally larger than the agricultural component of the local community labor force, although in several such communities the farming sector (small, large, and medium scale) is an important means of livelihood and source of economic activity. As a consequence, agricultural interests (including wage labor, some small farmers, and medium and large scale farm opera­tors) tend to constitute a significantly large, although not usually a major­ity proportion, of the local community labor force. This pattern is illustrated by the sample of labor force distributions shown in Table 7.2.                             

The third category of urban communities consists primarily of manual wage labor and a not insignificant minority of white-collar workers. These range from large tourist or market towns with populations between 20,000 and 40,000 residents to much smaller market centers. Here the division of labor is more developed, occupations are more narrowly specialized, and trade unions have a strong presence in the larger enterprises in the modern sector of the economy. The infrastructure of commerce and administration is obviously most developed in the urban centers which means that there is a relatively larger component of service and government workers among wage labor than in the more rural communities. The larger urban areas are stratified into well-defined neighborhood districts and communities, many of which have a distinct class component. These patterns of class differenti­ation of local urban communities therefore allow for a separate measure­ment of urban vote patterns as a whole as against the vote trends for specific “class communities” which correspond approximately to the cate­gories of class groupings developed in chapter one. 

Table 7.3 illustrates, from a sample6 of polling divisions, the occupations spread of the labor force in the urbanized areas in three of the more rural parishes. In these urban centers, the residential pattern is not as sharply stratified by large blocks of class homogeneous communities as is the Kingston and St. Andrew metropolitan area. Analysis of the class patterns of voting within the urban communities will therefore be confined to the metropolitan area of the capital city. 

Earlier analysis of class voting patterns relied entirely on occupational criteria by which to classify and rank urban metropolitan communities on


a class scale. The data presented in chapter one raises some doubts about the precision of an entirely occupational classification of class-communities given the overlap of income categories and living standards among manual occupations. In spite of the overall correlation between the occupational hierarchy, income, and living standards, it would seem that criteria of classification that measure more directly the poverty and levels of living among the urban communities would provide a more precise clas­sification of class categories for the respective urban metropolitan com­munities. The criteria used will therefore be a composite index7 of overcrowding and unemployment levels. The index was used to construct the following seven point objective scale of class-communities made up of clusters of polling divisions that lie within the larger areas of districts listed in Table 7.4. These clusters of polling divisions all represent local neigh­borhoods in the larger communities. 

In examining the relationship between class and voting in the urban met­ropolitan area, only the last three elections were included (1967, 1972, and 1976) due to the extensive internal migration which changes the class com­position of the local communities over periods of time. This is especially so as the period between the 1950s and l970s was a period of extensive middle- and lower-middle-class housing expansion8 that affected all classes. Government low—income housing also created new communities over the period. In the case of the twelve more rural parishes a more stable class community pattern permits an examination o a wider spread of elec­tions over time and this will include the five elections between 1959 and 1976. 

Table 7.5a sets out the average vote for the two parties in the three main types of peasant, main road, and urban communities in the twelve more 


Groupings of Class Communities Based on Poverty Index


(1)     Very poor working class communities



Hannah Town, Rennock Lodge Seaward

Pen, Down Town West, Down Town East,

Rae Town, Down Town Central, Denbam

Town, Greenwich Town, Cockburn

Gardens, Trench Town, Jones Town,

August Town, Cassava Piece, Majesty

Pen, Waterhouse, cedar Valley.


(2)     Poor working class communities


Allman Town, Fletcher’s Land, Franklyn

Town, Campbell Town, Bay Farm, Whit-

field Town, Mountain View Avenue,

Whitehall, Hope Tavern, Grants Pen,

Woodford Park, Tivoli Gardens.


(3)     Better off working class communities


Newton Square, Springfield, Bournemouth, Rollington Town.




(4)    Lower middle class communities



Washington Gardens, Molynes Gardens, Vineyard Town, Zaidie Gardens, Norman Gardens, Harbour View,



(5)     Middle class communities


Arcadia, Mona Heights, East Kings House, Trafalgar Park, Hope Pastures,



(6)     Upper middle class communities

Billy Dunn, Cherry Gardens, Long Lena, Barbican Heights.



  rural parishes which exclude the Kingston and St. Andrew metropolitan area. Bearing in mind the fact that the J.L.P. won the 1962 and 1967 elec­tions, and the P.N.P. the 1959, 1972 and 1976, a number of clear inferences emerge from these voting trends. First of all, the party that wins an election always wins a majority of the main road and peasant vote. In that regard, the national party majorities in these two more rural community types are an accurate indicator of which party is likely to win an electoral contest. This does not apply to the urban communities where the P.N.P. consist­ently wins a majority, although it is a fluctuating majority. 

The actual differences between the community areas as regards the party shares of the vote are not very large but they follow a predictable pattern. The rank order of party strength between the three types of areas has been similar for all three elections. For the J.L.P., the peasant community vote is always the largest proportion followed by the main road communities, with the urban community vote being always the smallest J.L.P. vote pro­portion. The order is, of course, reversed for the P.N.P. with the urban

                                                                                    TABLE 7.5A

Party Vote by Area for Twelve Rural Parishes IN %

                          Communities                          1959                          1962                          1967                          1972                          1976


                                JLP                                 47.6                                54.5                          55.25                             46.7                          47.9

                                PNP                                52.4                                45.5                         46.75                             53.3                           52.1 

Main Road:

                                JLP                              42.25                                52.75                             51.5                            42.75                           45.2

                                PNP                             57.75                                47.25                             48.5                            57.25                           54.8 


                                JLP                             41.4                                46.9                                46.7                                37.9                          39.25

                                PNP                            58.6                                55.1                                  55.3                                62.1                         60.75

  areas being always the largest vote proportion followed by the main road areas and the peasant communities in that order. 

The size of the differences in the party vote proportions for the three community types is, however, mainly less than ten percent between the highest and lowest party vote proportions. Although the differences are predictable, they suggest a high level of overall national uniformity in the vote and small margins of majorities between the parties in the responsive communities. This apparent uniformity may well be a product of the can­celling effect of an even spread of strong J.L.P. and P.N.P. communities which disguises party vote concentrations at the community level. It is therefore necessary to examine the actual distribution of party tendencies within the respective community types. 

An analysis of the spread of party vote tendencies among the three types of communities indicates that in the small peasant and main road com­munities between fifty to fifty-five percent of the communities have strong party vote tendencies amounting to twenty percent, or larger vote margins for one of the parties. The remaining forty-five to fifty percent of these more rural communities have weak partisan preferences with vote margins between ten and nineteen percent, or no clear partisan tendency indicated by party vote margins of less than ten percent. The relatively even and seemingly uniform spread of party strength in the various communities as a whole is due partly to the cancelling effect of concentrated J.L.P. and P.N.P. strength in half of the communities, but also is attributable to the small vote majorities held by the two parties in the other half of the com­munities in the twelve more rural parishes. 

In the urban communities, the party vote pattern is somewhat more con­centrated with approximately sixty percent of the communities showing a strong party tendency primarily in the direction of the P.N.P. but with a minority having a strong J.L.P. tendency. The remaining forty percent of urban communities show weak party vote majorities. 

The uniformity of the swing pattern and direction revealed in Table 7.5b also reflects the evident central tendencies and national pattern that char­acterize the electorate and underlies and qualifies the class and community differences that exist in voting trends. The realigning elections (1962, 1972) reveal a uniform large swing for the winning party for the three types of communities that unseated governing parties. The stabilizing elections (1967, 1976) also show an equally uniform swing pattern favoring the opposition party but not large enough to affect the balance of seats.                                                .

The vote competitiveness index in Table 7.5c is based on the ratio between the size of party vote swings and the size of the respective vote safety margins (or the margins necessary to preserve a vote majority for the leading party). Any index of less than one means the average size of swings are larger than the average safety margin, which implies that party majori­ties on average are vulnerable to change from one party to another. Indices of greater than one are safe from the threat of average vote swings. 

The indices shown in Table 7.5c confirm the high level of competitive­ness in the small peasant and main road communities. By contrast, the party majorities in the urban communities dominated by the P.N.P. are



Vote Competitiveness 

          % Comuetitiveness Index 

                                    Peasant Communities                         .82 

                                    Main Road Communities                   .71 

                                    Urban Communities                         1 .21

   relatively safe from the normal or average party vote swings in these urban areas. The fact that sixty percent of the Jamaican vote is located in the rural areas underscores the role of the rural voter in maintaining the com­petitiveness of the Jamaican two-party system. Further urbanization of the society could therefore have the effect of reducing the high level of voter competitiveness in the Jamaican electorate. The decline of the modern sec­tor of the economy in the l970s, however, suggests that the rapid urban drift of the rural population experienced in the post-World War II period may well generate a substantial decline in the future. 

The political sociological basis of the greater competitiveness of the rural vote rests on four important factors. First, the rural areas receive a disproportionately smaller share of public expenditure and infrastructural development due to the greater political influences of the urban classes. More importantly, the urban areas represent the nerve centers of local and central government. Second, in the channels of patron-broker-client rela­tionships that develop, all governing parties tend to integrate more exten­sively into urban networks as against rural networks that tend to get left behind in the center-periphery, town and country biases that characterize the party system. 

Third, middle-class party leaders, including elected representatives from rural areas, tend to become absorbed into bureaucratic tasks and profes­sional jobs that pull them to the urban nerve centers of commerce, produc­tion, and administration. This process of urban encapsulation develops communication and credibility gaps between rural elected representatives and their main road and small peasant constituents that heighten aliena­tion against incumbent parliamentarians. Rural based parliamentary seats are therefore more vulnerable than urban seats to opposition party chal­lenges at the polls. Fourth, rural communities are not as easily dominated by violence-prone local party machines that intimidate urban voters in many poor areas and establish hegemony over well-defined political terri­tory. Voting is therefore more free from local-community party terrorism although the difference is one of degree rather than kind as the rural areas have become more and more subject to political violence. 

Stability, predictability, and an underlying central tendency limiting the evident differences in party vote strength in the three types of communities are clearly the major features of the vote patterns revealed. In the light of the earlier discussion of party traditions between the parishes which divide between P.N.P., J.L.P., marginal and unstable tendencies, it is necessary to analyze the vote pattern of the three community types in these divergent parish environments. This is necessary to distinguish the effect of parish traditions on the vote from the effect of the community types, and to con­trol for the fact that parish influences may be obscuring or disguising some of the vote differences between the three community types.

As Table 7.5a makes very clear, the differences between P.N.P. and J.L.P. strength in the small peasant communities, the urban communities and the main road communities are consistent and predictable but they are very small. The question that must now be posed is whether a control for types of parishes (P.N.P., J.L.P., marginal, and unstable) would show up larger differences for some parish types as against others. The other related question posed is whether the apparent consistent differences between the party vote strength of the respective community types is not influenced and perhaps exaggerated by the fact that some community types tend to be more heavily concentrated in some parishes so that the parish party tradi­tions may influence the findings on community party voting.

Table 7.6 separates the four parish types (P.N.P., J.L.P., unstable, and marginal) in which the class communities are located and distinguishes the parish type locations for each grouping of small peasant, main road, and urban communities. Only in the P.N.P. and marginal parishes does the rank order of party vote strength for the P.N.P. conform to the national pattern of urban communities having the largest vote proportions, fol­lowed by main road and small peasant communities, in that order. With­out exception, however, in each of the five elections in all four types of parishes, the P.N.P.’s vote proportion is highest and conversely the J.L.P.’s vote proportional is lowest in the urban communities as against the rural and main road communities. The differences are consistent, but nevertheless, relatively small. In this respect, the voting pattern of the com­munities revealed in the grouping b-y parish types is not fundamentally dis­similar to that reflected in the earlier data on the overall national pattern. 

A comparison of the 1959 P.N.P. vote margins with those of 1976 in P.N.P. parishes reveals some important changes and trends in relative party strength between the earlier and the later periods of P.N.P. ascendancy. In all three categories of P.N.P. communities, there is a five to seven percent decline in the P.N.P. share of the vote. On the other hand, P.N.P. vote strength has increased in all types of communities over the same period in the unstable, marginal, and J.L.P. parishes. This pattern could be interpreted to suggest a declining trend in the differences in relative party strength between the four main types of parishes. The time period is, how­ever, much too short to arrive at any definitive conclusions about the trends.

This trend is further confirmed in Table 7.6 which compares the size of community party vote differences (controlling for parish effects) with those attributable to parish voting traditions. The differences in P.N.P. vote proportions shown in Table 7.6 between J.L.P. and P.N.P. parish communities of similar community type reflect the impact of parish party vote traditions, while the differences between urban and rural communi­ties (shown separately for J.L.P. and P.N.P. parishes) reflect the impact of the class-community differences. The differences due to parish party tradi­tions were large up to the 1967 election but they have declined substantially thereafter in 1972 and 1976. On the other hand, the differences due to class-community factors have been modest in size though consistent in pattern. This points tentatively to a growing trend towards declining parish region­alism in the party patterns and to an increasing central and uniform ten­dency in voting preferences throughout the twelve more rural parishes in Jamaica. 

A comparison of the P.N.P. proportion of the vote in the periods of J.L.P. ascendency also reveals some interesting findings which point to the strength of the competitiveness of the Jamaican two-party vote. In the J.L.P. parishes and the unstable parishes, the J.L.P. established large majorities over the P.N.P. during periods of J.L.P. ascendency in the small peasant and main road communities. In the P.N. P. periods of ascendency, small margins divide the J.L.P. and P.N.P. majorities in the small peasant and main road communities in the J.L.P. and unstable parishes. In con­trast, the urban community party vote margins tend to generate majorities for the winning party in the J.L.P. and unstable parishes. The similarity between the J.L.P. and the unstable parishes is based on the fact that the latter were J.L.P. parishes, which have shifted to P.N.P. dominance in the l970s. 

In the marginal parishes, the small peasant party vote consistently yields small majorities for the winning party, regardless of which party is in the ascendency. The main road vote in these parishes, on the other hand, always generates P.N.P. majorities with the size of the majorities declining to insignificance during the J.L.P. periods of ascendency.

The main road and urban communities in the P.N.P. parishes, always yield appreciable P.N.P. vote majorities regardless of which party is in ascendency. In contrast, P.N.P. majorities in the small peasant areas decline to very small margins in the periods of J.L.P. ascendency. 

The entire picture of the party vote in these twelve more rural parishes adds up to a dynamic pattern of competitive two-party voting that is stable and predictable and is built around interelection swings in two-party strength that are relatively large compared to the modest size of the party majorities at the community level.

As a region of the country consistently producing P.N.P. majorities since the 1949 election, the parishes of the Kingston and St. Andrew play a central role in Jamaican electoral politics with their large population of 547.6 thousand residents in 1970, representing approximately thirty per­cent of the island’s population. Table 7.7 sets out the pattern of two-party voting in these parishes since the first universal adult suffrage election in 1944. Although P.N.P. majorities have not been consistently large they have been unbroken between 1949 and 1976 for both parishes, with 1976 being the largest P.N.P. vote margin ever attained in the overall metropoli­tan area vote. 


Party Voting in Kingston and St. Andrew (1944-1976) 

% Share of Popular Vote

Kingston                      1944                    1949                1955                   1959               1962               1967           1972          1976
   PNP                               35                      61                    56                        62                   57                  51                54              61
   JLP                                58                      39                    37                        38                  42                   49                 46             39
St. Andrew
 PNP                                32                    50                    53                          56                  55                    53                 62             64
  JLP                                49                     48                    45                          44                 41                     47                38               36

  * includes rural St. Andrew  vote. 

The parishes which form part of a single Municipal Government Authority consist primarily of an urban and suburban metropolitan area surrounded by a small rural farm population in the hilly interior of St. Andrew. As is evident from Table 7.8, the Kingston and St. Andrew par­ishes are fairly typical of the superiority of P.N.P. over J.L.P. voting strength in the major urban centers of the country. Montego Bay in the parish of St. James and Savannah-la-Mar in the parish of Westmoreland, are however, more consistent and stronger P.N.P. strongholds than the Kingston and St. Andrew capital city. If indeed the country were to follow this urban voting pattern, Jamaica would be operating a one-party domi­nant system controlled by a P.N.P. monopoly of state power rather than a delicately balanced two-party system. 

The overall party vote strength in the Kingston and St. Andrew area sug­gests a pattern of stability and predictability similar to the vote trends in the more rural parishes. However, this is clearly not the case as can be seen from the voting trends of the respective socio-economic communities dis­cussed earlier and presented in Table 7.9. All socio-economic areas exclud­ing the poor working class and the hillside peasant communities reflect very high levels of vote instability at varying points in time. These large fluctuations in the party vote pattern have a compensating or cancelling out effect that presents the appearance of a stable vote pattern at the aggre­gate level. 

Among the urban class-communities in the metropolitan area, it is of interest that highest levels of vote instability are found among the poorest and the affluent communities. The underlying reasons differ, however, although they relate directly and indirectly to the phenomenon of cliente­listic politics. 

The first important factor to be considered is that the poorest communi­ties tend paradoxically to follow the party in power for two quite different reasons. When a governing party has begun to lose its grip over its majority coalition and defections begin to occur, the poorest stratum of the urban society begin to defect in large numbers either by withdrawing into apathy or antiparty militancy, or by shifting support to the opposition party on its upward claim to power. On the other hand, when a victorious party has taken over power and sets about consolidating its clientelistic base of sup­port, the poorest stratum of the urban society is available for absorption and integration into tightly knit urban patron-broker-client machines which dispense jobs, welfare, housing, and other such benefits from the state’s resources. 

In the period of J.L.P. ascendency in 1967 that party had a majority among the very poor in the metropolitan area primarily through the urban party machines of second-level party leaders, Eddie Seaga and David Tavares who controlled the violence-prone western Kingston and southern St. Andrew area. Predictable defections occurred among the less tightly organized poor communities (that fall outside of the more regimented power domains) in 1972, as the militant populist appeal of the P.N.P. struck chords of responsiveness among the urban poor. Between 1972 and 1976, the P.N.P. consolidated a massive majority among the poorest socio­economic communities by a combination of tightly knit patron-broker­client networks, intimidating violence at the community level, housing allocation and reallocation, and the integration of P.N.P. socialist symbo­lism with the emergent social militancy of the urban ghettoes. The margin is, however, probably inflated by a three to five percent level of fraudulent manipulation of the electoral process in these areas.’0 

The poorest working-class communities reveal a different pattern. There is a secular trend towards P.N.P increased strength that was most likely influenced by the more populist and militant symbolism of the P.N.P. over the period and the broad working-class majority support for the welfarist and socialist objectives and commitment of the P.N.P., particularly since 1974, and the preeminent popularity of the P.N.P. party leader, Michael Manley, up to early 1977 among the majority of the poor non-unionized working class. 

The better-off working class and the lower-middle-class communities have traditionally had a one-party P.N.P. tendency since the mid l950s. This sector of the larger working class formed the early strength of the N.W.U., the P.N.P.’s trade union affiliate, and have warmed to the more intellectual leadership of the traditional P.N.P. They represent that newer generation of labor that is the cream of the skilled part of the labor force. Their political and social attitudes incline them to middle-class aspirations for their children and deep respect for the intellectual leadership of the P.N.P. To them the P.N.P. represents an emphasis on education; leaders with brains and advanced education; radical thinkers who seek to remove the barriers to upward mobility posed by the old class structure; and respectable upper-middle-class professionals who have the knowledge to run government properly

Of interest is the fact that both the better-off working class and the lower-middle-class communities show a significant decline in P.N.P. vot­ing in 1976. Between 1972 and 1976 the P.N.P. vote in the better-off working-class communities dropped by 16.5 percent while the P.N.P. vote in the lower-middle-class communities declined by as much as 24.2 per­cent. Preelection surveys suggest that the decline was due primarily to dis­enchantment with P.N.P. management of the economy among the more privileged strata of the labor force. This lower-class and upper-working ­class decline in P.N.P. support in 1976 was probably influenced by the vocal antigovernment criticism of the P.N.P. by the middle class, which serves as a reference group in some areas for the upward mobility aspiring sections of the lower-middle class and the aristocracy of labor. Equally important is the extent to which the trade unions (including the N.W.U., the union affiliate of the P.N.P.), whose membership includes a large pro­portion of the labor aristocracy, and the lower-middle class have become more and more distant from a P.N.P. government whose policies and priorities seemed concerned more with the interests of the very poor than with the interests of the better-off strata of labor. 

P.N.P. voting is likely to decline even further within this segment of the urban class structure. The next election can be expected to remove all tra­ces of the P.N. P. one-party tendency in the aristocracy of labor that was so entrenched in the voting pattern of the decade of the 1960s. Another important contributing factor to this development is that the newer gener­ation of leftists in the party do not project images of competence, rational thinking and practical command over policy matters in the eyes of this more literate segment of labor. These leftists, on the contrary, are viewed by a growing section of the aristocracy of labor as sources of intimidating but meaningless rhetoric which frightens capitalists, thereby leading to loss of jobs and industrial insecurity and uncertainty. 

The vote pattern in the middle- and upper-middle-class communities is a clear example of extreme voter instability. In the case of the middle-class communities, a slender P.N.P. majority in 1967 develops into a massive P.N.P. majority of 62.8 percent in 1972, and in turn converts to a signifi­cantly large J.L.P. 19.6 percent majority in 1976. To a more significant extent the instability of the party vote in these more exclusive suburban communities reflects their “high elasticity” of response to adverse govern­ing party policies and policy statements. 

The middle-class communities traditionally yield a P.N.P. majority based on the status factor by which the P.N.P. attracted the label of “edu­cation”, respectability, and “brains”, and the J.L.P. that of “ignorance” and “illiteracy” during the early preindependence period. This status factor played an important role in the middle-class vote of the 1950s and cemented P.N.P. party loyalties among the later generation of the middle-class voters who came to adulthood in the 1960s. The middle-class P.N.P. vote has always, however, been lower than the P.N.P. lower-middle-class vote.

The apparent weakening of the top J.L.P. leadership in the period lead­ing up to the 1972 election, the allegations of government party corruption, the abuse of power by the J.L. P. government, the increasing level of crime,~ and the questioning of the government’s record in the area of civil rights all produced an issue vote” which added to the normal P.N.P. middle-class vote majority. By 1976, however, the middle class had begun to react nega­tively to fears of communism, P.N.P. leftist rhetoric, uncontrolled levels of violent and often sadistic crimes, leftist P.N.P. abuse of the middle class, increasing levels of class antagonism that was blamed on P.N.P. party ideology, a sense of being betrayed by the P.N.P. and its leader Michael Manley, and the increasingly closer ties between the governing P.N.P. party and Castro’s Cuba. As a result, the P. N. P. middle-class vote majority vanished in 1976 and the J.L.P. earned a majority similar to the P.N.P.’s 1967 majority.

The upper-middle-class communities which are more affluent and house the capitalist class and high income managerial and professional elite behaved in similar fashion for the same basic reasons. The main differen­ces, being that these more affluent communities are traditionally J.L.P. in orientation but defected to the P.N.P. in 1972 only to return to a solid J.L.P. majority in 1976. 

Of significance is the fact that in 1972 the swings in the vote between these class communities were uniformly in the direction of the winning party, the P.N.P. The central effect of divisive ideological symbols and class issues in the 1976 election was to increase the P.N.P. vote among the very poor and poor working-class communities and to reduce the P.N.P. share of the vote in the middle- and upper-class communities.

In contrast to the more stable though competitive pattern revealed at the community level in the twelve more rural parishes, the urban metropolitan area vote has been shown to be volatile, unstable and responsive to central and salient issues that realign party support among the main socio­economic groupings. Underlying the seemingly stable P.N.P. majorities are wildly fluctuating class voting trends. 

The peasant vote in the rural farm communities of the hilly interior of St. Andrew is indistinguishable from the rural vote in the other parishes. It is highly competitive, yielding small majorities and tends to follow quite pre­dictably whichever party wins an election. It is essentially part of the rural sixty percent of the Jamaican electorate whose competitive voting tradi­tions have kept the two-party system alive. While the metropolitan area vote fluctuates considerably at the local community level, the overall aggregate trend has been an uncompetitive one-party P.N.P. tendency based on the large concentration of strong P.N.P. communities. It is doubtful if many of these concentrated areas of strength will be eroded even by the economic crises of the 1970s although large swings will con­tinue to occur. 

*Note not all the tables have been included.


I.See Carl Stone, Electoral Behaviour and Public Opinion in Jamaica (Kingston, Jamaica: ISER, University of the West Indies, 1974), and “The 1976 Parliamentary Election in Jamaica,” Journal of Commonwealth and Comparative Politics, Vol. 14, No. 2 (1976).

2.See Carl Stone, ~Regional and Community Voting Patterns in Jamaica,” forthcoming in Journal of Jnteramerican Studies and World ~4/jiñrs, (Nov­ember 1978).

3.The sample includes ten polling divisions for each parish which together represents approximately 1,500 to 2,000 voters per parish.

4.Isolation is used here only in a relative sense as all these communities have some contact and communication with other communities. However, their remoteness means that the frequency of such contacts is very low given bad roads, weak communication channels, and distance from urban centers.

They are therefore less “open” than other community types and are more inward-looking.

5.This sample also includes ten polling divisions per parish.

6.This sample includes five polling divisions per parish. These pollingdivisions tend to be twice as large in number of voters as those in the rural areas.

7.The index is made up as follows:“% of community households with more than four persons per room + 9~ of commu nity labor force unemployed.”

The two items were combined due to the fact that they correlated to the extent of a .85 (Pearson R.) correlation coefficient and had mean values of approximately similar size (19.6 and 16.3).

The index was correlated with the percent of the community not exposed to secondary level education and the result was a .88 (Pearson, R.) correlation coefficient. The index is therefore considered to be both valid and sensitive as an indicator of intercommunity socio-economic differences.

8.For a discussion of housing development over the period, see Ann Norton, Shanties and Skiscrapers: Growth and Structure of Modern Kingston

(Kingston, Jamaica: ISER, University of the West Indies, Working Paper No. 13, 1978).

9.In two constituencies, St. Andrew south and St. Andrew southwestern, more persons voted than the names appearing on the voters’ lists. The level of increase in voter registration in some polling divisions was also extraordinar­ily high.

10.See Carl Stone, Electoral Behaviour and Public Opinion.