,the Lok Dal, and the Middle-PeasantCastes: An Analysisof the 1977and 1980 ParliamentaryElections in *

Paul R. Brass

[HE DRAMATICCHANGES that have taken place in Indian politics I in the past fiveyears have theirroots in the post-Independence politicalhistory of northIndia, particularlyof the stateof Uttar Pra- desh,the largeststate in the Indian Union and the home of fourof fiveof the primeministers of the country.It has been widelynoted that the "Emergency"imposed upon by Mrs. Gandhi fortwo yearsbetween 1975 and 1977 affectedthe northIndian statesmore than the south and that Mrs. Gandhi's overwhelmingdefeat in the 1977 electionsand the success of Janata was principallya con- sequenceof the nearlytotal sweep of the parliamentaryseats in that electionin everystate in the north.By the same token,the abilityof the Congressto recoup mostof its lossesin the northIndian Hindi- speakingstates by winning144 seatsthere in 1980,compared to only 2 in 1977,made it possiblefor the Congressand Mrs. Gandhi to re- gain powerat the Center. In fact,it is no exaggerationto say thatthe conflicts for control of

*Researchfor this article was made possibleby a grantfrom the Special CurrencyProgram of the SmithsonianInstitution, which provided funds for me to visitIndia in December 1979. I am especiallygrateful to FrancineBerkowitz of the SmithsonianInstitution for her part in fa- cilitatingarrangements for this grant and the trip.I also owe thanksto CarolynBrown for the use of hervan, driver,and researchassistant-Prahlad Misra-who touredwith me in western U.P.; to Dr. D.P. Singh,Professor and Head, Departmentof Public Administration,University of Lucknow,who providedme hospitalityin Lucknow; and to his children,Ravi and Ma- hendraSingh, who touredwith me in Oudh and the EasternDistricts. Finally, I want to thank JohnWood, Allen Kornmesser,and an anonymousreviewer for Pacific Affairs, all of whom read the firstdraft of thisarticle and made detailedcriticisms and suggestionsthat substantially af- fectedthe finalresult. The responsibilityfor statements of fact,interpretation, and opinionis, of course,my own. 5 PacificAffairs the governmentof India that have been going on forthe past five years,during which the Emergencyand the 1977 and 1980 elections have been dramaticfocal points, have been almostentirely a playing- out on the national stage of social and political conflictsthat have theirorigin in the northIndian statesof Uttar Pradesh and Bihar, whichcontain between them approximately 150 millionpeople. The leading political actors and forceswho have articulatedthose con- flicts,excepting Moraji Desai, all have come fromU.P. and Bihar: Mrs. Gandhi, ,Raj Narain and H.N. Bahuguna from U.P.; JayaprakashNarayan and JagjivanRam fromBihar; and the Jan Sangh, with its principalstrength in U.P., Bihar, and Madhya Pradesh.The issuesthat have overlaidthe strugglefor power at the Centeralso have derivedlargely, though not exclusively,from per- sistentpolitical and social problemsin thenorth: Hindu-Muslim con- flictand violence,widespread student unrest, the differentialimpact ofthe Green Revolution on regionsand ruralsocial classes,the failure of industrializationto provide off-farmemployment for the rural poor,and the consequentincreasing discontent of the low castes and landlesslaborers in the countryside. The significanceof the northin the Emergencyand in the 1977 and 1980 electionshas been widelynoted, but analysisof the sources of its importancein the social and political conflictsof the region have been neglectedin the undue and misplacedattempts to read intothe election results a greatvictory for Indian democracyor to ex- plain themin termsof the specificissues articulated during the cam- paigns.The 1977 electionresults were hailed by mostobservers of In- dian politicsas a virtuallyunprecedented restoration of a democratic systemby popular vote and as a confirmationof the deep com- mitmentof the Indian populace,rich and poor alike,to the values of democracyand parliamentarism.Specific factorsassociated with Mrs. Gandhi'sregime also werepinpointed as particularlyimportant in her defeat: the sterilizationcampaign, demolitionof squatter houses,the curtailmentof civil libertiesin general,and the increased prominenceof Mrs. Gandhi's son, Sanjay. Similar commentswere made about the 1980 elections.Once again, journalistspraised the Indian public fortheir commitment to democracyby throwingout yetanother unpopular regime. Again also, specificfactors were iden- tifiedas especiallyimportant to the result-namely,prices and short- ages of basic commoditiesand the discontentof the Indian public withthe "politicsof defection." Yet, clearlythe more sweepinginterpretations of the 1977 and

6 The 1977 and 1980 ParliamentaryElections in UttarPradesh 1980elections are wrong:since Mrs. Gandhi neverexpressed any gen- uine regretsfor imposing the Emergencyregime, it is not possibleto see bothelections as expressingdeep commitmentby Indian votersto democracy.As forthe morespecific explanations, they are not wrong, only superficial:they ignorethe deeper patternsrevealed in these electionsthat have developedduring the past thirtyyears of north In- dian politicaland social historyand that set the stage forthe Emer- gencyand its overthrow,as well as forthe returnof Mrs. Gandhi to power. The purposeof this article is to examinethese underlying patterns ofbehavior of the leading social forcesin U.P. in theelections of 1977 and 1980 and to trace the developmentof the social conflictsthat were expressedin thosetwo elections.Although this article focusses on thestate of U.P., manyof the patterns described here apply also to Bihar.' The remainderof the articleis divided into fourparts. The firstpart summarizes the main featuresof the electoral history of U.P. and offersan explanationfor the rise of the and its sweep of the 1977 elections.The second part examinesthe break-up of theJanata coalitionand the settingof the stage forthe 1980 elec- tions.The thirdpart presentsa detailed analysisof the electoralhis- toryof fiverural constituencies in U.P. The conclusionsare presented in the fourthpart.

THE ELECTORAL HISTORY OF U.P. AND THE RISE OF THE JANATA PARTY Table 1 presentsthe resultsof the seven parliamentaryelections in U.P. Three featuresof thoseresults should be noted.The firstcon- cernsthe overall decline of supportfor the Congress:having achieved its peak electoralsupport in 1952,when it polled 53 per cent of the vote,it began to declinethereafter to 46 per centin 1957,38 per cent in 1962,and to less than one-thirdof the popular vote in 1967. Al- thoughthe Congresspopular vote shareincreased markedly in 1971, it droppedeven more severely in 1977and thenreverted to somewhat above one-thirdof the popular vote in 1980.It is especiallyimportant to note that the decline of the Congresswas mostpronounced in a tractof eightagriculturally prosperous districts near Delhi, between the Ganges and the Jumna rivers,known as the Upper Doab. This

I See HarryW. Blair,"Rising Kulaks and BackwardClasses in Bihar: Social Change in the Late 1970s,"Economic and Political Weekly, XV, no. 2 (January12, 1980),pp. 64-74, and Prad- han H. Prasad, "Rising Middle Peasantryin NorthIndia," Economicand PoliticalWeekly, XV, nos. 5, 6, and 7, Annual Number(February 1980), pp. 215-19. 7 TABLE 1. PERCENTAGE OF VOTES POLLED BY POLITICAL PARTIES IN UTTAR PRADESH PARLIAMENTARY ELECTIONS, 1952-1980

PoliticalParty* 1952 1957 1962 1967 1971 1977 1980 Congress 52.99 46.29 38.20 33.04 48.56 24.99 35.90 Congress(0) (Foundedin 1969) 8.58 - - SocialistParty 12.94 5.67 8.64 - 0.24 - - Kisan Mazdoor Praja Party (KMPP) 4.90 - - - Praja SocialistParty (PSP) 15.34 10.35 3.74 0.23 SamyuktaSocialist Party (SSP) (Foundedin 1964) 10.27 4.10 Jan Sangh 7.29 14.79 17.57 22.58 12.28 Hindu Mahasabha (HMS) 1.91 - 1.39 - 0.07 0.12 ScheduledCaste Federation/ RepublicanParty of India (RPI) 2.88 - 4.27 4.07 - - 0.16 CommunistParty of India (CPI) 0.35 1.67 3.63 3.75 3.70 1.10 1.63 CommunistParty of India, Marxist(CPM) (Foundedin 1964) 0.71 0.19 0.10 0.10 Swatantra (Founded in 1959) 5.04 4.77 0.05 BharatiyaKranti Dal (BKD) (Founded in 1967) 12.70 Janata/BharatiyaLok Dal (BLD foundedin 1974,Janata in 1977) 68.08 22.57 (BLD) Lok Dal (Founded in 1979) 29.02 Independentsand Unsuccessful Parties 16.76 16.24 10.91 17.08 9.31 5.61 10.62 Total 100.02 100.00 100.00 100.01 100.01 100.00 100.00 Sources:Compiled fromofficial reports of the ElectionCommission of India and * fromG.G. Mirchandani,The People's Verdict (New Delhi: Vikas, 1980). All partiesthat won at least one seat in one electionare included.The KMPP also is includedin 1952,though it did not win a seat, to facilitatecom- of parison the 1957 PSP votewith the votesof its two pre-mergerprogenitors, the SP and the KMPP. The figurefor the Socialist Partyin 1957 is forthe Lohia Socialists,who brokeaway fromthe PSP and foughtthe 1957 electionsas Independents. The 1977 and 1980 ParliamentaryElections in UttarPradesh tractis also the home of ChaudhuriCharan Singh and the centerof his Jat caste in U.P., the leading owner-cultivatorcaste not only of the Upper Doab, but of the adjoiningstate of Haryana as well. Fi- nally,it is also themost politicized region of thestate with the highest turnoutrate in everyelection. In the parliamentaryconstituencies of thesedistricts, the median vote share forthe Congressdeclined from 58.2 per cent in 1952 to 30.3 per cent in 1980. In 1952,the Congress won everyseat in thesedistricts. In 1980,the Congress won only2 out of 14,whereas the Lok Dal won 11 and Janata 1. The leading causes of discontentin this regionarose fromtwo sources:the continueddominance of politicsand patronagein most of thesedistricts by theelite Brahman and Rajput castesin the coun- trysideand by urban tradingcastes in the towns;and the failureof the governmentto providedesired incentives of capital inputs and pricesto increasethe profitability of agriculture for the main cultivat- ingcastes, especially the Jats. From 1957 until 1969,the discontentof themiddle proprietary castes in thisregion was expressedin the form of votingfor independents and othersmall parties;2but no singleop- positionparty was able to organize this discontentuntil Charan Singh defectedfrom the Congressin 1967 and foundedin 1969 the BharatiyaKranti Dal (BKD) as a partyof agrarianinterests, repre- sentingthe peasant proprietors. A second noteworthyfeature of the electoralhistory of U.P. has been the relativeweakness of the radical leftistopposition. However, the Socialists,particularly the radical wing,did establisha consistent base ofsupport that stabilized around the 10 per centlevel in thecen- tral and easterndistricts of the statewhich, in contrastto west U.P. and the Upper Doab, have been agriculturallyless advanced and have a verylarge population of marginal landholders and landlessla- borers.In thesedistricts, as elsewherein thestate, the Congress organ- izationin the 1950sand 1960swas dominatedby theelite proprietary castes of Brahmansand Rajputs. The radical wing of the Socialists here appealed, with considerablesuccess, to those who were either excluded fromor werejunior partnersin the Congresscoalition- namely,the middle cultivating castes of Ahirs (Yadavs)3 and Kurmis; 2 Space considerationspreclude the presentationof the regionaldata on which this and otherstatements about regionalvoting behavior are based. However,for regional voting pat- ternsfor the U.P. LegislativeAssembly elections from 1952 to 1977,which are similarin impor- tant respectsto the parliamentaryresults, see Paul R. Brass,"The Politicizationof the Peas- antryin a NorthIndian State: Part I,"Journalof Peasant Studies, VII, no. 4 (July1980), pp. 395- 426 and "Part II," in ibid,VIII, no. 1 (September1980), pp. 3-36. 3 Ahirsand Yadavs are interchangeableterms for the same caste categoryin U.P. The term "Yadav" is used virtuallyexclusively in Bihar,more often than not in easternU.P., and less of- 9 PacificAffairs a higherstatus proprietary caste, the Bhumihars;and the landlessla- borers. The thirdimportant feature of the electoralhistory of U.P. has been therise of the Jan Sangh as theleading and best-organizednon- Congressparty in thestate, with a popularsupport base above 22 per centin the 1967 elections.The Jan Sangh was a partyof militantIn- dian nationalism,which drew its symbolsfrom the Hindu religion and the Sanskritand Hindi languages,and whosestrength was con- finedto northIndia, includingPunjab, Haryana,U.P., Madhya Pra- desh,and Bihar. It, and especiallyits principalorganizational prop, the RashtriyaSwayamsevak Sangh (RSS), werewidely considered to be anti-Muslim,as theywere usually opposed to educationaland lin- guisticconcessions demanded by the Muslim minority.In U.P., the Jan Sangh-like the Congress-drewits main supportfrom the elite castes of Brahmans,Rajputs and Kayasthas,and fromthe Vaishya castes in the towns. Moreover,like other non-Congressparties in U.P., it developed a strongregional support base-in this case, the centralplains districts of Oudh, once thehomeland of the greatland- lordsof U.P., knownas the talukdars.For some time the partydrew heavilyupon thesupport and resourcesof several of the talukdarifam- ilies,but eventuallyit succeededin establishingitself more broadly in thisregion as the principalvoice forthe leadingproprietary commu- nities,including many fromamong the backwardcultivating castes of Ahirs (Yadavs) and Kurmis. In Oudh as well as other regions, then, the Congresssupport base among the importantland-con- trollingcommunities had been eroded as earlyas the 1960s. Beforewe can comprehendthe social significanceof the rise of the BKD and theJanata party,we need to examinethe consequences for the Congresssupport-base of the partysplits that occurredbetween 1967 and 1969. In 1967 Charan Singh leftthe Congress.For many yearshe had been the principalspokesman in the U.P. government on behalfof rural,peasant interestsand values and had developed a networkof relationshipsin the U.P. districtsamong the middle-caste groupsin thestate, particularly Jats and Yadavs. Followinghis defec- tion fromthe Congressin the U.P. legislature,he joined forceswith theopposition to formand lead the firstnon-Congress government in the historyof the province.Although only seventeen persons left the Congresswith him in April 1967,Charan Singhhad thebroader sup- ten in Oudh and westernU.P., wherethe term"Ahir" is stillcommon. Ahir is the traditional censusname forthis caste, and Yadav the preferredname used by politicizedand sociallymo- bile segmentsof the caste.

10 The 1977 and 1980 ParliamentaryElections in UttarPradesh portof the middle proprietarycommunities, particularly in western U.P. This factwas demonstratedby thedramatic rise of his party,the BKD, to second positionbehind the Congressin the 1969 legislative assemblyelections and again in the 1971 parliamentaryelections, by itsemergence as theleading party in theUpper Doab in the 1969 leg- islativeassembly elections (with 31 per cent of the vote),and by the strongpositive correlations between its vote share and the percentage of landholdingsabove fiveacres in theplains districtsof thestate as a whole.4 The defectionof Charan Singh in 1967 was farmore significant forthe futureof the Congressin U.P. than was the major splitwhich occurredin the Congressin 1969.Also, because of the criticalimpor- tance of U.P. nationally,it was to be verysignificant for political de- velopmentsin the centralgovernment in the 1970s.Insofar as U.P. is concerned,the Congresssplit of 1969 involvedthe departureof the Gupta faction,a largelyurban-based political machine, but one with connectionsalso to importantlanded groups-particularlyRajputs- in the countryside.After the split,Mrs. Gandhi attemptedto reab- sorb the Charan Singh forcesinto the Congress(I) and even sup- porteda minorityBKD governmentled by him in 1970. However, whenCharan Singh refused-orproved unable to deliverthe parlia- mentaryBKD delegation-to supportthe Congressgovernment at the Center,Mrs. Gandhi withdrewthe supportof the Congressin U.P. fromCharan Singh'sgovernment. Gandhi's landslidevictory in the 1971 parliamentaryelections made it unnecessaryfor her to seek an alliancewith Charan Singhthereafter. From 1971 onwards,there- fore,she soughtto maintaincontrol of U.P. throughthe remnantof the Congressorganization in the state,which was leftin the hands primarilyof Brahmanpoliticians and theirallies. However,it deservesto be notedthat, despite the defectionof the Charan Singh forcesbefore the 1969 electionin U.P. and the loss of the Gupta factionthereafter, the total support for Mrs. Gandhi's Congresswas consistentat one-thirdof the popular vote in both the 1969and 1974state assembly elections and, afterthe dramatic fluctu- ations of the 1971 and 1977 parliamentaryelections, remained at about that same level in the parliamentaryelections of 1980. How can thisstability of the Congressvote-albeit at a low level-be ex- plainedin theface of theerosion of its base amongthe middle propri- etarycastes? The Congresswas able to hold onto its one-thirdshare

4 Brass,"Politicization of the Peasantry,"I, p. 419 and II, pp. 18-20.

11 PacificAffairs of the totalvote partly by retainingits supportamong the influential ruralBrahman and some Rajput castes,and partlyby drawingupon varied and heterogeneousgroups of supportersthat any large politi- cal partyis bound to collectbecause of its superiororganization and, in the case of the Congress,its controlof governmentpatronage. At the same time,the partymanaged to increaseits supportamong the ruralpoor, the landlessand the Scheduled Castes,by such measures as allottingland to Scheduled Castes forcultivation and house sites, establishingthe Small Farmers'Development Agency and the Mar- ginal Farmerand AgriculturalLabour Agency,and abolishingforced labor. Thus, underMrs. Gandhi's domination,the Congressin U.P., whichhad been alwaysconceived by both scholarsand journalistsas a great,broad-based party of the Center,became a party of ex- tremes-but ofopposite extremes-comprising within its foldthe old dominantlandlord and leading proprietarycommunities of Brah- mans and Rajputs and the rural poor and landless. Largely dis- affectedfrom the Congress was the band of middle cultivating groups,generally less influentialand less widespreadthan the Brah- man and Rajput groups,but forminga broad and diversealliance of varyinglocal importance. Needlessto say,the above descriptionunderstates the complexity of the supportbases of the various political partiesin U.P., which have oftenvaried fromconstituency to constituencyand electionto election,depending upon thestate of local rivalries,the castes of local candidates,and the interrelationshipsamong local, provincial,and nationalleaders. Nevertheless, I believe it to be valid as an accountof the core regionaland caste supportbases of the U.P. partiesthrough the 1969 elections.Between 1969 and 1977, two steps were taken whichbuilt upon the shiftsin partysupport bases noted above and ultimatelyled to the greatJanata sweep in the 1977 elections.The firstwas the dismantlingof the Socialist movementin U.P. princi- pally by Raj Narain5and the decisionto mergeits most vital seg- ment,the Samyukta Socialist Party (SSP), withCharan Singh'sBKD in 1974.This mergerhad theeffect of givingthe BKD morebalanced supportbetween western and easternU.P. In the process,though not entirelybecause of the alliance withthe SSP, the BKD-by now re- named the (BLD)-extended its support base among the middle cultivatingcastes, particularly among the Ahirs

5 See Paul R. Brass, "Leadership Conflictand the Disintegrationof the Indian Socialist Movement:Personal Ambition, Power and Policy,"Journalof Commonwealth and ComparativePoli- tics,XIV, no. 1 (March 1976), pp. 19-41. 12 The 1977 and 1980 ParliamentaryElections in UttarPradesh (Yadavs), who are thelargest cultivating caste in thecentral and east- ern districtsof the state. The BLD also succeeded in winningsome supportin areas wherelow-caste groups are concentrated. The secondstep toward the grandelectoral sweep of 1977was the formationof the Janata partyitself which, insofar as U.P. was con- cerned,meant the addition to the BLD of the Jan Sangh strength. Since the Jan Sangh supportbase was strongestin the centraldis- trictsof the stateand among the middle and richpeasantry, the Ja- nata was now in a positionthat only the Congresshad occupied be- fore. It had become a coalition with strongand relativelyeven supportacross the length and breadthof thisvast and populousstate. Not only had it gained a solid base of supportamong particular castes,particularly Jats and Ahirs(Yadavs), but it also had gathered into its fold large segmentsfrom the middle and rich peasantryof othercommunities. Moreover, it receivedsome supportfrom Sched- uled Castes and Muslimsbecause of developmentsduring the Emer- gencythat turnedmany fromthese groups away fromthe Congress and towardsJanata-particularly, the demolitionof squatterhouses, the rumorsabout forcedsterilizations, the killingof Muslimsin Mu- zaffarnagar,and the call by the Shahi Imam of theJama Masjid in Delhi to supportthe Janata party.However, the principalsocial sig- nificanceof the Janata partyin U.P. was thesuccessful formation of a broad-basedparty of the Centerwith the middleand richpeasantry at its core.

THE BREAK-UP OF THE JANATA COALITION AND THE PRELUDE TO THE 1980 ELECTIONS The Janata victoryin the 1977 electionswas built principally upon a coalitionof threemajor social forces-the middle and rich peasantry,the Muslims,and the Scheduled Castes who had become disaffectedfrom the Congressduring the Emergency.That coalition broke apart because of both externalpressures and internalcon- tradictionsthat have theirsources in the social fabricof northIndia and the multiplecleavages of thatsociety with which coalition lead- ersmust cope in orderto succeed.The Janata retainedits central core of supportamong the peasantrythroughout its two-yearrule; but such supportdepended largely on the identificationof the peasantry with Charan Singh, the leading advocate of and most articulate spokesmanfor this much-maligned social category.The depthof that supportwas clearly demonstratedduring the period that Charan Singh withdrewfrom the Janata government,between June 1978 13 PacificAffairs and January1979. In December1978, in thelargest and mostimpres- sive demonstrationever carried out in Delhi fora politicalcause, five millionkisans (peasants) expressed their firm commitment to Charan Singh and his pro-peasantpolicies. As Home Ministerbefore his exit from the Cabinet, Charan Singh was not able to accomplishmuch of significancefor the peasantrydi- rectly,even thoughhis ideas providedmuch of the guidingforce be- hind theJanata government'spolicies in favorof agricultureand ru- ral industries.Throughout his tenurehe remaineddissatisfied with the overall budget allocation foragricultural research and produc- tion-relatedmeasures, which he thoughtshould be doubled. He also failedto persuadethe governmentto supporta higherprice for sugar cane. His advocacyof such measures,however, was well knownto his supportersand made his withdrawalfrom the governmentappear to be a consequenceof its failureto do enoughfor the peasantry. Afterhis reinductioninto the Cabinet as Deputy PrimeMinister and Finance Minister,Charan Singh was able to implementtwo measuresof directbenefit to the peasantry.One was a reduction,by 50 per cent,of the excise dutyon artificialfertilizer. The otherwas the transferof the excise duty on tobacco fromthe growerto the manufacturer,a move whichbenefited mainly the kisansin southIn- dia. Both these measuresoccurred in a budget that otherwisebore down heavily-particularlyon the urban sectors-withexcise duties and othertaxes on consumergoods. In theseand otherways, Charan Singhand his followersmade clear theirconviction that cities tended to get betterbreaks in such mattersas taxation,resource allocation, pricepolicies and availabilityof credit-and showedtheir determina- tion to do somethingabout it. Charan Singh'sbase among the middlepeasantry of northIndia was also reinforcedby theselection of people frommiddle and "back- ward" castes6as chiefministers of threenorth Indian states-Chaud- huri Devi Lal, a Jat, in Haryana; Ram Naresh Yadav, an Ahir,in U.P.; and Karpuri Thakur,from a caste of barbers(Hajam), in Bi- har. These stategovernments in turntook measures on behalfof agri- culturaldevelopment, and in supportof the backwardcastes in par- ticular,that ensuredcontinued support for Charan Singh and his

6 term"backward" castes is commonlyused in politicallanguage in India to referto the middle castes,whether or not theyare landed, landless,or come fromartisan and service castes.In general,I preferthe moreneutral term "middle castes" forthese groups, and "middle proprietarycastes" forthe landed groupsamong them;but sometimesit seemsappropriate to use the term"backward" castes also, as herewhere the referentsare to both the middle,typi- cally landed castesof Jats and Yadavs, and the lowermiddle artisancaste of barbers.

14 The 1977 and 1980 ParliamentaryElections in UttarPradesh allies in the north.Some of thosemeasures-such as the Bihar policy of reservinga percentageof places in governmentservice and in schools and colleges forthe backwardcastes-were also quite con- troversialand drovethe wedge moredeeply between the upper-caste groupsand the middlecultivating and artisancastes. Duringhis brieftenure as PrimeMinister of India, Charan Singh was not able to do much more forthe peasantry.However, he sup- portedthe establishment of a veryhigh cane-price in northIndia that partiallyoffset the considerablelosses because of the poor monsoon rains in the summerbefore the 1980 elections.This measurecould not have been more timely:the electionsof January 1980 were held preciselyduring the periodthat cane is harvestedin northIndia. Though theJanata-or ratherCharan Singh-was able to retain the supportof the middlepeasantry, it had alreadylost its hold over the other two elementsof its 1977 coalition even beforethe dis- integrationof the coalitionat the leadershiplevel in July 1979. Mus- lim supportwas weakenedby the failureof the centralgovernment and the governmentsin U.P. and Bihar to make any significantcon- cessionsconcerning the use of the Urdu language,by the slownessof thecentral government in framinga new AligarhMuslim University Act thatwould guranteeethe futureMuslim character of the Univer- sity,by the appointmentof a non-Muslim,Minoo Masani, to head the new MinoritiesCommission established by the Janata govern- ment,and by an increasedincidence of Hindu-Muslimcommunal riots.The supportof the ScheduledCastes was weakenedby the fail- ure of theJanata leadershipto name JagjivanRam as Prime Minis- terand, moreimportantly, by the growingattention given to the ten- sion betweenScheduled Castes and middle proprietarycastes in the countryside,especially in northand centralIndia. The pressand Mrs. Gandhi were givingwide publicityto severalparticularly violent in- cidentsin whichScheduled Castes werekilled at Belchhiin Bihar,at Pantnagarin Agra in U.P., and at otherplaces.7 Althoughthe collapse of the Janata coalitionwas due primarily to competitionamong the top leadersfor preeminence and the Prime Ministershipand strugglesfor power among the originalparty con- stituents,social contradictionsin the coalition also were important duringthe severalcrises that occurred during its two yearsin power. Of particularimportance in this regardwas the conflictfor control over the chiefministerships of the northIndian states,in which the

7For descriptionsof these and otherincidents, see Udayan Sharma,ed., ViolenceErupts (New Delhi: Radha KrishnaPrakashan, 1978). 15 PacificAffairs chiefprotagonists were the formerBLD groupof Charan Singh and the formerJan Sangh group.As already noted,Charan Singh and the BLD were able to place in poweras chiefministers in Haryana, U.P., and Bihar personsfrom backward castes. Eventually, however, all threewere displaced in a continuingstruggle for power. Throughoutthese strugglesfor control over the north Indian statesand over theJanata partyorganization, issues concerning the relativepolitical and economicstatus of elite,backward, and Sched- uled Castes and concerningHindu-Muslim relations were ever-pres- ent.Eventually, the followers of Jagjivan Ram tendedto allywith the Jan Sangh in conflictswith Charan Singh'sBLD group.Members of the Scheduled Castes,who oftensuffered at the hands of dominant peasant-proprietorgroups from the backwardcastes, were clearly op- posed to any measuresenhancing the influenceof thosewhom they alreadysaw as locallypowerful and whose politicalspokesmen were theirrivals for power in statepolitics. Consequently, they sought an alignmentwith the Jan Sangh, both because theyhoped thatJagji- van Ram would becomePrime Minister with Jan Sangh supportand because theyopposed concessionsto the backwardcastes. For their own part,elite-caste followers of the formerJan Sangh objected to the BLD group'spolicies favoring reservation of jobs in servicesand collegesfor backward castes, as decreasingtheir own opportunitiesfor advancement.The BLD leaders,caught in a crossfirebetween both ends of the caste order,attempted to divertattention from the back- ward caste issue to the Hindu-Muslimissue, by describingthe con- flictas an attemptby the RSS elementsin theJanata to gain control of the partyand destroyits secularcharacter. It is a measureof the significanceof thesetwo contradictionsin northIndian society-in the caste orderand in Hindu-Muslimrela- tions-that as theJanata coalitionfell apart sociallyas well as politi- cally,the Congressof Mrs. Gandhi rebuiltits old coalitionof the late 1960sby capitalizingon the discontentof the elitecastes, the Sched- uled Castes, and the Muslims.It is movementback and forthfrom withinthese three broad groupingsthat largely explains the dramatic shiftsin votingfrom 1971 to 1977 and from1977 to 1980.

FIVE CONSTITUENCIES IN UTTAR PRADESH The developmentof thesocial contradictionsdiscussed above, the ways in whichthe politicalparties have been affectedby these con- tradictionsand have triedto controlthem to theirown advantage, 16 The 1977 and 1980 ParliamentaryElections in UttarPradesh and theways in whichsuch phenomenahave been perceivedby vot- erscan bestbe broughtout by a detailedexamination of several rural constituenciesin U.P. Five constituencieshave been selected,drawn fromfour regions of the state.' (The detailedelection results for each of the constituenciesfor 1977 and 1980 are givenin Table 2.) Each constituencychosen illustratesa differentaspect of the main social conflictsthat have been prominentin U.P. politics,but thereare also certainpatterns common to all. The historyof each constituencywill be discussedand thenfollowed by an analysisof the strategiesof the partiesin the 1980 electionsand of the likelyvoting behavior of par- ticularcategories of votersbased on interviewswith voters, non-vot- ers,and knowledgeableobservers carried out in December 1979 be- forethe elections.The interviewswere not random in the scientific sense and do not providea basis foran accurate account of the de- tailed votingresults in each constituency.They have been used to providesome insightinto the ways in which personsfrom different castes,communities, and economiclevels perceive the partiesand the social conflictsin U.P. The interpretationsof eventsin each constitu- ency,therefore, which are closer to guessesthan to scientificjudg- ments,are lessimportant than the broaderhistorical patterns and the similaritiesin the statementsof categoriesof votersacross the several constituencies. NainiTal Constituency Naini Tal constituencycomprised in 1980 all of Naini Tal dis- trict,including both plains and hillsregions, and a small part of Ba- reillydistrict. The districthad a populationof 790,080in 197 .9 The constituencyhad an electorateof 645,203 in 1980. There is onlyone town in the districtwith a population above 50,000-Haldwani. Nearly 18 per centof the populationof the districtin 1971 was from Scheduled Castes, and another6.65 per cent was fromScheduled Tribes.There are severalunusual social and demographicfeatures of the districtthat are relevantto itselectoral behavior. One is the very

8 Four ofthe five constituencies are in districtsin whichthe field research was done in 1961- 62 thatled to myFactional Politics in an IndianState: The Congress Party in UttarPradesh (Berkeley: Universityof CaliforniaPress, 1964). I have also done previouswork in Naini Tal on another project.Consequently, the analysesand judgmentsmade about the influencesof varioussocial forcesin the 1980elections are based upon informationcollected in interviewsin both 1961-62 and 1979,and also in occasionalinterviews and meetingswith politicians from these districts in the interveningyears. 9Contemporarydemographic data forthis section were derived from Government of Uttar Pradesh,Election Directorate, Population of UttarPradesh (Allahabad: Superintendent,Printing and Stationery,U.P., 1973). 17 TABLE 2. ELECTION RESULTS FOR FIVE PARLIAMENTARYCONSTITUENCIES IN UTrAR PRADESH, 1977 AND 1980 j." Percentage Votes ofTotal Constituency!Year! (Turnout percentage) Candidate Caste Party Polled ValidVotes Naini Tal 1980 (51.68) N.D. Tewari Brahman Cong. (I) 163,117 50.58 BharatBhushan Agarwal Janata 58,695 18.20 Pratap Singh Rajput Lok Dal 49,506 15.35 8 OtherCandidates 54,069 15.86 Total Valid Votes 322,387 99.99 Naini Tal 1977 (60.45) BharatBhushan Agarwal BLDa 196,304 61.70 K. C. Pant Brahman Cong. (I) 111,658 35.09 2 OtherCandidates 10,208 3.21 Total Valid Votes 318,170 100.00 Aligarh 1980 (50.22) Indra Kumari Rajput Lok Dal 128,353 38.49 GhanshyamSingh Rajput Cong. (I) 110,375 33.10 SangramSingh Rajput Janata 61,158 18.34 23 OtherCandidates 33,613 10.08 Total Valid Votes 333,499 100.01

Aligarh 1977 (64.79) Nawab SinghChauhan Rajput BLD 280,811 70.85 GhanshyamSingh Rajput Cong. (I) 90,053 22.72 4 OtherCandidates 25,492 6.43 Total Valid Votes 396,356 100.00

Baghpat 1980 (70.34) Charan Singh Jat Lok Dal 323,077 65.21 R.C. Vikal Gujar Cong. (I) 157,956 31.88 Dhara Singhb Gujar Janata 3,843 0.78 8 OtherCandidates 10,591 2.14 Total Valid Votes 495,467 100.01

Baghpat 1977 (74.83) Charan Singh Jat BLD 286,301 63.47 R.C. Vikal Gujar Cong. (I) 164,763 36.53 Total Valid Votes 451,064 100.00 TABLE 2-Continued

Percentage > Votes Total Constituency! of Year/(Turnoutpercentage) Candidate Caste Party Polled ValidVotes to Gonda 1980 (35.18) Anand Singh Rajput Cong. (I) 125,196 52.29 KaushalendraDatt Brahman Lok Dal 52,270 21.83 Satya Deo Singh Rajput Janata 38,849 16.23 7 OtherCandidates 23,106 9.65 Total Valid Votes 239,421 100.00

Gonda 1977 (45.25) Satya Deo Singh Rajput BLD 157,963 59.76 Anand Singh Rajput Cong. (I) 86,690 32.80 Gopal Chand Not Known Ind. 19,656 7.44 Total Valid Votes 264,309 100.00 Deoria 1980 (46.68) Ramayan Rai Bhumihar Cong. (I) 110,014 32.83 Ram Dhari Shastri Sainthwar Lok Dal 109,937 32.81 Ugra Sen Rajput Janata 81,337 24.27 5 OtherCandidates 33,823 10.09 Total Valid Votes 335,111 100.00

Deoria 1977 (55.12) Ugra Sen Rajput BLD 258,864 77.15 Vishwa Nath Brahman Cong. (I) 76,691 22.85 Total Valid Votes 335,555 100.00

Sources:1980 election returns are provisional and wereprovided through the courtesy of the ElectionCommission of India; 1977returns are fromGovern- mentof India, ElectionCommission, Report on the SixthGeneral Election to the House of the People in India,1977, vol. II (Statistical)(Delhi: Controllerof Pub- lications,1978). a = BLD BharatiyaLok Dal, whichwas theJanata Partyin U.P. in 1977. b Dhara Singh withdrewtowards the end of the campaign in favorof the Congress(I) candidate. PacificAffairs highproportion of elitecastes to the totalHindu population:accord- ing to the last caste censusin U.P. in 1931,the Brahmancastes com- prised19 per centof thepopulation, the Rajputs 26 per cent.'0A sec- ond demographicfeature is the veryhigh proportionof migrantsof thetotal population, composed of refugees from the PakistanPunjab, veterans,"political sufferers"" who have settledon tarailand re- claimedfrom jungle sinceIndependence, and tensof thousands of ag- riculturallaborers from eastern U.P. and Bihar. A thirdimportant featureof the district is thepresence of the Govind Ballabh Pant Uni- versityof Agricultureand Technology,which has been a leading forcein spreadingthe GreenRevolution in northIndia. It is also the largestlandholder in the region,having 16,000acres of its own, in- cludinga highlymechanized seed farmemploying some two to four thousandmostly migrant laborers. In April 1978,a labor disputeon the universityfarm led to an incidentin whichat least fifteenlabor- ers were killedby police gunfire.'2The killingswere politicallyvery significantat the time,since the Vice-Chancellorwas a Jat and was consideredto be close to Charan Singh (also a Jat) who was then Home Ministerin the governmentof India. The laborerskilled were presumedto be mostlylow- and lower-middlecaste persons, and Mrs. Gandhi lostno timerushing to thescene to identifywith them. At the same time,the incidentwas widelyreported to have been precipi- tated by Jat policemenantagonistic to the lowercaste laborersfrom easternU.P. and Bihar.On theother side of theissue, the bigger tarai farmersin the surroundingregion, especially those who themselves employedlarge numbersof laborers,were said to be sympatheticto the universityauthorities in the labor dispute,which centeredon a wage issue.Although many of thesetarai farmers received their lands fromprevious Congress governments, they now oftenidentify with differentsplinters of theold Congressor withparties other than Con- gress. Politically,Naini Tal constituencyhas been consistentlya Con- gressand a Brahmanstronghold, the only exceptionbeing the 1977

10 Censusof India, 1931, vol. XVIII: UnitedProvinces ofAgra and Oudh, Part II, by A.C. Turner (Allahabad: Superintendent,Printing and Stationery,1933), table XVII. " The term"political sufferers"refers to personswho reportedlysuffered some personal hardshipduring the Independencemovement and were,therefore, rewarded by the Congress governmentafter Independence. 2 On this incidentand its broadersignificance, see Sharma, ViolenceErupts, pp. 107-118, and Paul R. Brass, "InstitutionalTransfer of Technologyin Agriculture:The Land Grant Model and the AgriculturalUniversity at Pantnagar,"in RobertS. Anderson,et al., eds., Sci- ence,Politics, and the Agricultural Revolution inAsia (Boulder,Colorado: WestviewPress, forthcom- ing). 20 The 1977 and 1980 ParliamentaryElections in UttarPradesh elections.A BrahmanCongressman, C.D. Pande, won both the 1952 and 1957 elections.In the next three elections-1962, 1967, and 1971-the successfulcandidate was K.C. Pant (also a Brahman) the son of GovindBallabh Pant, and a ministerin Mrs. Gandhi's govern- mentfor some time.In 1977,Bharat Bhushan,from a businesscaste and formerlya memberof the Congress(0), won the seat on the BLD ticketwith 61.7 per cent of the vote. In 1980,the seat reverted again to a Brahman Congressman,Narain Dutt Tiwari, a former ChiefMinister of U.P. duringthe Emergency. Interviewsin thisconstituency in December 1979 suggestthat the Congressbase of supportwas principallyamong Brahmansand low- and middle-castelaborers, whereas the Lok Dal, whichcame in third witha Thakur candidate fromthe hills,was strongamong Thakurs and kisansin general.The Janata candidate,Bharat Bushan (who ul- timatelycame in second) was being supportedby the formerJan Sangh cadres,had some supportfrom persons of elite-castestatus in the plains, and was thoughtto have some supportfrom Scheduled Castes because ofJagjivan Ram's leadershipof Janata. The two most impressivefactors that emergedfrom these inter- viewswere kisan support for the Lok Dal and open statementsfrom Brahmanand laborervoters that they wanted the Emergencyback or that thingswere betterduring the Emergency.It was also reported that the big farmersin the area, many of whom were membersof a local kisanassociation, would supportthe Lok Dal. Reasons givenfor supportingthe Lok Dal includedspecific approval forthe economic policiesof Charan Singhand disapprovalof 's policy of favoringenforcement of land-ceilingslegislation. Sentimentin favorof the Emergencywas expressed,for example, by a Brahmankisan, who claimed that the governmentdid more for poor people duringthe Emergency and thatpolice did a betterjob of catchingdacoits (armed robbers). In a similarvein, laborers in a labor colonybeside the Agricultural University farm claimed that the farm laborershad receivedwage increasesduring the Emergency that were not implementedunder Janata, and that prices were lower then. They also complainedabout theshortage of kerosene. Moreover, they were reportedlyagainst the Janata candidate because of the police firingthat occurred on the Pantnagarfarm and because theyfelt that theJanata leadershad not come to see what had happened afterthe killings. AlthoughNaini Tal constituencyis ratheratypical both demo- graphicallyand geographically,several of the patternsrevealed in 21 PacificAffairs thisconstituency are characteristicof broadertrends in the stateas a whole.They includethe historical association between Brahmans and the Congress organization,broad kisan support for the Lok Dal among non-Brahmancastes, and the abilityof Mrs. Gandhi and the Congressto win supportfrom the Scheduled Castes and the poor.

AligarhConstituency

Aligarhconstituency comprises about halfof Aligarhdistrict. Al- though the town of Aligarh is within its boundaries, it is pre- dominantlyrural. It is one of the most interestingconstituencies in northIndia forseveral reasons. First, in thetown of Aligarh is the Ali- garh Muslim University(AMU), whosefounders, students and grad- uates played a criticalrole in the developmentof a modernMuslim politicalidentity in India duringthe nationalistperiod. The AMU continuesto be one of the leading symbolsof contemporaryMuslim political identityin India. The demand fora new Aligarh Muslim UniversityAct, which would ensure the preservationof the pre- dominantlyMuslim composition and Islamiccharacter of the univer- sity,has been a salientpolitical issue formany years. Just before the 1980 elections,the governmentof Charan Singh and Mrs. Gandhi both promisedto meetthis Muslim demand. Second, partlybecause of the presenceof the AMU, the townhas frequentlybeen the scene of Hindu-Muslimconflict and riots,which were recurrent during the two yearsof Janata rule.Third, the townand the surroundingcoun- trysidecontain a large populationof low-casteChamars. Moreover, the regionhas been a leading centerof politicalmobilization of the Chamar caste category-particularlyof theJatavs, an upwardlymo- bile segmentof thislarge caste. Fourth,Aligarh district was selected in the early 1960sby the governmentof India and the Ford Founda- tionfor intensive agricultural development. Along with other districts in the Upper Doab, it has been in the forefrontof the Green Revolu- tion in westernU.P. Fifth,two of the tahsilswhose boundaries lie withinthe constituency-Khairand Iglas-contain verylarge popu- lationsof Jats, the leading cultivatingcaste of westernU.P. and the caste to which Charan Singh belongs. There are also large popu- lationsof Brahmansand Rajputs, as well as a broad arrayof other castesand communities.Finally, the constituency has been an impor- tant centerof the Arya Samaj, a Hindu religiousreform movement that has often allied with the Jan Sangh in support of Hindu causes-such as cow protection-and thatdraws most of its members

22 The 1977 and 1980 ParliamentaryElections in UttarPradesh and supportersfrom the elite castes of Brahmansand Rajputs and fromJats. The diversityof this constituency has been reflectedin its political history.In contrastto Naini Tal, it has not been eithera Congressor a Brahman stronghold.In fact,the Congresshas not won the seat since 1957;and no Brahmanhas everwon on any partyticket. In the 1962 elections,B.P. Maurya,a Jatav,combining the support of Mus- limsand ScheduledCastes, won theseat forthe Republican Party.In 1967, the seat was won by Shiv Kumar Shastri,who was an Inde- pendentcandidate, a Rajput, and a memberof the Arya Samaj. In the 1971elections, running counter to the "Indira wave" of thatyear, Shiv Kumar Shastri was again victorious,this time on the BKD ticket.In 1977,another Rajput and a formerCongress leader in the district,Nawab SinghChauhan, was sweptinto office by the "Janata wave" witha massive70.85 per centof the popular vote.In 1980,the seat was won by anotherRajput-Indra Kumari,the wife of a big ex- zamindar of the Jadon clan-on the Lok Dal ticket.Thus, during thelast threeelections the constituency has been a strongholdof Cha- ran Singh's forcesin one formor another-firstthe BKD, then the BLD, thenthe Lok Dal. The strategyof theCharan Singh forcesin the last threeelections has been to combinethe supportof Jats and Thakurs (Rajputs)3 in the countrysideto formthe core of theirsupport base. They have done so by presumingthe supportof the Jats forany candidate of Charan Singh and by runninga Thakur on the ballot. On the other side, the Congressstrategy has been to relyon Brahmansupport for Mrs. Gandhi, to regain the support of Muslims and Scheduled Castes,and to attemptto splitthe Thakurs, who are dividedinto nu- merousdistinct clans in thisarea, by also runningThakur candidates. With the exceptionof the 1977 "Janata wave" elections,when the Congresslost its support base among Muslimsand ScheduledCastes, the strategiesof bothsides have largelyworked. The outcomehas de- pended upon the strengthof the respectivecoalitions, the degree to whicha candidateof eitherside can musteradditional support, and the extentto whichthe electorateis dividedby the appeals of strong third-partyor independentcandidates. Lookingfirst at the Congressstrategy in the 1980 electionsin Ali- garh,the Congresscandidate, Ghanshyam Singh, was a Thakur of the Chauhan clan, who had run forthe Congressin 1977 but had

1 The termsRajput and Thakur are interchangeablein U.P., Thakur being the titleused forpersons of Rajput caste,but also a substitutegeneric term for the caste category.

23 PacificAffairs polled only 22.72 per cent of the vote. In 1980, his vote share in- creasedby 10 per cent to 32.51. His major supportin both elections presumablycame fromsome Thakursand mostof the Brahmans.It is notablethat in neitherelection was therean independentBrahman candidatewho drewa largeshare of the vote.The additional 10 per centvote share in 1980probably came principallyfrom Muslims and ScheduledCastes. The Muslims were reportedlysomewhat divided in 1980. They were overwhelminglyopposed to Janata because of the recurrent Hindu-Muslimriots in Aligarhduring the two yearsof Janata rule and because of the identificationof theJanata partywith the former Jan Sangh in the mindsof many Muslims.Some Muslimsturned to the Lok Dal because of the supportgiven the latterby the two com- munistparties-the CPI and CPM-which have some Muslim fol- lowersin thisarea. Otherspreferred the Lok Dal to the Congressbe- cause theycontinued to resentthe police attackson Muslims which had occurredduring the Emergencyin the neighboringdistricts of Meerut and Muzaffarnagar.The Lok Dal opposition to the Jan Sangh and theattempt by Charan Singhto promulgatean ordinance establishinga new AligarhMuslim UniversityAct on the eve of the election' also may have inclinedsome anti-CongressMuslims to sup- portthe Lok Dal. At the same time,many Muslimsin Aligarhtown supportedthe Congressin thiselection because of the backinggiven to Mrs. Gandhi'sCongress by the Shahi Imam of Delhi's Jama Mas- jid, because of the returnto the Congressof H.N. Bahuguna (who is consideredto be an outspokendefender of the Muslim community), and because the Congressmanifesto clearly promisedthe mainte- nance of theminority status of AMU. As forthe ScheduledCastes, it was expectedthat most would simplyreturn to the Congressfold. The presumptionbefore the electionamong academic observers in Aligarhwas thatthe Congress would have supportfrom most Mus- lims,most Scheduled Castes, nearlyall Brahmans,and many Tha- kurs-altogethera strongercoalition than that of the Lok Dal. The factthat the coalition ultimately did notproduce sufficient votes for a Congressvictory was largelybecause theThakur, Jat, Ahir, and other backwardclass votes in the countrysidewent predominantlyto the Lok Dal, manyMuslims also votedfor the Lok Dal, and manySched- uled Caste people voted forJanata. The Lok Dal strategywas simplerthan that of the Congress.It

14 The ordinancedid not,however, receive Presidential assent. 24 The 1977 and 1980 ParliamentaryElections in UttarPradesh was decided, fortwo strategicallysound reasons,not to place a Jat candidateon theparty ticket. On theone hand,few people otherthan Jatswould vote forsuch a candidate;on the other,virtually all Jats would vote forany candidate of Charan Singh. The strategy,there- fore,was to put forwarda Thakurof a clan differentfrom that of the Congresscandidate, with the hope of winningbroad supportin the countrysideamong the two leadingproprietary castes of Thakurs and Jats and fromtheir allies and clients.The Lok Dal strategyproved the moresuccessful: its candidate won the seat with37.81 per cent of the vote,5 percentagepoints ahead of the Congresscandidate. BaghpatConstituency Baghpatconstituency in Meerutdistrict lies in theheart of the Jat countryof westernU.P. and in the forefrontof the Green Revolution in thispart of the state. It is also thehome ground of Chaudhuri Cha- ran Singh.Jats are the leading proprietarycaste herecomprising be- tween15 and 19 per centof thepopulation, according to the old gaz- etteer of the district. Other important proprietary castes are Brahmans,Rajputs, Tyagis, and Gujars. Chamars compriseapproxi- mately 12 per cent and Muslims between 15 to 23 per cent of the populationin differentparts of the constituency."5 In districtCongress politics, and oftenin electionsin thisdistrict, Jatsand Tyagishave been allied witheach otherin competitionwith Brahmans and the mostlyurban tradingcastes of Banias. In the countryside,there are also traditionalrivalries between Jats and Raj- puts and betweenJats and Gujars. The electoralhistory of the con- stituencyhas partlyreflected these alliance patternsand rivalries. Congresswon the firstthree elections in the progenitorsof this constituency-MeerutSouth in 1952 and Sardhana in 1957 and 1962-with urban Brahman and Bania candidates. However, by 1962 Chaudhuri Charan Singh had become disaffectedfrom the dominantgroup in Congressstate politics, and his own dominancein the districtalso was being challenged. Moreover,discontent was growingin the countrysideamong the kisansover the failureof the stategovernment to provideadequate supportfor agricultural devel- opmentand forfood-grain and sugar-caneprices. Consequently, the Congresscandidate's vote share in 1962 dropped to 32.86 per cent, and a ruralJat candidaterunning as an independent-allegedlywith

15H.R. Nevill,ed., Meerut:A Gazetteer,Being Volume IV ofthe District Gazetteers of the United Provincesof Agra and Oudh(Allahabad: Superintendent,Government Press, United Provinces, 1904),pp. 187-316. 25 PacificAffairs the surreptitiousbacking of Charan Singh-polled 27.22 per cent of the vote.In 1967,in a replayof the contestbetween the Congressur- ban Brahmancandidate and theJat independent,the latterwon the seat with an absolute majorityof 50.15 per cent. In the next three elections,the patternof rivalrychanged completely to an entirelyru- ral competitionbetween Ram Chandra Vikal, a man from the middleproprietary caste of Gujars,who ran on the Congressticket in all threeelections, and Jat candidatesrunning on the BKD/BLD/ Lok Dal tickets,respectively. In 1971,Ram Chandra Vikal defeated the BKD Jat candidatewho had won the seat in 1967.Then in 1977 and 1980,Charan Singh himselfcontested and defeatedVikal with the largestmajorities ever achieved in the constituency-63.47per centin 1977and 64.45 per centin 1980.In a sense,therefore, the elec- toralhistory of thisconstituency capsulizes several important aspects of the broader post-Independencehistory of state politics:namely, the transitionfrom urban to ruralleadership, the risein importance of the middle agriculturalcastes, and the competitionbetween par- ties forsupport among thesemiddle castes. Interviewsand newspaper reportsfrom Baghpat constituency forecastthe followingtrends during the 1980 election.16Most Brah- mans had voted forRam Chandra Vikal in 1977 and would do so again in 1980. Brahman respondentsexpressed a favorableview of the Emergencyand theirfaith in Mrs. Gandhi. It was expectedthat the middle landed castes, such as Ahirs (Yadavs) and, of course, Jats-but not Gujars-would vote forCharan Singh. Some Sched- uled Casteswho wereinterviewed-Julahas and Chamars-expressed a preferencefor Mrs. Gandhi and hostilitytowards Charan Singh. It was assumed,once theJanata candidatehad withdrawn,that nearly all the ScheduledCaste votes,except those controlled by Jats,would be cast forthe Congress.Muslim voteswere expectedto be divided, withmost going to the Congresscandidate. In general,the electoralcontest in Baghpat seemed to turn pri- marilyaround threefactors: divisions among the leading proprietary castes;conflict between the ScheduledCastes and the leadingpropri- etarycastes, especially the Jats; and the issuesof the cane-priceand the availabilityof needed inputssuch as diesel,electricity, and water, which were of concernmainly to the landed castes in the constitu- ency.These threefactors are representativeof broadertrends in elec- toraland politicalconflict in the northIndian countrysidein recent

16 See, especially,Statesman (New Delhi), December 28, 1979; NorthernIndia Patrika, Decem- ber 28, 1979; and OverseasHindustan Times, January 17, 1980.

26 The 1977 and 1980 ParliamentaryElections in UttarPradesh years.It should also be noted that theseissues and conflictsintersect with each otherin such a way as to preventa polarizationof class conflictin the countryside,to perpetuatethe complexintercaste al- liance patternsthat have alwayscharacterized north Indian political behavior,and to maintainthe economicand politicaldominance of the landed castes,which is not affectedby electoraldivisions among themeven thoughthe Scheduled Castes and the Muslimsmay hold the balance of popular votesneeded forvictory in an electoralcon- test. Conflictamong the leading proprietarycastes was evident in Baghpat particularlyin the underlyingopposition between Jats and Brahmans.Sure of the Brahmanvote, however, the Congressdid not feelit necessaryto run a Brahmancandidate in thisconstituency. In any event,such a decisionwould have been counterproductivefor the Congressin Baghpat,where the Brahmansare a less powerfulrural social class thanthey are in someother districts of U.P. and wherethe supportof some elementsfrom the middleproprietary castes is essen- tial forvictory. Consequently, the Congressran Ram Chandra Vikal again in the hope of dividingthe middle castes. The Janata party pursueda similarstrategy at firstin also puttingforth a Gujar can- didate. The withdrawalof theJanata candidate duringthe election campaign seemed to add considerablestrength to the Congress(I) candidate by preventingthe division of Gujar votes and also of ScheduledCaste votes,some of whichmight have gone to Janata be- cause of the identificationof that partywith the name of Jagjivan Ram. The Congresscandidate, therefore,seemed to have a strong base withthe solid supportof Brahmans,Gujars, Scheduled Castes, and Muslims. Againstthis formidablecoalition, the Lok Dal reliedprincipally upon the supportof theJats and the middleproprietary castes allied withthem, especially the Ahirs (Yadavs) and also, accordingto one newspaperreport, the Thakurs (Rajputs) and otherlanded commu- nities."The practicalissues used to unitethese castes were the sugar- cane priceand theavailability of agriculturalinputs and otherscarce commodities.The high cane-pricebenefited Brahmans and the Gu- jars as much as otherlanded castes in Baghpat,but thesetwo com- munitieswere dissuaded fromallying with other proprietary groups on economicgrounds by the appeal to caste loyalty,which inclined themtowards the Congress.

17NorthernIndia Patrika, December 28, 1979. 27 PacificAffairs The ScheduledCastes shouldbe regardedmore as pawns than as a powerfulforce in theirown right.All the landed castesshare an in- terestin thwartingthe economicand political demands of the low castes,but the votesof the latterare numerousand can be decisivein many constituenciesin U.P. In Baghpat, it has been reportedoften that the landed castes-especially Jats and Gujars-do not permit the low castesto vote as theywish, but cast theirvotes for them. Al- thoughboth Jats and Gujars have been accused of this kind of in- timidation,the Congresshas had considerablesuccess in identifying theJats and Charan Singhas the leadingexploiters of the Scheduled Castes. Givena freechoice, therefore, it is likelythat most Scheduled Castes in Baghpat would have cast theirvotes forthe Congress(I) candidate.In Jat-dominatedareas, however, many Scheduled Castes may have voted forCharan Singh or had theirvotes cast forthem. Regardlessof how Scheduled Castes voted,economic and political powerin Baghpat,as elsewherein the northIndian countryside,re- mains in the hands of the dominantlanded castes,who may contest againsteach otherin the politicalarena, but who neverthelessshare commoneconomic interests. GondaConstituency Gonda constituencyin Gonda districtlies in the formerprovince of Oudh, in a regionwhere politics and rural economyhave tradi- tionallybeen dominatedby great landlords-known here as taluk- dars-and theirdescendants. In the area of Gonda constituencyin particular,the mostpowerful talukdari family has been that of Man- kapur,a Rajput estatewhich once collectedthe revenue from 149 vil- lages. Althoughthe heir to the Mankapur estate,Anand Singh, no longeris entitledto collectthe land revenues,his father'sand his own good managementand politicalskills have made it possiblefor him to maintaina verystrong economic position and to extendhis politi- cal influencebeyond the area thatwas formerlycontrolled directly by the estate.The passage of land-ceilingslegislation and numerousat- temptsby the stategovernment to confiscatethe lands underthe di- rectcontrol of the Mankapur estate have resultedin muchof the land formerlyheld by Anand and his fatherbeing lost. However,Anand Singh stillcontrols some land'8 and also owns several agro-business

18 Some yearsago, Anand Singh transferred500 acres of land to an educationaltrust. The stategovernment challenged the transferin the courtson the groundsthat it was a bogus trans- action,alleging that Anand stillcontrols the land indirectly.Anand is fightingthe case in court, insistingthat the transferwas a genuineone.

28 The 1977 and 1980 ParliamentaryElections in UttarPradesh enterprises.Moreover, the palace of the Mankapur estate has been maintained-as well as a flockof managersand personalretainers, who become campaignmanagers for Anand Singh or the candidate of his choiceat electiontime. Most importantfor the politicalsuccess of Mankapurhas been theperpetuation and extensionof its influence in the countrysideby continuingthe princelytraditions of dispute settlement,relief of thejust grievancesof the people who seekhelp in dealingwith the local administration,and distributionof patronage, especiallyin hard times. Mankapur is not the only politicallyactive estate in Gonda dis- trictor in the constituency.The heirsto the Gonda estate-the sec- ond largestformer talukdari estate in Gonda district-also have been politicallyactive in thisconstituency, formerly in alliance withMan- kapur but latelyin opposition.Moreover, Anand Singh's uncle has controlledextensive landholdings and has participatedin districtpol- itics,sometimes in oppositionto Anand and his father. The constituencyis virtually100 per centrural, containing only a fewsmall townswithin its boundaries.Its largestcastes and commu- nities,in rankorder, are Brahmans,Muslims, Koiris, Kurmis, Ahirs, Kahars, Chamars,and Rajputs,of whomthe mostinfluential are the Brahmans and Rajputs followedby the middle-peasantcastes of Ahirs,Kurmis, and Koiris.However, in all electionscaste has been a less importantfactor in thisconstituency than the resourcesbrought to bear by the heirsto the greatestates and theirpolitical rivals. Most of the electoralhistory of the Gonda constituencyhas cen- teredaround the effortsof the Mankapurscions to demonstratetheir politicalpower in orderto retaintheir economic resources. In fact,in all electionsexcept that of 1952,whether or not Anand Singh himself has contestedthe parliamentaryseat and irrespectiveof the political partieswhose candidates have been put forthin theconstituency, the contesthas been betweenthe Mankapur forcesand theirrivals. Sev- eral of thoseelections also have been "prestige"contests, bitterly con- tested,and involvingcandidates of considerableimportance in state or nationalpolitics. Anand Singh himselfhas contestedthe electionthree times: suc- cessfullyin 1971 on the Congress(0) ticketagainst his uncle on the Congress(I) ticket;unsuccessfully in 1977 on the Congress(I) ticket; and successfullyin 1980,again on the Congress(I) ticket.Running againstthe Raja ofGonda on theLok Dal ticket,Satya Deo Singhon theJanata ticket,and sevenother minor candidates, Anand won the last electionwith a huge marginand an absolutemajority (50.95 per 29 PacificAffairs cent) of the votes.The resulthinged partly on the relativeresources and popularityof the contendingscions of the twocompeting houses of Mankapur and Gonda, and partlyon caste coalitionssimilar to those in otherconstituencies in the state. In both respects,Anand Singh had the advantage. Despite the fact that Gonda constituencyhad been gerryman- deredin such a way as to dividethe central core ofAnand Singh'sin- fluence,the Mankapur forceshad built a broaderbase of supportin the area over the past elections.Moreover, the caste combinations also favoredMankapur in this election.The Lok Dal strategyin Gonda constituencywas to win the supportof Brahmans,the largest caste in the constituency,on the strengthof the factthat theircan- didate, the Raja of Gonda, was a Brahman;and to combine Brah- man supportwith that of backwardcastes, particularly Ahirs, who are also numerousin the area. In fact,however, interviews and per- sonal observationsin the constituencysuggest that the Mankapur forceshad broader influencethan the forcesof the Raja of Gonda and that the Lok Dal's strategyof combiningBrahman and back- ward-castevotes did not work effectively.Insofar as the backward castesare concerned,it was generallyconceded that the Ahirswould votefor the Lok Dal candidate.However, the other great middle cul- tivatingcaste of Kurmis,who are also numerousin Gonda constitu- ency,were reportedlydivided, because theirmost prominent leader, Jai Ram Verma fromthe adjacent districtof Faizabad and pre- viouslya supporterof Charan Singh, had joined the Congress(I). Brahmans,too, were divided; but it is probablethat Anand Singhre- ceived mostof his supportfrom Brahmans and thathe securedmore Brahmansupport than the Raja of Gonda. It was also reportedthat therewere some divisionsamong Muslimsand thatthe tradingcom- munityand some otherurban groupswould vote forthe Janata can- didate, who was a memberof the RSS. It was expected that the majorityof theThakurs, the ScheduledCastes, the middleand back- ward castes other than the Ahirs and Kurmis, and the Muslims would vote forthe Congress. The followingimpressions emerge most strongly from interviews and observationsin Gonda constituency.The firstis thatthe election contestin 1980, as in most previouselections, centered around the conflictsbetween the descendantsof the great landed estatesin the district.Second, in thiscontest not only the lower-castegroups, but also the middlecastes, were essentially pawns in an arena dominated by menwith great resources. Third, the Congress in thisconstituency, 30 The 1977 and 1980 ParliamentaryElections in UttarPradesh despiteits desire to projectan image as theparty for the poor and the landless,was dependentupon the Mankapur forcesfor its victory.Fi- nally,the coalitionput togetherby the Congressstrikingly illustrates a generalaspect of Congresssupport in U.P.-namely, that it is a partyof extremes,combining the upperand lowerlevels of the social and economicorder in itscoalition, but lackingmuch support among the middleproprietary castes.

DeoriaConstituency

Deoria constituencyis one of threeparliamentary constituencies in the verydensely populated districtof Deoria, which has always been-and remainstoday-virtually entirely rural. The leading pro- prietarycastes in Deoria traditionallyhave been the Brahmans,Raj- puts,and Bhumihars,but the mostnumerous castes in thedistrict are the middlecultivating castes of Ahirs(Yadavs), followedby the low- casteChamars and Brahmans.The densityof populationand the ab- senceof other work opportunities except through emigration have led to greatpressure on the available land. The ratioof agriculturalla- borersto the total workingpopulation in the districthas increased duringrecent decades. According to the 1972census, nearly one-third (32 per cent) of the workingpopulation were agriculturallaborers.'9 Alongwith most of theother Eastern Districts of U.P., Deoria has been a leadingcenter of Socialistpolitics. Here as elsewhere,however, the Socialist movementwas set back and ultimatelyeliminated as a seriouspolitical force by persistentfactionalism and defectionsto the Congress.Nevertheless, the politicsof the districtand of Deoria constituencyhave come to be dominatedby personswho were once prominentSocialist leaders,though they are now membersof the Congress,Lok Dal, or Janata. Because of its Socialistpolitical history and because of the large concentrationof the middle cultivating castes of Ahirs,Deoria constituencyis an ideal site to considerthe electoralsignificance of the alliance forgedin 1974 betweenCharan Singhand the BKD, on the one hand, and Raj Narain and the SSP, on the otherhand-and the impactof that alliance on the 1977 and 1980 elections. In thepre-Independence years, the CongressSocialist Party was a major forcein the Congressorganization in Deoria district.In 1952,

19 Censusof India, 1971, Series 21: UttarPradesh, Pt. II-A: GeneralPopulation Tables, by D.M. Sinha (Delhi: Controllerof publications,1975), pp. 318-19. 31 PacificAffairs the Socialistparty won one-thirdof the fifteenLegislative Assembly seats in the districtand one of the threeparliamentary seats. Ramji Verma won the Deoria District(East) parliamentaryconstituency in the 1952 electionson the Socialistticket; in 1957,as a memberof the Praja SocialistParty (PSP), he won the Deoria seat with52 per cent of the votes. In 1962, Deoria constituencywas considered"safe" enough to give to AshokMehta, the mostprominent national leader of the PSP in the country.This time,however, Vishwanath Pande won the seat forthe Congressin a four-waycontest with a pluralityof 40 per cent of the votes.The seat remainedin the hands of the Congressduring the next two elections.The Socialist parties,which were previously major forcesin the constituency,were severelyweakened by the de- fectionof the followersof AshokMehta to the Congressin 1964 and by a successionof splitsand mergersin the remnantsof the Socialist partiesthereafter. In the electionof 1967, the candidate of the Sa- myuktaSocialist Party (SSP) polled only 27 per cent of the vote in Deoria constituency,and in 1971 anotherSSP candidate polled only 13 per cent of the vote. Between 1952 and 1971, the characterof the electoral contest changedin two importantrespects. The firstchange, already noted, was the decline in the strengthof the oppositionto the Congress- particularlythe Socialistopposition. Another more dramatic change was taking place in successiveelections, however, that was more threateningto the Congress:the entryof backward-classcandidates into this constituencyand the adjacent constituencyof Salempur, and theattempts by SSP candidatesin particularto mobilizesupport fromthe backwardcastes. In Deoria constituency,the successfulPSP candidatein 1957 was a Kayastha,while the second-placecandidate was a Brahman who wenton to win theseat in 1962.In 1967and 1971,the successful Con- gresscandidate was a Bhumihar.The second-placecandidate in 1971 was a Yadav, who polled only 13 per cent of the vote on the SSP ticket.However, in 1977 and 1980 thebackward castes made theirin- fluencefelt in a verydramatic way. In the 1977 election,the BLD- formedfrom the merger of theBKD and the SSP-put forwardUgra Sen, a long-timeradical Socialistleader of the district.Though him- selfa Thakur,he had built up strongsupport among the backward castes,especially the Ahirs.In that "Janata-wave"election, he de- featedVishwanath Pande by a massivemajority, with 77 per cent of thevote. In 1980,after the split in theJanata, Ugra Sen ran again on 32 The 1977 and 1980 ParliamentaryElections in UttarPradesh the rumpJanata partyticket. However, the main contestwas be- tweenthe Congress (I) candidate,Ramayan Rai, a formerPSP leader of the districtand a Bhumihar,and the Lok Dal candidate, Ram Dhari Shastri,a memberof theSainthwar caste.20 In a veryclose con- test,the Congress (I) candidatewon witha bare pluralityof 77 votes. Amongknowledgeable observers, party election agents, and vot- ers interviewedin this constituencyin December 1979, therewas a highdegree of consensus that most of the backward-caste votes would go to the Lok Dal candidate.Even theJanata partyworkers, though hopingto getsome votesfrom the backwardcastes, completely wrote offthe votesof the Yadavs and the Sainthwars,"all" of whom,they said,would votefor Ram Dhari Shastri.Interviews with Yadav kisans in the constituencyyielded the same impressionand also made it clear that,by votingfor the Lok Dal candidate,they were voting for Charan Singh. It is importantto note, however,that this kind of middle-castepro-Lok Dal sentimentbecame attenuatedor non-exis- tent in this constituency-aselsewhere-among non-Yadav middle castesand the poorermiddle castes. Insofaras the upper-castevotes were concerned,there was a fair degreeof consensus among those persons interviewed about the likely votingbehavior of these groups.It was presumedthat most Brah- mans would vote forthe Congress(I). There was agreementbetween the electionagents of both the Janata partyand the Lok Dal that mostThakurs would vote forthe Janata candidate,Ugra Sen, him- self a Thakur. It was also generallyassumed that the Bhumihars would vote forthe BhumiharCongress (I) candidate. As forthe Muslims and theScheduled Castes, Janata and the Lok Dal hoped to draw votesfrom these traditional Congress (I) support groups,but theirhopes seemedexaggerated. The aggregateelection returnssuggest that the Scheduled Caste votes were split between Congressand Janata, with Congressreceiving the largershare. It is not clear to what extentor how the Muslim voteswere divided. What does emergemost clearly from the interviews in thisconstit- uencyand fromthe electionreturns is that,even thoughthe Lok Dal candidate lost,the middle castes,especially the Yadavs, made their influencefelt very decisively. There was simplyno doubt in theminds ofany ofthe persons interviewed, including the opponents of the Lok Dal, that the Yadavs would vote solidlyfor Ram Dhari Shastri.A

20 The Sainthwarsare mostlycultivators who, thoughthey claim Thakurstatus, are gener- ally classifiedas a backwardcaste. 33 PacificAffairs secondprominent feature of the electoralcontest in Deoria constitu- encywas thatthe Congress here, as elsewherein U.P., won theseat by combiningopposite ends of the social order-Brahmans and Bhumi- hars,on theone hand, and ScheduledCastes and the Muslim minor- ity,on the other.Finally, this was one constituencywhere the Lok Dal nearlywon because the middle castes, primarilythe Yadavs, were united while the upper castes,the Scheduled Castes, and the Muslimswere, in varyingdegrees, divided between Congress (I) and Janata.

CONCLUSIONS The followingconclusions emerge from this surveyof fivecon- stituenciesin U.P.2" First,to the extentthat theseconstituencies are representativeof the state as a whole,the landed castescontinue to be politicallydominant in the U.P. countrysideand constitutethe cen- tralcore of supportfor all leadingpolitical parties. The predominant landed castes in U.P. include the elite proprietarycastes of Brah- mans,Rajputs and Bhumihars,and the middlecastes of Jats, Ahirs, Gujars,Kurmis and othersmall groupsof local importance.The elite castes, especially, have retained their traditional importance as candidatesand sourcesof supportin theseconstituencies. Ten of the fifteencandidates of the three leading parties in these five con- stituenciesin 1980 were Brahmans,Rajputs or Bhumihars;one was an Agarwal;and fourwere middle castes.Four of the fivesuccessful candidates were of Brahman, Rajput or Bhumiharcaste and the fifth-CharanSingh-was a Jat. It also shouldbe notedthat in three of the fiveconstituencies the main contestswere among personsof elite-castestatus-a Brahman,a Rajput and an Agarwal in Naini Tal; a Rajput againsta Rajput in Aligarh;and two Rajputs and a Brahman in Gonda. In only one constituency-Baghpat-was the main contestbetween two personsof middle-castestatus. Here, how- ever,the Jats have forlong been thedominant caste both in numbers and in controlover the land; theyhave traditionallyoccupied the ec- onomicposition-if not thestatus-that Brahmansand Rajputs have held elsewherein U.P. Only in Deoria, in otherwords, was therea di- rect contestbetween a representativeof an elite caste (Bhumihars) and a backwardcaste (Sainthwars).Even here it is somewhatmis- 21 Althoughthe constituenciesare not representativein a demographicsense, I believethat, takentogether, they illustrate the major trendsand tendenciesin contemporaryU.P. electoral politics.Also, my interviewsin 1979 were not confinedto the fiveconstituencies discussed above. The conclusions,therefore, are based also on mybroader impressions derived from inter- viewsin severalother constituencies. 34 The 1977 and 1980 ParliamentaryElections in UttarPradesh leadingto describethe contestin thisway, since the Bhumiharshave traditionallyoccupied an ambiguous position in the Hindu hier- archy,and the Sainthwars,though of Kurmi origin,claim Rajput status.All in all, therefore,what is most strikingabout the leading candidates is that they came overwhelminglyfrom the chief land- owningcastes in each constituency,whether of high- or middle-status ranking. Especiallyworthy of noticeis the prominenceof personsof Raj- put caste in theseelectoral contests. The Rajput and Thakur castes have traditionallybeen the leading proprietarycastes in U.P. Before zamindariabolition, they were the principallandowners in mostdis- trictsin theprovince-particularly in Oudh, wherethey usually held more than half of the land. In the fiveconstituencies discussed in thisarticle, Rajputs contributedthe largestnumber of leading can- didates-seven out of fifteen-andwon two of the fiveseats. They were leading contestantsin all constituenciesexcept Baghpat, and theyran as nomineesof one or more of the threemajor partiesin theseconstituencies-on the Lok Dal ticketin Naini Tal; on the Lok Dal, Congress,and Janata ticketsin Aligarh;on both the Congress and Janata ticketsin Gonda; and on theJanata ticketin Deoria. In the aggregatein thesefive constituencies, Rajput candidatesreceived 51.32 per cent of the votespolled by the leading candidates. The second importantconclusion from this surveyof five con- stituenciesis a corollaryof the first,namely, that despite the electoral significanceof the ScheduledCastes and the Muslimsas swingforces, particularlyfor the Congress,they are reallypawns in electoralcon- testsin whichthe resourcesand the predominantlocal supportbases are controlledby the landed castes,whose economic interests are sim- ilar whateverthe statusdifferences among them.Two otherfeatures of the political positionand behaviorof the Scheduled Castes and Muslimsshould be notedhere. One is thattheir support for the Con- gress,which has been the preferredparty for both of thesecategories of voters,has fluctuatedacross time and space. In Aligarh,for ex- ample,the Congress lost the support of both groups in the 1962, 1967, and 1977 elections.More generally,the success of the Congressin particularconstituencies and in the state as a whole has depended upon itsability to retainthe support of most voters from among these two categories.The Congresscannot affordto base its appeal exclu- sivelyon the supportof thesetwo categories,however, for power in the countrysideremains with the landed castes. The last point to be noted about the low-castevoters and the 35 PacificAffairs poorervoters of middle-castestatus is that,in manycases, their votes can be purchased.The impressionis unavoidable that many,if not most,poor personsof low-castestatus are lookingfor protection and patronagefrom anybody who comes by. Most low-castepersons see the Congressas theirprotector and vote accordingly,unless they are frightenedby such issues as sterilizationor are intimidatedby the landed castes.They are not,however (as theyare in Kerala), an inde- pendentpolitical force and do not show signsof becomingone. A thirdset of conclusionsconcerns the supportstructure and re- sourcebase of the Congress.The centralcore of supportfor the Con- gressin U.P. comes fromthe Brahman proprietarycastes and, al- thoughthere are some exceptions,the Brahmanshave tendednot to divide theirvotes. Save forunusual circumstances,Brahman votes will go to the candidateof the Congress,whatever his caste,and will not be divertedto non-Congresscandidates of Brahman caste. It is myimpression from these constituencies and frominterviews in many otherconstituencies throughout north India that the generalization applies to small Brahman kisansas well as to the more prosperous ones. There is a broadergeneralization to be made concerningthe Con- gresssupport base, however,which is thatthe Congress depends upon elite castes and the wealthyfor victory in its electioncontests while projectingan imageof itself as theparty of the poor and landless.It is also truethat the Congressin 1980 and at othertimes has won broad supportfrom among the low-and middle-castelaborers and poor ki- sans.The Congress,because of the manyameliorative measures it has taken on behalf of these groupswhen in power and because of its carefullycultivated image as the protectorof the poor and the land- less,has become the preferredparty for most of thesepeople in U.P. It is because of its mutual dependenceon elementsfrom both upper and lowerlevels of the economicand statushierarchies in U.P. that I have describedthe Congressas a "coalitionof extremes."Moreover, it shouldbe obviousthat, whatever its rhetoric, a partyso constructed cannot be an instrumentfor revolutionary change in the northIn- dian countryside-orprobably at all in India. The Lok Dal emergedfrom the 1980 parliamentaryelections as thesecond largest party in U.P., lessthan seven percentage points be- hindthe Congress in thestate's total vote share. It was also the princi- pal oppositionto the Congressin mostconstituencies in the state-as it was in fourof the fiveconstituencies analyzed in thisarticle. Inter- viewsin all fiveconstituencies left little doubt that therewas a solid 36 The 1977 and 1980 ParliamentaryElections in UttarPradesh core of kisansupport for the Lok Dal throughoutthe state.The soli- darity of the Yadav kisanswas especiallyevident, particularly in Gonda and Deoria constituencies.The supportof the Jats in west- ernU.P. was, ofcourse, taken for granted; but thatof theYadavs was more important,for the lattercaste is much largerand more wide- spreadin U.P. Thus, in the 1980 elections,the Lok Dal withits core of supportamong the Jat and Yadav castes was pittedagainst the Congresswith its core of supportamong the Brahmans.The support of all otherelite- and middle-castegroups was divided among the threeleading parties in differentconstituencies, and sometimesin the same constituency. The riseof the Lok Dal and its predecessors-theJanata and the BKD-has been associatedwith the increasingprominence of middle castesboth as votersand as candidates.The latterphenomenon was especiallyevident in Baghpat,where, although they were divided, the leading candidates were all of intermediate-castestatus, and in Deoria, wherethe middle castes were united behind the Lok Dal can- didate. The growingpolitical importanceof the middle castes in thesetwo constituenciesover the last seven electionshas involveda movementaway fromboth urban candidates-as in Baghpat-and candidates of elite-castestatus. However, the significanceof these trendsshould not be exaggerated,for as has been noted above, the elite castes remain politicallymore importantthan the backward castes in the northIndian countryside.Economically, the most im- portantfact is the universaldominance of the landed castes in the U.P. countryside.The political divisionsamong these castes reflect not onlythe persistingimportance of statusdifferences among them; theyalso reflecta feelingamong the leadersof the middlecastes that the elitecastes, who have dominatedmost local institutions,are bar- ringtheir advancement, which in effectmeans barringtheir access to controlover these institutions-thebanks, the cooperativesocieties, and the schools-and the resourceschannelled through them. If theseinterpretations of thepolitical divisions among the landed castesare correct,then the political process of the northIndian coun- trysideis likelyto continueto be characterizedmore by tacticsof coa- lition,division, and cooption of particularcaste groups and their leadersthan by class-castepolarization, for it is primarilya conflict overpolitical control of economic resources that is occurring,and not so much a conflictbased on differenteconomic interests. The future politicalstrategy of the Congress,therefore, will probablybe to di- vide the middle castes. As long as political power in the provincial 37 PacificAffairs and nationalcapitals is dependentupon adult suffrageelections, the leading contendersfor power are likelyto continueto attemptto build inter-castecoalitions that cut acrosseither status or economic differences-orboth. Class polarization,therefore, is unlikelyto occur as long as the Indian regimeis based upon open, competitiveelec- tions. This concludinganalysis has focussedso faron questionsof caste and class. It remainsto considerboth the importanceof the issuesin the campaignand the significanceof the 1980 electionfor the future of Indian democracy.On the face of it, the electionseemed to turn overwhelminglyon the issuesof highprices, scarcity of essentialcom- modities,the sugar-caneprice, and the availabilityof agriculturalin- puts (particularlydiesel) needed by thekisans in theprevious growing season. Several of these issues cut evenly across lines of caste and class. Only the verybiggest farmers, with ample financialresources and good politicalconnections, were able to obtain sufficientsupplies of diesel.Yet, the votersclearly did not respondevenly to the issues. Whetheror not a voterblamed governmentfor the economic diffi- cultiesand scarcities-and,if so, whichgovernment-depended more on the caste statusof the respondentsthan on theireconomic posi- tion.The break-upof the Janata coalitionalso made it convenientfor differentcategories of votersto be selectivein allocatingblame. Brah- man votersfavorably disposed to the Congressnaturally blamed the Janata governmentin general,including both the rumpJanata party and the Lok Dal. Yadav kisans,who were inclinedtoward the Lok Dal, however,blamed onlythe Janata governmentand excusedCha- ran Singh,who had been in officeonly a shorttime. Brahman and Yadav kisans who harvested cane in November-Decemberhad equally good reason to thank Charan Singh for the high price of cane, but only the Yadavs articulatedsuch feelings.In otherwords, the issuesin the campaignwere as much excusesfor voting behavior as reasonsfor it. The reallycentral issue in the campaign-in U.P. at least-was whetherthe votersidentified with the middle cultivating castesor withthe Congresscoalition. The strongundercurrent in this campaign,therefore, was the questionof the politicaland economic advancementof the middlecastes. To giveprecedence in thisway to caste overeconomic issues is not to deny the importanceof the latter.People who supportedCharan Singh and the Lok Dal-particularly thosefrom the middlecastes- did so fora combinationof caste and economicreasons. The cultivat- ing kisansof intermediate-castestatus with economically viable hold-

38 The 1977 and 1980 ParliamentaryElections in UttarPradesh ings believed that Charan Singh and the Lok Dal would promote theireconomic interests. Moreover, his hold on thesecastes was solidi- fiedby the beliefamong the middle castes that he was one of them and thathe favoredtheir social advancementas well. A Brahmanor Thakurwith the same programas Charan Singhwould not have had the same appeal to thesegroups. Anotherfeature of the electionand voterreaction to it in U.P., whichwas mostapparent in interviewsacross the state, was thatmore voterswere orientedto the national leadersand to the partylabels than to the local candidates.It would not,however, be correctto in- ferfrom this fact that the traditionalparochial attitudes of the north Indian electoratehad been overcomeand replacedby national per- spectives.Rather, it was more a case of the voters"parochializing" the nationalcandidates and issues.Often, in fact,the votersdid not knowthe names and castesof the parliamentarycandidates, because theparliamentary constituencies are verylarge and diverse.They did know-or thoughtthey knew-the castes of the national candidates fromU.P. and Bihar,with Charan Singh being somewhatof an ex- ceptionin thathe was seen as simplya memberof a backwardcaste. Thus, in a sense,the nationalleaders have becomesymbols of the lo- cal identificationsand aspirationsof the northIndian voters.Con- sequently,Mrs. Gandhi'smanoeuver in 1971of separating the parlia- mentaryfrom the legislativeassembly elections has had the effectof focussingvoter attention on national leadersand parties,but it has not freedthe parliamentary elections from local, parochialissues and concerns.In somerespects, in fact,it has promotedthe spread of such issues.22 A finalobservation concerns the optimisticconclusions that have been drawnfrom the resultsof the 1977 and 1980 electionsabout the commitmentof the Indian public to democracyand competitiveelec- tions.Such conclusionsdo not seem warranted.In fact,voter com- mentsabout theEmergency in theinterviews conducted in December 1979 suggestthat this issue,even more than the issuesof scarcities and high prices,was used to rationalizevoting behavior that had other motivations.Voters who identifiedwith the Congresseither wereprepared to excuse Mrs. Gandhi forher past errorsand accept

22 It would be interestingto know to what extentthe legislativeassembly elections have been influencedby similartendencies, that is, whetherin thesemuch smallerconstituencies- wherethe names,castes and personalitiesof the local candidatesare moreintimately known to the voters-votingcontinues to be influencedmore by local coalitionsand combinationsthan by the broaderidentifications that are now prominentin the parliamentaryelections. 39 PacificAffairs her statementsthat she would not impose an Emergencyagain, or welcomedthe prospectof reimpositionof an Emergency.Brahmans in particular,some Rajputs, and some low-castepersons openly de- clared that theywould like to see the Emergencyback or, at least, thatthings were better during the Emergency-which,for these per- sons,primarily meant law and orderand stable prices.Opponents of the Congress(I) were naturallyalso vociferousin theirattacks on Mrs. Gandhi and the Emergencyregime-which for them meant forcedsterilization, detention, and otherforms of harassment.It did not,however, appear to thiswriter in December 1979 that attitudes foror againstthe Emergencywere a primarymotivating factor in the votingdecisions of mostvoters. The significanceof the 1977 and 1980 electionsin U.P. must be soughtelsewhere-in the persistentunderlying trends and tendencies thatare maskedif undue attentionis paid to the "issues"articulated by candidatesand voters.Two such trendsare particularlyimpor- tant. The firstis the attemptby the political partiesto turn some caste and communalcategories into votingblocs, providinga stable basis of supportacross time and space. Insofaras the attemptis suc- cessful,it involvesfor Hindus the elevation of the caste category to greaterimportance than local jati as a political factor.In the 1980 elections,three caste categoriesin particular seem to have votedas blocs-Brahmans,Jats, and Yadavs. Brahmansvoted over- whelminglyfor Congress (I), and Jatsand Yadavs forthe Lok Dal, of- ten even when presentedwith reasonablygood candidates of their own caste in oppositionto a candidate of theirpreferred party but not of theirown caste. Muslimsand Scheduled Castes,too, have of- ten votedas blocs-usually forthe Congress.Congress probably won mostof the votesof thesetwo groupings,but therewas some loss of Muslimvotes to the Lok Dal and of ScheduledCaste votesto Janata and the Lok Dal. All othercastes were more divided both acrossand withinparticular constituencies. The tendencytoward bloc votingis partlyreinforced and partly counteractedby the dual appeal of the Congressto richand poor. Its practiceof distributingpatronage and protectionto securethe sup- portof therich draws into its fold persons and groupsfrom the upper layersof all castes.Its adoptionof programsand policiesfor the poor has the effectof consolidatingthe supportof some Scheduled Caste groupsand of drawingaway fromthe Lok Dal segmentsof jatis and caste categoriesof intermediatestatus. The secondmajor underlying pattern in thesepast twoelections is 40 The 1977 and 1980 ParliamentaryElections in UttarPradesh reallyan aspect of the first-namelythe politicizationand increased cohesionof themiddle castes of peasants, who formedthe bulk of the supportfor Janata in 1977 and forthe Lok Dal of Charan Singh in 1980. Some of them supportedCongress (I) also, but most did not and were in conflictwith both ends of the spectrumof supportthat formedthe basic Congresscoalition. This has been the principalun- derlyingconflict in northIndia-between the middle peasantryand all othersocial forces.It was not changed by the 1980 elections.In fact,it was sharpened.The major politicalissue in northIndian poli- tics in the immediatefuture is whetherthat conflictwill intensifyor will be deflectedby politicaltactics that divide middle-casteleaders and groups.The futureof the Indian parliamentarysystem depends more on the outcomeof that strugglethan on eitherthe presumed commitmentof the Indian public to democraticvalues or on tran- sient issues such as sterilization,high prices,and shortageof basic commodities.

Universityof Washington, Seattle, U.S.A., December 1980

41