A Case of Rampaging Elephants: The Politics of the Middle Classes in Small-Town Pakistan

Abstract This article draws on a year of ethnographic fieldwork with traders in a Pakistani agricultural commodity market. It analyses their business and wider networking strategies to show how they – as a segment of Pakistan’s middle classes – perceive and interact with the state in the process of accumulation. Ordered by custom, contract, and selective engagement with state functionaries who also engage them selectively, traders’ economic activity is inextricably bound up with political practices that defy democratic principles. Neither a concern with the public good nor programmatic politics is visible in what traders do; the state is viewed as an instrument of accumulation while itself appearing to have no project of its own separate from the local dominant classes; and collective organisation both substitutes for the regulatory state and staves off its attempts at enforcement. These trends further militate against viewing the middle classes as catalysts of democracy and have important implications for development strategies seeking to reform the state.

their numbers and visibility, both of which have increased greatly as a result of policy decisions fuelling private consumption and the rapid and extensive development of the urban real estate sector since the late 1990s (see, for example, Haq 2012;Anwar 2018;Mallick 2018).Secondly, they are culturally distinct from the Western-educated, often socially liberal middle classes that staffed the state in Pakistan's first four decades and worked as salaried professionals in the formal sector.By comparison, the new middle classes bring with them significantly more vernacular, non-Westernised political and cultural visions and practices rooted in Islamic and nationalist idioms; in this they represent important discontinuities with the past and a conscious repositioning within a changing global order.
The political economy literature has sought to understand the impacts of the rise of these classes for the Pakistani state, society, and economy.Has the coterie of military, landed elites, high bureaucrats, and corporate capitalthat scholars have long argued dominate Pakistanadmitted a new member to the ruling coalition?And what does the steady ascent of this social force portend for democracy in a country with a long history of authoritarianism and deep, divisive hierarchies that order social life?Debate continues on the position of the middle classes in the structure of power, but certain of their political impacts are less contested: by and large, Pakistan's middle classes tend not to challenge an order characterised by structural violence, inequality, and particularist ideologies upheld by the politics of patronage.
This article contributes to the burgeoning literature on the relationship between a growing middle-class fraction that straddles rural and urban (rurban) and the Pakistani state.The literature on the politics of the middle classes in developing countries has tended to focus on its more visible, metropolitan segments, eliding an important aspect of internal differentiation within the group.Students of agrarian change and small towns in South Asia know that the rural and rurban middle classes are growing at least as fast as their big city counterparts (see, for example, Harriss-White 2008; Jodhka 2016; Jan 2019).Aslany's (2019) statistical analysis of rural-urban variation in the Indian middle classes corroborates this contention, demonstrating that a considerable segment of the middle classesand in fact the majority of lower middle-class householdsis rural.But even within the metropolitan areas that scholars privilege as field sites, it is the waged workforce that commands most attention.The heterogeneous nature of the middle class is demonstrated by its members belonging to occupational groups ranging from "the upper tiers of the managerial-professional classes to the clerical and lower-level secretarial workers" (Fernandes 2006, 93-94).The propensity to focus on salaried groups may be justified by the fact that salaried employment as a source of household income is heavily over-represented amongst the middle classes (Aslany 2019).However, this excludes a segment with a wide political footprint, one that has interests contradictory to its salaried counterparts.Theorised as the "intermediate classes" by Kalecki (1969), Jha (1980), and Harriss-White (2003) in the Indian literature, this strata comprises the small capitalists found in every city, town, and village, in the bazaars, in wholesale markets, in the transport sector, and of course in agriculture.What sets them apart from the big bourgeoisie is size, structure, and operational domain.They are small, with an organisational structure embedded in local social institutions, often catering to markets that corporate capital does not.What sets them apart from the salariat is their structural location in the economy.As owners of capital, they do not fear inflation, shortages of essential goods are a boon for greater profits, and tax avoidance is endemic, draining the fiscal power of the same state from whom the salaried middle classes demand infrastructure and expect basic services.It follows then that the politics of the intermediate classes demand interrogation.This article examines agricultural traders' business and wider networking strategies and focuses on their articulation with the state as they pursue accumulation through market exchange.The article is structured as follows: it first introduces agricultural traders as a rurban middle-class fraction and sketches them briefly as economic and social actors operating in the informal economy. 1 It then details four key business strategies that bring them into close contact with the state: the cultivation of strategic relationships, capture of local regulatory institutions, involvement with constituency politics, and collective action.
By looking closely at what traders do in the pursuit of power and profit in these four areas, the analysis provides detail of the undemocratic practices flowing through and shaping the political economy of a small town and beyond.Neither a concern with the public good nor programmatic politics is to be found in the practices of this group, the state is viewed as an instrument of accumulation while itself appearing to have no project of its own separate from the local dominant classes, and collective organisation substitutes for the regulatory state and staves off unwanted interventions such as taxation, quality control, and much else.The implications of this for the state's developmental capacity and autonomy are enormous.Development interventions continue to focus on building capacity and making better policy, but what this article describes is a socially regulated universe in which the state is far from ableand most importantly, willingto do what development experts ask of it.

Situating Rurban Traders as a Middle-Class Segment
Fieldwork for this article was conducted largely in 2011-2012 in central Punjab in a town pseudonymously called Punjabkot.More than 130 interviews were conducted in Punjabkot with agricultural traders, farmers, labourers, state officials, politicians, association leaders, and political brokers.In addition, 12 detailed trader case studies were developed. 2The town's economy is heavily dependent on agriculture and, as in most of central Punjab, small-and medium-sized landholdings are the norm in its hinterland. 3 Agricultural marketsor mandis as they are called in Pakistanare critically important sources of credit, information, and marketing opportunities for farmers operating at this scale, and the mandi in Punjabkot is amongst the 10 largest agricultural commodity markets in the province, having benefited from the economic liberalisation programmes of the 1990s.Shop numbers here almost tripled between 1998 and 2006 and approximately 320 traders operate in the market, together turning over approximately US$70 million annually.The range of turnovers is wide.The smallest have annual turnovers of about $40-50,000 a year, the largest exceed $1 million.Not all entrants are successful but the pull of the trade is strong and has attracted many agriculturalists with some capital and a desire for social mobility.This represents substantial social changelandowning agriculturalists have historically been higher in the status hierarchy than traders and non-agriculturalists.The inversion or equalisation of status is by no means complete but it is underway.Not only is trading considered to be more lucrative and reliable as a source of income than farming, it also makes possible a much-coveted base in town and holds out the promise of participation in urban middle class culture.Many traders are men of some standing in their villages, and almost all of them are middle class given the familiar and widely used definitional criteria of income, consumption, and social status. 4  These traders are also middle class in that they are small capitalists who employ family and non-family labour while themselves working.In India, this segment of capitalstraddling capital and labour and located in small townshas been studied variously as: "intermediate" by McCartney (2010) and Harriss-White (2003); "rural-commercial" in Harriss-White (2008); "regional" by Baru (2000); "fraternal" by Chari (2004); "provincial" in Upadahya (1997) and Rutten (1986); and as "bullock" capital by Rudolph and Rudolph (1987).The trajectories of accumulation of all these segments are characterised by similar features including capital's rural roots, the structuring roles of caste and family, interpenetration with the local state, and a tendency to diversification and multiplication rather than intensive growth.
More than two decades ago, Harriss-White (1996) observed that agricultural traders confronted the mass of the Indian peasantry and landless labour as "masters of the countryside."This is an apt observation in present-day Pakistan where agricultural traders called aarthis in Punjabiare of central importance to the system of agricultural production, lending inputs to small (5.0-12.5 acres) and medium-sized (12.5-50.0acres) farmers and either purchasing crops directly or arranging crop sales between farmers and buyers.Their functional necessity to the system, privileged nodal position within it, and the near-complete absence of organised labour and small farmers groups make the trade a profitable one, particularly for large traders. 5However, this economic power does not regularly translate into influence over policymaking, unlike big business (see Kochaneck 1983;Armytage 2019).Instead, they: defend non-competition with the help of the identity-based social networks through which market exchange is organized; they defend their economic interests through corporatist associations and with the help of their social contacts in the state; they manipulate local party politics (many funding all political parties, rather than being identified with any one); they enforce market contracts through authority relations based on caste, ethnicity, religion, gender, and so on, rather than through state sanctions; they consolidate local power by engaging in small acts of philanthropy or the provision of services to people in need of them, in parallel to the transfers of the state; and they also make profit by coercion and more recently the resurgence of primitive business practiceswhat F. Braudel called "elementary fraud" … (Harriss-White, Jan, and Amirali 2019, 209).
While these words describe the accumulation strategies of the Indian intermediate classes, they are also a remarkably accurate depiction of their counterparts in Punjabkot.Similar to the aarthis of Depalpur and Toba Tek Singh studied by Jan (2017) and Lahori bazaar traders studied by Javed (2018), Punjabkot's aarthis accumulate by: cultivating personal links with politicians and bureaucrats; establishing effective control of market regulatory institutions; displaying power and the capacity to do violence; exhibiting moral uprightness and Islamic piety; and engaging in organised collective action vis-a-vis the state, labour, and farmers.Striving to either grow or survive in a competitive market, these strategies are oriented towards increasing returns on investment and recovering and reducing costs.What aarthis do in pursuit of this flows freely and often unselfconsciously between the legal, extra-legal, and illegal, shoring up an undemocratic and unequal order in which the local state appears almost completely indistinguishable from the town's dominant social classes.
Certain political conditions have to be in place for systematic illegality and extra-legality of this sort to be routinised.Key state institutions and personnel must be complicit either through active collusion or wilful ignorance and in the absence of regulation by formal law, alternative governing logics, moral discourses, and authorities are necessary.Also, those whose labour forms the "nutrient base" of the economy" 6labourers and small farmersmust believe that the system works well enough for them given their alternatives.Akhtar (2014Akhtar ( , 2018) ) has argued that the Pakistani intermediate classes have, since the 1970s, come to play a critical role in stabilising the patronage-based political order.Akhtar (2014, 51) writes: "What appears from a macro-institutional standpoint to look like fragmentation and weakening is precisely what I call a hegemonic politics of common sense," one that is held together by the mediation performed by these groups.It is the mobility of the intermediate classes that enables this roleoriginating from the working classes, at times from historically low-status castes and kinship groups, and either rural or closely connected to the rural, their entry into the ranks of the urban, consuming middleclasses performs both a material and an ideological function.Linking dominant and subordinate classes through their inhabitation of the "middle," they are positioned at both ends of the patron-client dyad, building networks upwards while acting as patrons for the poor, connecting them to the political resources they gain access to.But equally importantly, they also act as ideological-cultural brokers "affirm[ing] the possibilities of upward mobility offered by urbanisation and informalisation, as well as the ruthlessness of a system that can just as quickly punish those who do not negotiate state and market in accordance with the rules of the game" (Akhtar 2018, 65).The argument is substantiated by recent field studies in Pakistan by Javed (2018) and Jan (2019: Ch. 7) and echoes some of the most persuasive perspectives from studies of Indian middle classes, succinctly summarised by Deshpande (2003, 139): "[t]he middle class is the class that articulates the hegemony of the ruling bloc; it both (a) expresses this hegemony by translating the relations of domination into the language of legitimisation; and (b) mediates the relationship between classes within the ruling bloc, as well as between this bloc and other classes" (Deshpande 2003, 139).
Class-theoretic analyses of Pakistan's power structure largely went missing after the Marxian contributions of the 1970s.Only recently did they receive a fillip, when a timely academic intervention by Zaidi (2014) outlined a research agenda for understanding power and politics in Pakistan outside elite circles and urged scholars to socialise the study of political economy in contrast to prevailing state-centric approaches.This agenda has been picked up with enthusiasm (see, for example, McCartney and Zaidi 2019).The analysis presented in this article is similarly motivated by the intention to illustrate the grounded functioning of the political order, one thatas this article began by observing is now host to an ambitious social force seeking to ascend the ladder.The remainder of this essay details Punjabkot's aarthis' business strategies that interface directly with public institutions, personnel, and resources and highlights their implications for state functioning and capacity.

The Power of Personal Connections
The nature of an aarthi's trade is such that he must maintain extensive networks and personalised relationships with farmers, other traders, labourers, state officials, transporters, and mills and factories.These networks are heavily structured by the institutions of family, friendship, biraderi (kinship), faction, and Islam, and by the practice of patronage which flows through and binds them all in corporate blocs of uneven and unequal relations of reciprocity.The strategic deployment of customary solidarities and the practice of patronage is as important as economic capital in this business and the largest aarthis extend their reach into big cities and provincial capitals, leveraging their contacts and diversified investments and looking for ways to expand their businesses and "big men" status.
Reciprocal obligations between social unequalsor patronage relationsare ubiquitous in South Asia.Piliavsky (2014, 159) argues that "the donor-servant relation [is] the basic formula through which people exchange things, exercise power and relate socially, and through which their identities effectively take shape."Patronage is a "living moral idiom" she says, one that cannot be reduced to rational decision-making and instrumental exchange.Osella (2014) goes further to note that in contrast to the West, South Asians aspire to "highly socially connected selves" that are imbricated in each other's social lives; patronage and brokerage are therefore corollaries of valuing "non-autonomy."This moral-cultural perspective highlights the many motivations that propel human decisionmaking and in so doing provides an important corrective to the reductionist accounts of mainstream political scientists. 7Yet the analysis of the effects of these relationships on the larger systemwho it works for, what conditions are necessary for it to function thusoften remains absent.In addition, in their desire to rescue situated, historically, and culturally determined social practices from condemnation by a (Eurocentric) universalism, many authors do not engage in debate about the consequences of patronage for egalitarianism.
However one understands it, patronage is the single most pervasive social institution that wraps around and orders relations across social space in Punjab, with a key function of accessing state power for private purposes.From landed elites to landless peasants, big traders to casual labourers, everyone tries to connect and feed at the public trough.Except for elites whose personal linkages reach directly into the state, access is generally mediated through brokers positioned between political representatives and the public (see Akhtar 2018, 33-34;Mohmand 2019, 20-41).And so it is unsurprising that Punjabkot's aarthisparticularly larger onesspend time and money building extensive networks thatin rare casesculminate in the Punjab Chief Minister's or a federal cabinet minister's office but more often connect them to well-placed brokers, local Members of Provincial Assembly (MPAs), Member of National Assembly (MNAs), and district bureaucrats.
Almost everyone encountered in Punjabkot explained that what mattered most in both business and politics were personal relations with powerful men.The case histories of the largest aarthis highlight, above all, the returns to cultivating good personal relations with a wide selection of well-placed persons.Malik sahib, 8 an established aarthi farming 60 acres of his own land in addition to the aarath, lightheartedly explained his business strategy thus: "We like to be self-sufficient: one retired army officer, one lawyer, now all that's left is a policewalla!But he [the lawyer] has good relations with the police and so that work gets done by him."He was referring to his two business partners, friends who had invested in the aarath.His family has been allied with the Pakistan Muslim League (Nawaz) or PML-N since 1985 and Malik sahib proudly declared that "we've been in power ever since we started politics."In 2012, his cousin's son was an MPA and the chairman of the District Market Committee (DMC)the local market regulatory authoritywas a biraderi member with whom a good relationship existed.Altogether, these connections helped Malik sahib in ways that he considered indispensable but was unwilling to disclose.
Lawyers, policemen, army officers, MPAs are all representatives of the state and their power to undercut, manipulate, and twist rules and regulations is incontrovertible. 9How are such connections made?Which solidarities do they mobilise?In Malik sahib's case, family and biraderi provided the scaffolding.This is typical of Punjab where family and kinship form thick threads connecting people up, down, and laterally across social space.Family is the thickest, and strategic marriageseither within or outside the biraderi/kinship group, depending on what the family wants to achieveare key to economic and political success while biraderi connections draw on "fictive kinship," or a form of political solidarity mobilising the idiom of family but rooted primarily in transactional relationships (Gilmartin 2014, 27-28).Patrilineal descent is important to defining biraderi boundaries but is not determining; what matters most is gift exchange and reciprocation (Alavi 1972).Together with Islam, ethnicity, and faction, family and kinship generate solidarities vital to market functioning.As has been extensively argued, far from dissolving customary institutions, capitalism has reworked them to support accumulation (see, for example, Nadvi 1999;Harriss-White 2003;Meagher 2006).Punjabkot's mandi is no exception; it is difficult to overstate the importance of these ties in the trajectories of accumulation of the commercial classes, although there is also a kind of "loosening" taking place in which a few low-status individuals without useful biraderi or family ties have joined the town's commercial elite. 10 An example of this upward mobility is Arshad sahib.He is one of the biggest traders in the mandi having worked his way up from a petty trading kammi (non-agricultural caste) background.In addition to his aarath, he has two poultry sheds, each representing an investment of Rs. 15 million ($150,000).In 2012, when poultry prices were good, he was turning a Rs.1.2 million profit per shed every 40-50 days.His father had been a small, itinerant trader and Arshad sahib got lucky and was employed in a public institution.However, he retired early, and in the two years immediately afterwards, he was "much busy, no work" making contacts in town and country before taking up the trading work his father had done.He rented a storehouse and started stocking small quantities of mustard seed, paddy, cotton, and other non-perishables, bringing them into the mandi when prices were high.Having built up a small amount of capital, Arshad sahib put his new networks to use and started buying at the auction and selling to mills and factories.Margins were higher in the days before the mobile phone, and he was able to secure recurring supply orders.He explained: "My success was that I understood this country early on.I didn't have a lot of money, but in Pakistan social contacts are most important.If you have them, 80% of the work is done." And money has followed from shrewd investments made possible by knowing the right people.In violation of formal procedure, a senior officer in the public electric distribution company put in seven massive electricity pylons to connect Arshad sahib's poultry farm to the main grid.Arshad sahib had previously leveraged his contacts with a senior district administrator to help this officer with a land encroachment dispute.Another big win was when a well-tended friendship with a former federal minister and intense lobbying with high-level officials in Lahore led to the siting of the new mandi in a place of his choosing.The former federal minister intervened in the process at Arshad sahib's request, and as soon as he learnt that it was going ahead as he had hoped, Arshad sahib started buying land along the main road as well as on the outer perimeter adjacent to the future mandi site.The mandi map-maker was bribed to add a gate connecting Arshad sahib's recently acquired agricultural land to the mandi: "The mapmaker said 'there's no gate here in the original design.'I said, don't look at the map man, look at this wad of cash in your hand and just put two lines there as I say!"The value of this land shot up 300% in four years, even before construction had started, and Arshad sahib had many plans for what he would do with his property, foremost amongst them being the construction of storage facilities and the development of a commercial plaza.
Arshad sahib's story illustrates the power of what is vernacularly called a good "approach."The disadvantages of his received caste and class position were overcomeat least economicallyand he was able to undertake illegal activities with impunity because he knew big men who derived their power from association with the state.In his own words, "free spending" (khula kharcha) and close attention to the personality and preferences of those he was cultivating helped him establish and maintain these relationships in the absence of ties of biraderi and family.

Regulatory Capture
Aarthis of Arshad's sahib economic and social stature are often senior members of trade associations, civic groups, and at times the local market regulatory body.Sitting on the lowest rung of the Agriculture Department's institutional hierarchy and receiving no funds from the provincial government for their day-to-day operations are the DMCs, public authorities charged with establishing, maintaining, and regulating mandis.Their primary purpose is to regulate the first point of sale between farmers and traders in ways that protect small farmers and labourers and ensure fair exchange under the mandate provided by the Punjab Agricultural Produce Markets Ordinance 1978.However, in a textbook case of regulatory capture, DMCs champion the interests of traders, constituency politicians, and public officials and the main activities in the DMC building are well outside official remit.Market fees and local taxes pay for food and drink, air conditioning, and even the transport expenses of the incumbent political party's loyalists who regularly converge in the Chairman's office for the panchayats that resolve disputes, and social gettogethers. 11 Control of the DMC is useful both for aarthis as a corporate group and for the wellplaced individuals able to leverage access to social networks that spin around politicians.Despite trade taking place in a licensed and formally regulated marketplace, much of the activity in and around the DMC is informal, meaning that it takes place outside the regulatory framework of the state, and often it is criminal in that it breaks the law. 12This pervasive informality and normalised criminality is readily apparent in the flouting of labour regulations laid out in the Agricultural Markets Act and national labour law; in the violation of stipulations on farmer remuneration; in the contravention of zoning regulations; in widespread and significant tax evasion; and in the brisk trade in fake licences.It is also observed in the practice of not issuing receipts for retail sales of fertiliser, in seed adulteration, in tampering with weights and measures, and in the flagrant violation of health and safety norms, under-invoicing, and price-fixing at auctions.These practices are known or suspected by everyone in the trade but as part of the received landscape they almost always go unchallenged.
How the DMC is staffed reveals a great deal about how the state is perceived, both by those within it and those outside.At the apex sit the Chairman and Vice-Chairman, ostensibly elected by an Executive Council but in practice appointed by the Punjab Chief Minister and his coterie of managers.Under the Punjab Agricultural Produce Markets Ordinance 1978, the Executive Council itself should be appointed by the elected district council or in the absence of thatrather vaguely -"the Government of Punjab."The law requires that it comprise nine farmers, five traders, one mandi labourer, and one person who is neither a farmer nor a trader.At the time of fieldwork, the Executive Council had been put together by the Punjab Chief Minister's son and the main consideration in its formation was accommodating local elite factions loyal to the PML-N.The positions of Chairman and Vice-Chairman were allocated accordingly and the Executive Council was exclusively composed of intermediate capitalists: all "farmers" owned at least 75 acres of irrigated land, there was no labour representative, and traders were large and included the president and vice-president of the Punjabkot produce market associations. 13 This capture of the DMC is not unconnected to the fact that it is a semi-autonomous organisation that receives no funds from the Department of Agriculture and so must generate its own running expenses.By contrast, the District Commissioner's Officethe allimportant state authority in Punjabkot that a local politician referred to as "mother of the district" (ziley ki ma)is a different beast.A central pillar of the provincial bureaucracy, well resourced, and with much wider and greater executive powers, the social forces vying for its control are much larger and more varied.Here too, individual aarthis and aarthis as a corporate bloc find firm footholds and bend state power to their advantage.A widely employed strategywith obvious implications for state autonomyis petitioning political representatives to intervene with the District Commissioner's Office on their behalf, usually to evade the implementation of the law.
The District Commissioner's Office is staffed by civil servants and senior postings are made by the Chief Minister's office in Lahore.As the balance of power between the political classes and the district bureaucracy has shifted in favour of the former, even senior bureaucrats now need to cultivate the good favour of local politicians to obtain and retain preferred postings (Cheema, Khawaja, and Qadir 2005;Mohmand 2008).Emboldened by this reversal in historical roles and propelled by the logic of the electoral system (see below), Punjabkot's political elites incessantly intervene to enable legal but also illegal, fraudulent, and criminal activity by their well-heeled, well-organised, or otherwise important supporters.One example will illustrate: in 2012 the newly arrived Assistant Commissioner of Punjabkot conducted a raid on a seed dealer in the mandi.A farmer who was also a lawyer had filed a complaint that the dealer had cheated him by supplying adulterated seed.Seven other complaints had already been received against this dealer.The Assistant Commissioner raided the store, impounded evidence, and took him into custody on suspicion of fraud.Within two hours, angry members of the seed association and Anjuman-e-Tajiraan (Punjabkot's federation of trade associations) had gathered outside the Assistant Commissioner's office, upon which the lawyer's District Bar Association also showed up and punches flew.It was only when the incumbent MPA with links to both associations arrived that matters were settled, and a deal was brokered in which the dealer was released and the lawyer compensated.The previous seven complainants were not mentioned and received nothing.The dealer's fraudulent activities went unsanctioned except for when the aggrieved party was able to mobilise vocal and visible support, physical violence, and political powerand even then he was not subject to the law which would have seen him fined and imprisoned for a minimum of three months.

The Business of Constituency Politics
The seed dealerand unfortunately for him, in this case his customer toohad a good "approach.""Approach" (in English) is a term used by illiterate labourers and high-level state officials alike in Punjabkot.To say "uss ki approach hey"literally meaning "he has approach"indicates the person has the contacts, stature, or direct personal relationships necessary to petition the powerful.No businessman in Punjabkot is without an "approach" or recourse to someone powerful, with the larger ones cultivating multiple relationships to this end.In the competitive political environment of Punjab's small towns, contributing to competing candidates' campaigns during elections and not openly taking sides in disputes is common practice.In contrast to how rural politics is conducted with overt displays of loyalty on the part of voters and village-level leaders during elections, the big businessmen of Punjabkot are much less inclined to proclaim their exclusive loyalties, even ifor precisely becausethey often have high stakes in the outcome of political games.As a large aarthi said, "[This] is a small town, everyone knows everyone else.So we keep good relations with everyone."This does not prevent them from backing specific candidates though, and it is common to hear of large traders referred to as a politician's "investor," meaning they have donated in cash or kind to individual politician(s).The most far-sighted are those who bankroll politicians with strong links in the bureaucracy so that even if they lose the intra-party or electoral race, they still gain valuable contacts in the bureaucracy. 14Investing in politicians is also good sense because politicians are tied to their popular support bases whereas district officials are transferred frequently and because politicians now informally hold sway over bureaucratic appointments at district level (see Cheema, Khawaja, and Qadir 2005, 28-29).
But while politicians are valuable, parties are less so.Recent research on political parties posits that "parties matter" (Siddiqui, Shafqat, and Mufti 2020, Introduction).However few studies actually document and explain if, why, and how they matter to voters (see, for example, Wilder 1999; Mohmand 2019).Prevailing opinion is that they matter to the extent that they enable supporters to access the state through vertical patronage relations (Wilder 1999;Javid and Mufti 2020, 189).This appears to be true also amongst Punjabkot's business classes, where party loyalty was stronger during the 1970s when they were keen to defeat the Pakistan People's Party's (PPP's) redistributive agenda and again during the two-party political competition of the 1990s.By contrast, party has seemed to matter less in recent years.An aarthi proclaimed: Businesspeople have no party madam!We liked Nawaz Sharif, one of the first things he did was finish [two taxes on goods movement].A Rs. 1,000 tax would turn into Rs.2,000, 3,000, because the man collecting it would ask for whatever he thought he could get away with.It was a real problem for us … Nawaz Sharif understood this because he's one of us, he knows all the problems.Does this mean that he supports Nawaz Sharif's PML-N then?"No, I supported [someone else] in the 2002 elections because we had good relations.Look, it's all about personal relations.No party, no association is strong.It's only as strong as the people in it" (emphasis added).
Despite the fact that most aarthis have an ideological affinity with the PML-N, Arshad sahib's contention that parties are "only as strong as the people in them" and therefore interchangeable is easy to understand.It is a mundane fact of Punjabi politics that political candidates frequently switch parties based on their calculations of personal success.The precedent for this was set in the 1980s with General Zia-ul-Haq's ban on political parties, three rounds of non-party local government elections, and non-party general elections in 1985 (see Waseem 1994;Wilder 1999).These exercises formalised and cemented the regime's relationship with constituencies that had mobilised against PPP leader Zulfiqar Ali Bhutto's populist and redistributive government and a new urban class of businessmen-turned-politicians emerged, sporting distinctly non-ideological allegiances and foregrounding patronage in the political sphere.The political landscape changed definitively and the changes have proved durable.
At the time of fieldwork, all major politicians barring one in the city constituency of Punjabkot had switched allegiances in the last 15 years.At the provincial level, this was most apparent in the run-up to the 2018 general elections when the Pakistan Tehreek-e-Insaaf's ranks swelled with PML-N supporters in seeming proportion to the displeasure that the PML-N leadership incurred with the military establishment.Much has been written about the rise of the so-called electables, a term that refers to constituency elites with money and a robust client base who shop around for the best political opportunities and almost always join the winning party (see, for example, Siddiqui, Shafqat, and Mufti 2020).Given the amounts of money involved in elections and the absence of party funding for campaigns, many of these electables are businessmen; in Punjabkot, the man who won the last two MNA elections is also the city's biggest agro-industrialist with roots in the mandi. 15He entered politics in 2013 as an independent having failed to secure the PML-N ticket and reportedly spent US$1.6 million in a whirlwind three-week campaign, distributing cash by the fistfuls, printing, distributing, and displaying slick posters designed in Lahore, commissioning campaign songs and videos set to Bollywood tunes, and delivering thousands of packets of food to the urban poor in incredibly organised fashion (Amirali 2018, 312).Quickly ensconced within the PML-N immediately after his (and the party's) victory, he retained his seat by a large margin in 2018.Many aarthis anticipated the win and believe that a new political dynasty is in the making.As one aarthi put it: "If this family does politics, the rest will be left in the dust.They have more money than the rest of them combined." Entering electoral contests as an independent is an increasingly popular way to access the windfalls of incumbency and the implications are worrying.Nearly 2,500 independent candidates registered as independents in the 2013 elections, twice as many as in the preceding two elections combined (Dawn, October 30, 2015).In Punjab, 40 were elected and thus won 11% of the provincial assembly seats.By early 2018, only five of these remained independent while the rest had joined a party (Samaa News, June 11, 2018).In 2018, more than 6,000 entered the race as independents, more than twice the 2013 number and more than the total number of candidates -5,500 -fielded by all political parties put together (Geo News, July 11, 2018).Should this trend continue, parties will increasingly become shifting agglomerations of local powerholders clustered around a leading family, something that is already the case but appears set to intensify.Cheema, Javed, and Naseer (2013) and Javed (2018) tentatively suggest that the changing demography of Punjab (more urbanised and with an increased number of seats) and a national media that makes it possible for voters to visually identify leaders, listen to speeches, and so on may help strengthen party identity at the expense of the electables.There are, however, no indications that this is happening amongst the intermediate classes of Punjabkot.

Collective Action
While cultivating strong networks is a key individual strategy of accumulation, collective organisation and action ensure the necessary regulatory infrastructure for business and defend aarthis' corporate interests against the state, labourers, and farmers.The story of the seed dealer narrated above demonstrates how quickly and effectively associational power can be mobilised.Often it is arrayed against the regulatory state at the point of implementation but just as frequently it throws its weight behind the state's ideological projects.Javed's (2018) research describes how urban bazaar traders in Punjab became a powerful mobilised force in the late 1970s and were pacified through incorporation into the power structure through the military regime's local government in the 1980s.The regime's Islamisation programme also fit well with traders' self-fashioned political identity as pious Muslims, and subsequent decades saw them consolidate their position, particularly in light of the trends discussed in the previous section in which political candidates are now self-financing, effectively making politics the exclusive preserve of businessmen and the landed classes.
Today tradersand not just aarthis, but wholesale and retail traders across the countryremain a highly mobilised group.Estimated at 3.5 million firms and generating almost 19% of GDP (Economic Survey 2020), wholesale and retail traders are a significant economic force.As an occupational group, according to one estimate, traders are second only to government employees in their propensity to take to the streets.They accounted for approximately 15% of all protests and strikes in the country between 2010 and 2015 (Javed (2019, 211).They have come out to decry taxation and the breakdown of public services but also to support core state ideologies of Islamisation and anti-Indian nationalism.Organised by locality and commodity within the All Pakistan Anjuman-e-Tajiran, traders' associations are powerful civil society actors.Business associations like these are either conceptualised as rent-seeking entities impeding economic growth as observed by Cheema (2003, 162) and Khan (2000, 19-20) or, from a Gramscian perspective, as corporate organisations enabling accumulation (Harriss-White and Basile 2004).In the latter perspective, and similar to Jessop's (2007) influential strategic-relational approach, the state is not besieged by rent-seekers or interest groups because the state is not assumed to possess a discoverable will independent of the groups and individuals that comprise it.However, this does not preclude the existence of conflict. 16Indeed, Punjabkot's wellorganised and vocal traders' associations have not endeared themselves to the town's administration.A state official remarked: "This trading class is like a rampaging elephant (tajir tabqa badmast haathi ban gaya hey)" while the Assistant Commissioner of the town rued that district administrators "are much less powerful than we used to be … In five years, mark my words, five years, these associations will be dragging us out on the streets."Their power is not only manifest on the streets though (see below), it is also visible in the erasure or much-weakened presence of state authority in the market.
The most important organisation in the mandi is the Mandi Committee (MC) or the collective association of the mandi's aarthis.Capital, biraderi, family, and faction determine its leadership, how things get done and decided, and which issues it chooses to take up (or not).It governs matters of collective concern such as contract enforcement, rules of the auction, use of space, and the entry of buyers.It sets wages for labour and remuneration for farmers and determines and enforces (or not) punishments in case of breach of contract.In short, it substitutes its own regulatory power in lieu of the state's.The MC president and senior vice-president are on the DMC's Executive Council, a number of aarthis are political "investors" and bring pressure to bear when needed, and local tax collectors are kept at bay through a system of regularised illegal monthly payments.The mandi therefore appears to operate as a kind of protected zone wherein the context of regular hoardingfertiliser dealers never issue receipts, farmers sell for prices that are determined by the rules of the buyers, and labour is poorly treated, poorly paid, and entirely unprotected by the law.Order and predictabilityessential for market exchange are ensured by the MC, which is itself organised by the power of capital and custom.Small aarthis are unconcerned with the MCs operation and do not seek to influence it except through collective factional politics in times of organisational crisis, but large aarthis consider it a crucial vehicle for their interests.
Every year or two, the cash-strapped state attempts to increase revenue collection from traders across the country by either expanding the tax net, providing incentives and assurances to encourage compliance, or taking a hard line and threatening punishment for non-filers.3.5% of total tax revenue was collected from retail and wholesale trade in 2019 despite these two activities comprising 19% of GDP (Economic Survey 2020).It is not easy to tax traders.Like traders in other sectors, the provincial aarthi associationthe Anjuman Arthian Punjab (AAP)and the Punjabkot MC have mobilised over the years to defend themselves against taxation. 17The most dramatic of these mobilisations occurred in 2011, when the Federal Bureau of Revenue autonomously and without consultation tweaked the Income Tax Ordinance 2001 to bring agricultural produce traders into the tax net.Previously, aarthis were paying no taxes on income or turnover.The reaction was immediate.The new rules went into effect during the cotton ginning season in January 2011 and aarthis in Punjab's cotton-growing southern districts responded by ceasing all sales.What ensued next was a display of concerted, organised power.Starting in Bahawalpur, the province-wide aarthi association announced a strike.On January 11, mandis across Punjab began shutting down in an unprecedented action that lasted 18 days.AAP office-bearers went town to town, sometimes addressing as many as six rallies a day.GT Road, a major transport artery running through the country from north to south, was blocked multiple times.DMC offices were bolted from the outside, aarthis ceased paying market fees, and DMC staff were intimidated into hiding.In short, Punjab's small-town economy and its hinterlands were thoroughly disrupted.The intensity of the protest was unprecedentednot since the unrest of 1977 which shook the Bhutto regime had mandis witnessed such widespread and continuous agitation.And it ended, as such events do, with a promise and a compromise: the new law would remain dormant and unenforced for the next three years and the AAP leadership would encourage their members to acquire National Tax Numbers and thus register in the tax system.Yet, 10 years later, this law is still unenforced and aarthis have yet to pay any direct taxes on turnover or income.
What should we make of this?Traders' street power is now an established feature of the urban and rurban political landscape (see, for example, Wilder 1999; Javed 2019).Even authoritarian regimes such as General Pervez Musharraf's (1999Musharraf's ( -2008) ) have been unsuccessful in their attempts to document and tax their activities (Javed 2019, 211).As a collective, they defend themselves against the long arm of the state, rarely seek to influence policy, andas Harriss-White (1993, 57) noted of Indian tradersact at the point of implementation.Their relationship with constituency politicians is one in which campaign contributions buy "states of exception" (Roy 2005, 147) and lead to the informalisation of licensed trade.Not surprisingly, the AAP president is an advocate of democracy precisely because the transactional political relationships it is founded on work well for this section of the middle class that has money to spend and laws to avoid.His reasoning is: Democracy [is better for us].This is my personal opinion, but we've also had many debates on this within ourselves in the Anjuman.We've lived through both, and in a military regime meeting a fauji [military officer] is like you're going to see God.They are locked away, we can't reach them without a huge effort.The time has to be fixed exactly, there's so much checking, and then you have to plead.And these days, with the security, we could never meet them.With a political [elected] government, it's easy for us.Round up five to 10 MNAs, MPAs, and we've got a meeting with whoever's in charge.Doesn't mean we always win, but for us, democracy is a ladder.The DMCs also behave differently during a military government, they are more aggressive.With a political government, we are more comfortable.

Conclusion
Twenty years ago, Deshpande (2003, 147) wrote: "[t]he task of identifying the different coherent fractions of the contemporary middle class is an urgent one.Undertaking this conceptual-empirical exercise may, in fact, render the "middle class" redundant" because we can only speak of a middle class if the segments comprising it have more in common with each other than they do with other classes.He proposes a few possible analytical slicings, each producing a different picture: an intelligentsia vs. middle-and lower-middle class slice to distinguish between idea-makers and idea-consumers; an occupational slice to distinguish class fractions based on types of cultural capital; a language slice, performing a similar function as the previous; and a regional slicerural-urban, globalised-metropolitan for exampleto distinguish spheres of operation.
What this article has done is identify a rurban commercial segment deeply rooted in the institutions ordering rural life and focused narrowly on its interactions with the state.The picture is of a politicsabove allof patronage and personal contacts, of family, faction, and biraderi.The circles of trust, accountability, and responsibility are no wider than close kin, immediate locality, and those with whom individuals tend instrumental, mutually advantageous relationships.The authority of the putatively autonomous, democratic state, with its impossible claim to represent the nation and all its fragments, faces fierce competition, a competition that it loses badly as norms and normative evaluations derived from custom lead to regulatory capture, enable and legitimise criminal economic activity, and undermine the principles of equality and justice that it is founded on.Politics stands emptied of ideological content, money is increasingly the basis of political power, and as Wilder (1999, 204) wryly observed, "lawbreakers, rather than law-makers are … what many voters are looking for when they cast their ballots."At the local level, the state seems to have collapsed into society, or to use a well-known idiom, the boundaries between it and society are so blurred as to be indistinguishable (Gupta 1995).This is not unexpected: midway through her book India Working, Harriss-White (2003: 99) asserts that "the Indian State is protean; its form is complex" and that these formsmost if not all of them -"are the self-reinforcing product of a pervasive reluctance to comply with regulative law."Quoting multiple authors, she goes on to underline just how widespread this dissonance between policy and practice is and the many ways in which it has been described and understood: Other students of the local State have written of "ontological incoherence," "multiple-layering," "negotiated grids," "normative pluralism" and "the coexistence of predatory and development States and fractions within States," of "the domination of the developmental State at the centre and the predatory State lower down," the State's "dissolving into the tissue of society at the local level," of "multiple manifestations and interpenetrations" and "particularistic embeddedness" (Harriss-White 2003: 99-100).
Twenty years on, watching this dissonance take its toll in yet another location prompts the same weary question: what is to be done about the fact that state, society, and economy are governed by particularistic ideologies that contradict the formal universalist logic of the state that purports to represent each citizen equally?In the South Asianist and Africanist literature the misfit between formal institutions and the realities of political legitimacy on the ground is recognised and discussed in tones ranging from resignation to celebration: "ordered disorder" (Gayer 2014), "the generalized informal functioning of the state" (Blundo and Sardan 2006, 5), andstriking a curiously celebratory tone -"the politics of the governed" (Chatterjee 2004).The literature on hybrid governance often essentialises informal regulation and misperceives coercion and political capture as popular legitimacy (see Boege, Brown, andClements 2008 anda critique by Meagher 2012).But even the most clear-eyed accounts of this pervasive misfit from within the literature on informality and predatory capitalism only briefly address the question of "what is to be done" (see, for example, Harriss-White 2003; Harriss-White and Michelutti 2019).
It is a difficult question, perhaps the most difficult.Answering it entails, amongst other things, paying attention to the norms that underlie social institutions and working out the extent to which they support mutuality, solidarity, care, and other socially constructive actions.How can these norms be mobilised to improve interactions with formal institutions and move society towards greater social justice and other positive social outcomes?An easier taskone which the analysis in this article prepares ground foris to understand the possibilities for political alliance-making.As Deshpande implies and as Marxian theory has always emphasised, class is a relational concept.Classes exist in relationship to other classes, and the individuals and groups comprising each class must manifest enough of the same interests, lived experiences, and/or collective capacities for political action to qualify as one class rather than another (see Wright 2018).The particular forms of late 20 th and 21 st century capitalism have forced us to journey far from the simple elegance of Marx's two-class model into a much messier landscape of overlapping locations and contradictory class positions.Durable political alliances are correspondingly more difficult to construct but arguably even more important than before because the diffusion of class positions means that no one class is powerful enough to effect social change on its own.This article illuminates the politics of a single fraction of the Pakistani rurban middle classesmore work is needed and it remains a matter of empirical investigation if and how the everyday politics of the salaried, urban middle classes are different than their commercial cousins.Looking at what is known about the Indian middle classes in works by Fernandes (2006), Baviskar and Ray (2011), and Deshpande (2003) amongst others, there is good reason to assume that there will be difference here too, and that this difference will be key to understanding Pakistan's future political development.

Disclosure Statement
This article draws on many of the themes discussed in the author's unpublished doctoral thesis (Amirali 2018).The author would like to thank Barbara Harriss-White for her generous support during the doctorate and Aasim Sajjad Akhtar and Neil Howard for their comments on this article which helped improve it.Financial support by the Pakistan Higher Education Commission and Saint Catherine's College, Oxford is also gratefully acknowledged.
No potential conflict of interest was reported by the author.

Notes
1.The informal economy is the composite of economic activities that take place outside the specified regulatory framework of the state.It is therefore always a deviation from the law, or as Roy (2005, 147) calls it, "a state of exception."This article follows Roy in understanding informality as produced by the state itself whenexercising its sovereigntyits officials grant exceptions and suspend the validity of its previous pronouncements.2. Data were gathered in eight visits lasting between 10 and 14 days at a time.Two follow-up visits were made in 2013 and 2014.3. Eighty-three per cent of land is farmed by those controlling less than 50 acres and more than half of all farms are smaller than five acres (Government of Pakistan 2010).4.There are no women agricultural traders in Punjabkot.5. Profitability varies widely.Annual average rates of interest can range between 60% and 80% (Haq, Aslam, and Chaudry 2013).At the same time, many smaller aarthis are themselves indebted and therefore not accumulating despite high rates of interest.6.This is a particularly apt metaphor for understanding unequal exchange and relations of dependence between different actors and sectors of the economy.I credit Barbara Harriss-White from whom I first heard it.7. Mainstream analysts tend to view clientelism as: an adaptive response to scarce material resources and weak formal institutions (Khan 2005); ineffectively centralised power (Gellner and Waterbury 1977;Bardhan 1988); and high levels of poverty and inadequate public service provision (Wilkinson and Kitschelt 2007;Bardhan and Mookherjee 2012).8. Sahib is a polite title for men.9. Theorised as the "shadow state," the "private interest state," and the "private status state" in India (Harriss-White 2003), in Africa as the "criminalization of the state" (Bayart, Ellis, and Hibou 1999), and the "privatization of the state" (Hibou 1999), and across geographies as the "everyday state," the use of public office to further private interests through selective implementation of the law is a widespread post-colonial phenomenon distinct from corruption and understood in these formulations as a mode of governance.10.Successful accumulation and lower caste social mobility has been observed in a small number of contexts.Hasan (2009, 22) describes it occurring amongst clans of Punjab's Mianwali district who were traditionally transporters and made it big in Karachi.Jan (2018) notes the rise of the Rehmanis in Okara, also seen in Punjabkot.
11.This holds true well beyond Punjabkot.Journalists have noted that "the offices and facilities of the market committees are being used in some areas as party offices for the ruling PML-N, according to sources familiar with the matter" (Express Tribune, August 15, 2011).12. See Harriss-White and Michelutti (2019) on the ubiquity of criminal activity in South Asian politics and economies.Contributions to this volume demonstrate how criminal violence and collusion with the state are critical to processes of accumulation and "criminal" activity is understood as that which is in breach of the law.13.When asked why there was no labour representative, council members laughed and one of the men jokingly motioned at a large landlord member across the room, saying "there's your labour, he's sweat more than anyone in this room today."This provoked uproarious laughter.clearly no one felt the need to make even a purely discursive concession to the interests of labour.14.Former Chief Minister of Punjab, Chaudry Pervaiz Elahi, was considered to have mastered this strategy, if in slightly unorthodox fashion.Multiple sources narrated that he had arranged for the education and upbringing of young girls from his area to eventually marry them off to high bureaucrats and thus strengthen his networks.15.The family still trades in the mandi, but the business has been run by a less fortunate cousin.
Similar to what Jan (2019) observes in Dipalpur, meaningful distinctions between landed and agrocommercial/industrial classes are often not viable.Almost all large landowners in Punjabkot have diversified their portfolios to include storage, agro-processing, or other agro-related activities.16.Roy's (2005) theorisation of informality as a "state of exception" provides a framework for understanding how conflict and complicity exist simultaneously.Public officials inhabit contradictory roles because informality is produced by the state and its officials through granting de facto exemptions from official rules, there is always tension between officially mandated practices and those which deviate and actually come to exist.17.For comparable commentary, see Javed (2018) on traders' anti-tax mobilisations in Lahore.