*Article* **Engaging Smallholders in Flower Agribusiness for Inclusive Rural Development: The Case of Yunnan, China**

**Jieming Zhu 1, Chen Chen <sup>1</sup> and Lie You 2,\***


**Abstract:** Serious farmland scarcities make smallholders a default mode for China's agriculture, which makes efficient and equitable rural development a great challenge. Tensions lead to alternation between autonomous family farming and coordinated collective agriculture. Rapid urbanization since the 1980s has strongly stimulated flower-growing and agribusiness in Yunnan, China. The organization for commercial flower-farming is, however, an issue. Officially promoted for collective farming, voluntary cooperatives are wrecked by the free-riding problem. Grower associations nevertheless spontaneously emerge, with the flexible entry and exit of members without binding jointassets and joint-ownership, which is facilitated by technological changes to the transaction. Empirical investigation in Tonghai, Yunnan, unveiled the institution of agribusiness–smallholder partnership for inclusive rural development. Smallholders have actively participated in flower agriculture, which has contributed significantly to the development of rural economies. The high casualty of micro-smallholders suggests that farm size is an important and crucial factor for sustainable farming. Effective rural development has to be supported by endogenous non-agricultural jobs so that farm size can be increased.

**Keywords:** flower agribusiness; smallholders; inclusive rural development; social well-being; social equity

#### **1. Introduction**

Due to farmland resources made increasingly scarce by burgeoning population growth, China has become a country with one of the lowest arable land areas per capita in the world. In 2016, China's agricultural land accounted for 10.9% of the world's total [1] (https://data. worldbank.org/indicator/AG.LND.AGRI.K2, accessed on 18 March 2019); however, its population was 18.6% of the world total [2] (https://www.prb.org/wp-content/uploads/ 2016/08/prb-wpds2016-web-2016.pdf, accessed on 18 March 2019), resulting in farmland holdings per capita 58% of the world average. Taking the level of urbanization into consideration, China's rural population's farmland per capita is only 9% that of Brazil's, another developing country (https://www.cia.gov/library/publications/resources/theworld-factbook/, accessed on 6 April 2019). Therefore, serious farmland scarcity makes smallholding a default model for China's agriculture.

Efficient and equitable rural development has always been a great challenge to agrarian China throughout upheavals in history caused by cyclical land aggregation and redistribution. Small landholdings because of land scarcity give rise to turbulent changes in agricultural organization. The collective People's Communes, responding to small landholdings created by the 'land to the tillers' reform in the 1950s, were disbanded after existing for three decades and replaced by autonomous family farming in the early 1980s. Commune farms were dissolved into fragmented smallholders again and fragmented families struggled to sustain small farms [3]. Farming cooperatives are actively promoted

**Citation:** Zhu, J.; Chen, C.; You, L. Engaging Smallholders in Flower Agribusiness for Inclusive Rural Development: The Case of Yunnan, China. *Sustainability* **2022**, *14*, 2614. https://doi.org/10.3390/su14052614

Academic Editors: Piotr Prus and Marc A. Rosen

Received: 7 November 2021 Accepted: 26 January 2022 Published: 24 February 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

top-down in order to address the rural spatial and social fragmentation, where rural elites play an important role [4]. This demonstrates that institutions conducive to efficient and equitable rural development are crucial in the setting of severe land scarcity. Intensive farming because of small landholding is productive to a certain extent, i.e., before agricultural involution sets in, which inhibits productivity. Cooperatives and collective self-governance are promising, but local contexts are critical for those schemes to be successful [5]. Contract farming between agribusiness firms and smallholders, the so-called 'core–satellite' model, is prevalent in many developing countries, but its contextual constraints need to be emphasized.

This paper addresses the question of productive rural development in the context of farmland scarcity, because the consolidation of farmland under collectivism or privatization is not conducive to inclusive rural development that should engage smallholders. We identify the institutions and mechanisms for engaging smallholders in rural development by presenting a case of flower-growing agribusiness in rural Yunnan, an underdeveloped province in China. Flower farming has been growing in Yunnan in tandem with rapid urbanization, which results in a large middle-income class with personal consumption commensurate with their tastes and lifestyles. Potentially, huge demand will develop in future; thus, how smallholders on the supply side should be organized in order to capitalize on the market opportunities is at issue. Due to the ubiquitous free-rider problem and substantial transaction costs, smallholders tend to be organized as associations, instead of cooperatives encouraged by the socialist authorities. Smallholder associations and agribusiness firms collaborate based on mutual interests without being bound by contracts, and the agribusiness–smallholder partnership is a helpful institution for inclusive rural development. Despite this, non-farm employment caused by urbanization is still considered crucial for increasing farm size and making rural development sustainable.

The research methodology adopted was an intensive case study that intended to discover insights how institutional change has occurred in the development of the flowergrowing agriculture in Yunnan. During 2017–2019, we conducted several on-site investigations in Dounan and Tonghai, Yunnan, to conduct interviews with smallholder farmers, smallholder associations and agribusiness firms engaged in flower agriculture. There was no overall population information available about the numbers of agri-business firms and smallholder farmers in Tonghai; therefore, we used the snowball technique for sample selection.

#### **2. Institutions for Efficient and Inclusive Rural Development in the Context of Small Landholding**

Efficient production and equitable development are two pressing and challenging goals for both developed and developing countries, although it is well recognized that pursuing these two together is a formidable task for the latter. It is indisputable that sustainable development for developing countries hinges on the efficiency of their economic development. Inefficient economic development wastes resources unnecessarily, adds to tension in social relations and heightens pressure on the eco-environment. Deficiencies of wealth often exacerbate social injustice and environmental unsustainability. Inequality undoubtedly generates social conflicts that undermine the purposes of productivity.

The harsh scarcity of farmland has always been one of the greatest challenges to China's rural development. The widening gap between the 'haves' and the 'have-nots' in terms of farmland holding created many landless peasants prior to 1949. The nature of landlords as a rentier class gave rise to a highly unproductive agricultural economy that made peasants poorer over time. Abject poverty weakened traditional social cohesion in the villages. A natural disaster, such as a flood and drought, was often the straw which broke the camel's back, and peasant uprisings brought down old dynasties and ushered in new ones. The cycle repeated itself in China's history and prevented the country from advancement. Although the farmland holdings by the rich landlords were about 30 times larger than those owned by the poor peasants, which stirred up violent revolution for land redistribution in the early 1950s, the average farmland holding per household was merely 3.3 mu (one hectare is equal to 15 mu) that was achieved by the land redistribution in the course of the 'land to the tiller' reform [6]. So-called landlords, the richest and largest farmland holders among the rural social strata, owned only 26.3 mu (less than 2 ha) on average, considered smallholders in other countries. Even in India as the second most populous nation are farmers with less than two hectares of land considered poor, and thus qualify for cash benefits from the state (Econo-mist, 2019) [6–9]. Scarce endowments of farmland have made subsistence farming a principal mode of agricultural production that give rise to prevalent and persistent rural poverty in China. Techniques of intensive farming have increased land productivity and helped to slightly mitigate the abject poverty, but the growth rate of productivity could not match with that of the population. Agricultural output per land unit has increased, whereas output per capita has remained the same or even dropped; thus, the production mode was agricultural involution rather than evolutional change [10].

The 'land to the tillers' reform dissolved highly concentrated land holdings by the landlords and landless peasants received their fair shares. Nevertheless, subsequent household farming resulted in land fragmentation that prevented the scaled production and adequate provision of farming infrastructure, such as irrigation, that needed collective action. A series of top-down agrarian collective movements eventually created the People's Communes in 1958, when only recently privatized farmlands were converted into collective land ownership. Peasants began collective farming on consolidated farmland with economies of scale. No longer was it smallholder farming, but collective farming, although egalitarian, did not give strong incentives to its members for wealth creation. Subsistence farming have remained a predominant mode of rural economies. Mired in political turmoil and ideological battles, the whole countryside was on the brink of economic collapse and social breakdown by the late 1970s. People's Communes had to be disbanded in the early 1980s and replaced by the Household Production Responsibility Scheme (HPRS), that leased farmland evenly to village households. Village households became basic farming units, and smallholder farming was restored again after three decades of failed efforts in bringing up agricultural productivity.

The HPRS has proven much more productive than People's Communes, but the tension between smallholders and scaled farming lingers. Agricultural involution led by smallholders' intensive farming eventually makes marginal productivity falling to zero. This does not hold promise for sustainable rural development. Voluntary farmers' cooperatives are thus advocated by China's Cooperatives Law (2007) as an alternative for smallholders to collectively scale up their production [11]. As an organizational arrangement, bottom-up cooperatives are characterized as voluntary and open memberships, autonomous management and members' active participation in the economic management [12]. Although a cooperative can serve its members' economic needs in certain social contexts, it as an organization has intrinsic problems breeding members' short-termism. Short-termism is caused by the free-rider problem, horizon, and transaction costs in decision-making [13,14].

When property rights are not transferable and not clearly in the cooperatives of open membership, the free-rider problem occurs [15]. Ambiguous property rights encourage the over-use or abuse of resources where effective control is absent. The horizon problem arises when the productive life of collective resources is longer than the timespan, and a shareholder can claim on the dividends generated, because the shareholder may leave or die [16,17]. Thus, there is a disincentive for members to contribute to long-term investments and an incentive for members to maximize the short-term dividends. Due to the democratic member-controlled system, there could be high transaction costs in reaching management decisions, and the cost of group decision-making increases with the size and diversity of the cooperative [18]. Cooperative management may not be able to react to market opportunities in a timely manner, likely leading to management inaction. An official estimation is that about 40% of China's rural households nationwide belonged to cooperatives by 2015 [19] (http://www.xinhuanet.com//politics/2015-12/08/c\_128509962.htm, accessed

on 20 March 2019). Nevertheless, an in-depth investigation discovered that as many as 96% sampled cooperatives were neither genuine nor successful [20]. The findings are not unexpected. Ostrom [21] advocates the management of collective resources with neither centralized government control nor privatization. However, contextual variables, such as the relative scarcity of the resource and the size of the collective involved, are emphasized as critical for effective collective self-governance [22]. High population density and land scarcity constitute a tough challenge to collective governance over common land resources.

Engaging smallholders for productive agriculture remains a great challenge, whereas small farms are still recognized as important for rural development and food production [9,23]. Contract farming has been considered beneficial to smallholders in general, because such arrangements could facilitate the transfer of technology and provide access to markets to smallholders [24]. Through empirical investigation, Miyata, Minot and Hu [25] determined that the incomes of smallholders who grew apples and green onions in Shandong, China, increased when participating in contract farming. There are also caveats about the drawbacks of those schemes. Relationships established by contracts could be problematically unfavorable to the growers, and the poorest farmers could be excluded in some cases [26,27]. Certain economic, technical and social conditions are crucial for contract farming models to work beneficially for smallholders [28]. Mistrust and a lack of transparency can disrupt contract farming [29]. It appears that market-oriented contract farming is more workable than collective-based cooperatives, because the former tends to have a better access to markets that are crucial for commercial farming.

Figure 1 shows the historical path along which the farming modes have evolved. Farming institutional changes are driven by problem-solving initiatives as well as socioeconomic changes. Smallholder farming creates problems of farmland fragmentation and lacks economies of scale, whereas collective farming (such as the People's Commune and cooperatives) is either low in productivity or breeds the free-rider problem. On the demand and supply sides, urbanization has created a demand for commercial agricultural produce, and has also generated non-farm jobs for peasants who leave farming and pass on their farmland to fellow smallholders, thus increasing the economies of landholding scales. Lowder, Skoet and Raney [30] discovered that the average farm size decreases in most developing countries and increases in many developed countries because of urbanization and industrialization.

**Figure 1.** The evolution of farming modes driven by problem-solving initiatives and socioeconomic changes.

The actual form of institutions for partnerships between smallholders and agribusinesses should depend on the type of commercial commodities, their related market conditions, and the size of family farms. Agribusiness emerges in China responding to robust

demand from the regional, national and even international markets, along with a rising urban middle-class and improved transport connectivity because of consistent investment by the state in infrastructure. Farming is undergoing profound restructuring such that vertical integration occurs in the production process of agricultural commodities where agribusiness firms are linked to farmers [31,32]. Firms have access to markets and productive technologies for commodity quality, whereas farmers have secured farmland thanks to collective landownership. A new institution, agribusiness–smallholder associations, has evolved from the dysfunctional cooperatives, linking smallholders to agribusiness firms so that smallholder farmers can productively participate in the rural development.

#### **3. Emerging Flower Growers and Agribusiness in the Course of Rising Middle-Income Classes**

Yunnan, situated in a high-altitude and low-latitude region, is endowed with a favorable climate which is conducive to flower-growing, although it is a hilly province with scarce arable land for its population. Only 3.3% of its mountainous territory (15,285 square kilometers) are basin plains where the 2.4 million people (2018) can inhabit and farm in Yuxi, one of its prefectures [33] (see Figure 2; http://www.yuxi.gov.cn/yxszfxxgk/tjgb/20 190312/1025161.html, accessed 5 April 2019). Yunnan is one of the poorest provinces in China because of its underdevelopment in industrial economies and the rudimentary infrastructure of inter-regional road networks. Poor connectivity and isolation have hindered Yunnan's development. Historically, farmers did not grow flowers as commercial crops because there was no effective market demand when urban consumers consisted of small and scattered populations.

**Figure 2.** Location of Yunnan in China, with Yuxi and Tonghai shown.

In Tonghai, we interviewed 96 smallholder farmers, 13 smallholder associations and 7 agribusiness firms that are engaged in flower-growing agriculture. It was reported that flower-growing started by a peasant named Hua Zhongyi in 1983 on a tiny family land plot of 0.3 mu in Dounan township in the suburbs of Kunming, the provincial capital of Yunnan [34]. This was soon followed by a few villagers in the vicinity after a street market evolved in Dounan, where Kunming urbanites came to buy the flowers from the local growers. That was the beginning of Yunnan's flower plantation, at the humble size of 40 mu in Dounan by 1987. Thirty years later, the flower plantation areas had increased in size to 1,561,500 mu in 30 counties by 2017 (see Figure 3). Flower farming accounted for 11.6% of the total agricultural outputs in 2011, and the figure grew to 13.0% in 2017 [35,36]. The flower-growing cluster around Dounan mainly produces fresh-cut flowers that are transacted daily in Dounan Market within a three-hour driving distances, whereas farming clusters in the west region grow flowers for petals processed into products that do not rely on Dounan for swift delivery to customers.

**Figure 3.** Location of 30 flower-growing counties in Yunnan. Note: The black dot indicates Dounan Market.

The demand for bouquets of fresh flowers has been strengthened by a rising urban middle-class along with rapid urbanization since the 1980s. The China Development Research Center of the State Council estimated that middle-income classes accounted for about 10.0% of the total population in 2002, and the figure rose to 28.9% in 2016 (http: //www.drc.gov.cn/xscg/20180911/182-224-2896926.htm, accessed on 22 March 2019). During the period 2002–2016, the absolute number of middle-class people increased by 270 million, based on the above estimations. One of the most significant events for Yunnan's flower agribusiness has been the establishment of flower markets that link supply to demand. As early as in 1993, a few Dounan villagers ventured into marketing and trading in large cities such as Chengdu, Wuhan, Guangzhou, Beijing and Shanghai, where they set up shops promoting sales of Dounan flowers. At its peak, Yunnan contributed 70% of fresh-cut flowers to the markets in the major Chinese cities. This specialization added value by fetching higher prices of flowers than what could be obtained in Kunming, while those traders leased their own farmland to the fellow villages at home (APKPG and KRICH, 2018). The node linking flower supply to potentially substantial demand is the Dounan Market. The initially rudimentary street market had evolved into an indoor marketplace at a size of 3.7 ha in 1999, and the Kunming International Flora Auction (KIFA) trading center started operation in 2002, when flower transaction volumes had increased substantially. By reducing potentially high transaction costs associated with perishable commodities, KIFA has proven to be the most effective measure promoting flower transactions that increased in value by 15 times, growing at 20.2% per annum between 2003 and 2018 [37]. Another auction market, DFETC, commenced operation is 2015, when KIFA was reaching full capacity. Dounan provides crucial market accessibility to Yunnan's flower agribusiness.

#### **4. Organization for Flower-Growing Micro-Smallholders: Cooperatives or Associations?**

Tonghai is a county in Yuxi prefecture, about 140 kilometers to the south of Kunming (see Figure 2). Approximately 77% of its territory (721 square kilometers) is composed of mountainous areas; thus, farmland is extremely scarce against its farming population (see Figure 4). Farmland holding is 0.7 mu (0.05 ha) per capita and 2.1 mu (0.14 ha) per rural household on average [38]. Its agricultural industry consists of smallholders that are considered to be micro-size, as a result of socialist collectivism that distributes farmland equally. A few farmers started growing flowers in the late 1990s, spurred by the Dounan Market, which is easily accessible. Bi Chaochun was one of them. After two years of flowergrowing on his own tiny land plot, Bi believed that smallholders should be organized collectively to embark on market-oriented commercial farming, and initiated the Lianqing Rose Cooperative together with 42 other micro-smallholders on a total aggregated area of 76 mu (5.1 ha) farmland in 2004 (1.8 mu/member), so as to improve agribusiness efficiency by increasing economies of scale. There were 45 flower-growing cooperatives, genuine or nominal, established during 2002–2017 in Tonghai, according to revelations by an informant. In total, 4791 micro-smallholders joined those so-called cooperatives, the largest having 582 members and smallest having 5 members [39]. By estimation, the farmland scale per cooperative was about 220 mu (15 ha) on average; the largest was 1220 mu (80 ha) and the smallest was 10 mu (0.6 ha).

**Figure 4.** Locations of sampled agribusiness firms and cooperatives/associations in Tonghai.

With only a tiny fund of CNY 30,000 (USD 4500 at the exchange rate in April 2019) collected from the members as collective investment, Lianqing managed to rent a disused and dilapidated temple for his office and storage. The cooperative organized sessions for members to learn about commercial flower farming and managing technical problems, such as temperature control and pest-prevention. Fresh-cut roses were assembled, packaged and transported to Dounan by the cooperative, and sold under the same brand that was the namesake of the cooperative. By 2007, Lianqing had expanded to 140 members and 260 mu farmland (1.9 mu/member). Due to tiny household landholding farming, growers tended to farm the land intensively with heavy applications of fertilizers and pesticides so as to increase flower productivity. After two to three years' intensive farming, the land became salinized, and the quality of crops declined drastically. Farmers would not have any income if lands had to lie fallow for a year or two, or farming continued and yielded low-quality flowers. Members whose lands were not yet salinized still produced good-quality roses. Collective sales distributed sale proceeds equally per bundle between members, regardless of quality variations in the roses. It thus generated the free-rider problem, and it did not give incentives for members to keep up the good quality of rose growing. Free-riding was exacerbated by the expansion of the cooperative that was too diversified with too many members. The cooperative could not hold members together. By 2014, Lianqing was all but dismantled, broken down by the problems similar to People's Communes, although Lianqing was spontaneously organized in a bottom-up manner. The same fate befell Lianxin (51 members initially), another cooperative interviewed.

We interviewed 13 so-called cooperatives in early 2019, and found that only two of them, Lianqing and Lianxin, were genuine cooperatives with collective investment and cooperative assets (see Figure 3). It is not surprising that most of the cooperatives are not genuine, elaborated by Hu, Zhang and Donaldson [20] in their extensive investigation. Many nominal cooperatives are actually associations. Most organizations for flowergrowing micro-smallholders are therefore associations in nature. Associations are flexible in terms of the entry and exit of members, because there are no binding joint-assets and joint-ownership. Associations provide services paid for by members who use them with no other obligations. Sold in Dounan under the brand provided by the association, flowers are collectively transacted through the association's trading account, and the same free-rider problem can occur, as with the cooperatives. Although associations have a mechanism to remove undesirable members who provide inferior products to spoil association's collective sales, this exercise could not be easy in village settings where people tend to be socially close. However, those with better flowers can leave the associations, and only 'lemons' stay on. The sustainability of associations depends on the collective management of flower quality.

Based on the list compiled by the Agricultural Office of Tonghai County Government, 45 cooperatives or associations were founded during 2004–2017; 32 of them are still in operation, and 13 have closed. A significant event occurred on 2015, when another auction center, DFETC, was opened. This ushered in a technological change that offers every flower smallholder an individual trading account; thus, flowers are no longer transacted collectively under association accounts. This solves the free-rider problem that destroys cooperatives and associations. The survival rate of associations has improved remarkably since 2015. Before 2015, the survival rate was only 62.5%, and after 2015, it increased to 92.3% (see Table 1). This technological change allows free association, and thus helps hold micro-smallholders together as farming communities. Our on-site investigation revealed that associations do benefit farmers greatly, through the sharing of farming techniques and expertise between fellow members. The number of members frequently communicating with fellow members for information exchange accounts for 82.3% of the total, with only 4.2% who never communicate with other members. Without the free-rider problem besetting the organization, associations of free membership become real social communities of smallholders for mutual help and collective learning (see Table 2).

**Table 1.** Survival rates of cooperatives/associations.


Source: Authors' 2019 fieldwork.


**Table 2.** Sources of consultation to smallholders for flower-farming techniques and expertise.

Source: Authors' questionnaire survey of smallholders in January 2019.

#### **5. Institution for Inclusive Rural Development: Agribusiness–Smallholder Partnership**

Agribusiness firms usually provide flower-packaging and flower-delivery services, quality seedlings and training and consultation to the smallholder associations. Between 2006 and 2015, seven flower-farming firms grew out of associations (see Table 3). The initial organizers of those associations were fledgling entrepreneurs who managed the associations by providing services in addition to their own flower-growing. The services expanded after a few years in business, which helped produce agribusiness firms. Associations have become incubators in this case. In addition to this paradigm, there are two other models for firm–association unions. One is agribusiness firms setting up associations to recruit steady clients as an important part of business operation, and the other is firms linking up associations as joint-partners (see Table 3). Still providing services of flower packaging and transportation to Dounan Market, agribusiness firms have expanded their landholdings so that smallholders are transformed to commercial farms. Ten agribusiness firms have aggregated 8080 mu (539 ha) farmland in total, each having 800 mu on average; the largest at about 3500 mu, and the smallest at 50 mu. Large-scale farming allows firms to invest in farming facilities and R&D for growing high-quality flowers. Large firms also diversify into specialization in developing rose seedlings that are provided to smallholder farmers, so as to extend the flower production chain that enhances flower-farming productivity.

**Table 3.** Formation of agribusiness–association partnerships.


Source: Authors' fieldwork conducted in January 2019.

Based on a sampled survey conducted in January 2019 of 96 flower-farming smallholders, we discovered that the average landholding area per farmer is 3.8 mu (0.25 ha). Almost half have as little as 1.7 mu, with the rest clearly renting farmland to increase their own family land areas from villagers who do not farm (see Table 4). Farmland transfer by leasing has been facilitated by an institutional change to collectively owned farmland [40]. According to China's constitution, there is a dual landownership scheme: urban land belonging to the state, and rural land being collectively owned. Since the founding of the People's Republic in 1949, farmland owned by landlords has been confiscated and re-distributed to landless peasants under the Rural Land Reform. In 1956, the People's Commune movement abolished the private ownership of farmland and replaced it with

collective ownership. Although rural land is collectively 'owned' ostensibly, rural communities 'own' the land on the condition that it is only used for their own production (farming and non-farming) and habitation. It is considered that in the countryside, lands are held by villages based on the Marxist principle of land as 'a means of production', which suggests that land should not be considered an economic asset. The notion of land being 'a means of production' grants land use rights only to the holders; the collective has neither the right to derive income from land by letting it out, nor the right to change its form and substance by developing it for non-agricultural uses without approval from the government at the county level or above. The right to alienate collective land for urban uses is restricted in situations where the other party in the transaction is the state. Therefore, collective rural land as assets is owned by the state. Only recently has agricultural land leasing been allowed. The conversion of farmland to nonagricultural uses for leasing remains illegitimate and controversial. Farmland transfer by leasing allows subsistence farming to evolve into commercial agriculture, in the setting of urbanization that creates non-farm jobs for rural–urban migrants. Some farmlands are left behind for sub-leasing as those farmers have migrated to cities to take up urban jobs.

**Table 4.** The size of landholdings for the sampled smallholders.


Source: Authors' questionnaire survey of smallholders in January 2019.

Substantial expenses occur upfront for building greenhouse facilities, and daily maintenance costs on fertilizers and pesticides are two great concerns by those farmers. There is a marked cleavage between the size of landholdings in relation to concerns over the costs of fertilizers and pesticides. Those farmers under heavy cost pressure with land areas under 4 mu account for 70% of that cohort, and with between 4–8 mu 58% of that group. Only 36% of those above 8 mu are worried about these expenses. It is likely that smaller farms will utilize more fertilizers and pesticides in order to raise productivity. The concern over the upfront outlay on building greenhouses is negatively correlated with the number of years growing flowers. Three-quarters of the farmers who have been involved in flower-growing for 1–2 years strongly feel the burden. The ratio declines to 50% for those who have farmed for 3–9 years, and down further to one-quarter of those who have worked for over 10 years. This suggests that crucial and expensive greenhouse facilities are gradually improved over the years, along with the savings accumulated.

The years spent as a flower-farming smallholder is positively correlated with the size of the landholding (see Table 5). The longer the farmers have been involved in growing flowers, the larger their land holdings. This suggests that the casualty rate should be high for smallholders in their early years of farming, due to their tiny farmland plots and the intensive farming approach. The initial outlays are relatively high in relation to the outputs restricted by the small size of the farmland. Inadequacy in facilities is compensated by the heavy usage of fertilizers and pesticides. This unsustainable farming method expedites land salinization, and the quality of crops degrades in a downward spiral. In the KIFA auction center, high-quality roses (classes A and B) only accounted for 30%, and low-quality (classes C, D and E) comprised the remaining 70% in 2018 (unpublished report by KIFA, 2019). The average prices of low-quality roses are only 46% of those of the high-quality ones. Those low-quality roses are most likely supplied by smallholder farmers; the data reveal that 70% of flowers in the Dounan Market are provided by smallholders [34]. A vicious cycle is created by unsustainable intensive farming and ended in shrinking incomes, a kind of involution associated with small-landholding. In the first few years, many microsmallholders are squeezed out of the flower market. The composition of associations is fluid, with a high frequency of entry and exit of members, which also explains why flower-growing cooperatives are not sustainable. The chance of smallholders surviving the first five years is only one in three (see Table 5).

**Table 5.** The number of years in flower-farming and the landholding of micro-smallholders.


Source: Authors' questionnaire survey of smallholders in January 2019.

The severe scarcity of farmland and resultant micro-smallholdings represent great challenge to sustainable and inclusive rural development. Associations seem more applicable than cooperatives to flower-farming by micro-smallholders. Agribusiness firms with large farmland stocks have thus become crucial anchors for flower-growing agriculture. With financial capacity, firms have invested in techniques of wastewater recycling and soilless cultivation, which prevent the pollution of soils, so that land degradation is mitigated. Agribusiness firms also make great efforts in cultivating new varieties of roses, which allows growers to command price premiums. According to a report by KIFA, there were 200 kinds of roses available in the auction market in 2016, and the figure rose to 356 in 2018, a valuable contribution by the flower-growing firms (unpublished report by KIFA, 2019). Auction centers also serve as protectors of intellectual property over newly invented flower varieties. Royalties are automatically deducted from the proceeds of sold flowers, therefore the practice promotes research and invention. Flower-growing firms have become steady providers of quality fresh-cut flowers in a market that is expanding daily.

Due to volatile fluctuations in market prices and the perishable nature of fresh-cut flowers, contract farming is not prevalent for flower-growing smallholders. The agribusiness– association model works well. Agribusiness firms provide commercial services and professional advice, whereas farmers' associations provide a platform to smallholders for mutual help and fellowship. Smallholders are effectively engaged in rural flower-growing development, although smallholders need to work competitively and productively to gradually expand the size of their landholdings. We have derived an agricultural development index (ADI): the formula for ADI is 0.8 × agricultural product per farmworker/50,000 + 0.2 × farmland per farmworker/50, where CNY 50,000 agricultural product per farmworker is designated as 100 and 50 mu farmland per farmworker as 100. This tool measures farming-based rural growth. The ADI stood at 27.2 in 2010, and it rose to 71.8 in 2017, a substantial improvement. Flower agribusiness is promising for rural development.

#### **6. Sustainable Rural Development: Endogenous Non-Agricultural Jobs**

Flower-growing agribusiness has changed the rural economy of Tonghai significantly. Agricultural outputs per capita increased by 181.7% between 2010 and 2017. However, Tonghai is still predominantly an agrarian economy, where the agricultural labor force represented 61.6% of the total labors in 2017, only declining slightly from 66.0% in 2010 [41]. The state of micro-smallholding remains little changed. Smallholder subsistence farming is not promising in leading peasantry to prosperity. Non-agricultural economies seem the only way to reduce the number of farmers so as to increase the scale of farms for high-density land-scarce regions, paving the way for effective rural development. The shift of the agricultural labor force into non-agricultural sectors is considered de-agriculturalization [42], and many land-scarce agricultural countries have made efforts in searching for non-farm income sources [43]. Sharma [44] claims that India's urban settlements are keys to development and inclusive growth in its rural areas.

Rural industrialization since the early 1980s has had great economic and social impacts on China's countryside under the auspices of grand economic reforms. The collective township–village enterprises flourished, providing industrial employment to those who were underemployed in farming [45]. Although collective enterprises reached the zenith in the mid-1990s and declined thereafter, non-agricultural jobs grew continuously. Most of those privatized enterprises remained in operation in the rural areas. In 2016, Kunshan county, in the Yangtze River Delta, saw only 8.0% of its total laborers working in the agricultural sector, and Nanhai county, in the Pearl River Delta had as low as 4.7% after four decades of rural industrialization [46,47]. The Yangtze River Delta and Pearl River Delta are the two of the most dynamic industrializing regions; however, rural industrialization is not evenly developed across the country. A regional disparity is clearly displayed in a similar pattern as national economic development, where the eastern regions are the most dynamic and the western regions the least developed.

On the one hand, in situ urbanization provides non-farm jobs to peasants in dynamic regions that witness urbanization-led rural development. On the other hand, urbanization offers numerous economic opportunities to those seeking non-farm employment. According to the 1990 population census, the number of peasant migrant workers (intercounty migration) was estimated at 13.0 million, which accounted for 1.5% of the total rural population then. This ratio rose to 6.3% in 2000 [48,49]. The number of peasant migrant workers jumped to 168.8 million, representing 28.0% of the total rural population in 2015 [50,51]. Rural–urban migrants should have become better-off economically than they were in the countryside. Nevertheless, the hardships associated with migration have been overwhelmingly reported by both the conventional and social media. Those peasants in the less dynamic regions need to travel longer distance to find non-farm jobs. Among 166.1 million peasant migrants, 46.6% and 53.4% of them were inter-provincial and intra-provincial travelers, respectively. Rural–urban migrants seem to be traveling shorter distances recently. Rural non-farm laborers working in their hometowns stood at 36.8% of the total in 2009, which rose to 39.2% in 2015 [51,52]. Changing preferences suggest a new generation of rural migrants making balanced choices between economic benefits and social costs in a new set of conditions that are quite different from those made decades earlier. In Tonghai, 14.0% of rural labors took on non-farm jobs by moving to urban areas, and those inter-county migrants only accounted for 2.9%; and those within Tonghai accounted for 11.1% (2016) [53].

Yunnan is an agricultural province with weak secondary and tertiary industries, and 85.0% of its labors were involved in farming in 1980. The proportion only dropped to 50.8% in 2017 after nearly four decades of market-oriented development [36], whereas it was 11.8% in Zhejiang and 16.8% in Jiangsu in 2017, two provinces in the Yangtze River Delta [54,55]. Exogenous non-agricultural development led by outside investments has become few and far between since the new millennium. Localities, whether advanced or underdeveloped, need to rely on bottom-up endogenous growth initiatives. Liu [56] described how Taihu region in the Yangtze River Delta had developed silk and textile cottage industries since the Ming dynasty, driven by the ever-deteriorating farmland resources. Many small towns specializing in making and trading those products emerged, connected by canals for interlinkage. The processing of the agricultural produce added value and production chain was extended to allow specialization. Those towns became the rural centers of manufacturing and trading, stimuli for rural agricultural and non-agricultural growth.

The initial farm size is a crucial factor for the survival of smallholders. It has been discovered that the use of fertilizers decreases along with the increase in farm size [57]. Wu et al. [58] found that there is a strong inverse-correlation between the farm size and use intensity of agricultural chemicals in China, and "a 1% increase in farm size is associated with a 0.3% and 0.5% decrease in fertilizer and pesticide use, respectively, and an almost 1% increase in agricultural labor productivity" (p. 7010). The relationship also exists in other countries, based on an international panel analysis of 74 countries, also proven by an

empirical survey in Tonghai. The total stock of farmland is fixed, which suggests that jobs should be created to attract farmers out of flower-growing.

The production chain and value chain of flower agribusinesses can be extended so as to deepen specialization of the links in the chains. It is promising to see that some endogenous non-agricultural jobs have been created in the areas of flower-packaging, flower-delivery, e-commerce (i.e., selling flowers online), and the making of rose cakes and rose essences. Rose cakes are already available in the retail markets of major cities in China. Research and development in cultivating quality flowers, seeds and seedlings, as well as developing growing techniques, are crucial steps which further the deepening of the industry value chain and flower clusters. Flower firms' managers have complained about poor infrastructure and facilities that impinge on the progress of flower industries, and the difficulties in recruiting educated workers. Town amenities, or lacks thereof, have become one of the main stumbling blocks to attracting professionals to work in agribusiness firms that are usually located in rural areas. Improving town facilities and amenities should promote flower agribusiness and thus inclusive rural development.

#### **7. Conclusions**

Serious scarcities of farmland make smallholding a default model for China's agriculture, posing a great challenge to efficient and equitable rural development. Subsistence farming thanks to smallholders has given rise to prevalent and persistent rural poverty in China. It seems that intensive farming, although helpful, tends to lead to agricultural involution because of the slower growth in productivity than that of population. Tensions lead to alternations between family-coordinated farming with even landholding, and community-managed farming with aggregated landholding. Rapid urbanization under the grand economic reforms has changed the conditions for China's agriculture. A growing middle-income class triggers the development of commercial farming.

Flower-growing and agribusiness in Yunnan provide an enlightening case to shed light on how commercial agriculture should be organized with a backdrop of pervasive smallholders. Farming modes are adopted and heavily dependent on what agricultural produce is available and its accessibility to the market. Yunnan flower growing is a case for reference. Officially promoted so as to encourage collective farming, voluntary flower-growing cooperatives are crippled by the free-rider problem. Associations of growers nevertheless spontaneously emerge with the flexibility of entry and exit of members without binding joint-assets and joint-ownership. Technological innovation in the transaction market also critically facilitates the operation of associations. Partnership between agribusiness and smallholder associations have become the institution for inclusive rural development. Agribusiness firms with large farmland stocks are anchors for the flower-growing agriculture, whereas memberships of smallholder associations are fluid, with a high occurrence of entry and exit, because of unsustainable farming adopted by micro-smallholders. Although smallholders have actively participated in flower agriculture, which has contributed significantly to the rural economy, the high casualty of micro-smallholders remains an issue. Sustainable rural development must be supported by endogenous non-agricultural jobs so that exiting farmers can transfer lands to growers in order to increase the landholding size of farms. Through a detailed case analysis, this paper examines the endogenous development logic of farmers, agricultural films and agricultural associations, which enriches the theory of inclusive rural development.

**Author Contributions:** Conceptualization, J.Z.; Data curation, L.Y.; Formal analysis, J.Z., C.C. and L.Y.; Funding acquisition, C.C. and L.Y.; Investigation, C.C. and L.Y.; Methodology, J.Z.; Writing original draft, J.Z.; Writing—review & editing, J.Z. and L.Y. All authors have read and agreed to the published version of the manuscript.

**Funding:** National Natural Science Foundation of China (NSFC51978478); National Natural Science Foundation of China (NSFC51908350).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

