A review of rural transformation studies: Definition, measurement,and indicators

Dong WANG, Chunlai CHEN, Christopher FlNDLAY

Crawford School of Public Policy, Australian National University, Canberra 2600, Australia

Abstract The paper reviews studies of rural transformation in three dimensions-definition, measurement, and indicators-and summarises the findings of rural transformation research.The scope of rural transformation includes four elements -productivity, rurality, inclusiveness, and sustainability.Current concepts of the dimensions of rural transformation and their associated indicators are insufficient for policy decision-making, as they lack objectivity, feasibility, accountability,comprehensiveness, and comparability.Future research to develop new measures to assess rural transformation in developing countries is valuable.Furthermore, there is potential to explore the topic in some directions: urbanisation strategy, public intervention (i.e., institution, policy, and investment), gender inclusiveness, market creation, and international trade.

Keywords: rural transformation, agricultural transformation, structural change

1.lntroduction

Rural transformation involves sweeping changes.These include shifts in the practice of farming, a change in the use of rural areas, the movement of people across locations and sectors, and new dynamic interactions between the primary and other sectors.The rural transformation brings productivity improvements, develops the non-farm economy, and generates benefits for migrants and their communities of origin and destination (FAO 2017).It can be a force for greater inclusiveness.Our interest in this paper is the nature of rural transformation and how to assess its impact, especially its success.

The success of rural transformation can increase the welfare of the entire society in several ways: poverty reduction (Imaiet al.2017; Liuet al.2018; Benfica and Henderson 2021), displacement of farmers from farming (Zhanget al.2018, 2021), diversification and commodification of agricultural production (Keyder and Yenal 2011; Lanjouwet al.2013; Liet al.2016; Ecker 2018), food security and dietary transition (Masterset al.2018; Schmidtet al.2020), improvement of health and education (Nguyen and Tran 2014; Reardon and Timmer 2014; Brown and Guinnane 2018; Johnson and Taylor 2019), greater inclusiveness through the provision of justice and contributions to gender balance (Mu and van de Walle 2011; Zhouet al.2016; Heckertet al.2020), and ultimately, productivity growth and structural change of the whole economy (You and Sarantis 2013; Kaginet al.2016; Emerick 2018; Gong 2020).Conversely, the world cannot achieve the UN 2030 Agenda and 17 Sustainable Development Goals (SDGs)1The SDGs are to “eliminate poverty and hunger (SDGs 1 and 2), improve the health and education of citizens (SDGs 3 and 4),promote gender equality (SDG 5), improve the use and access to water (SDG 6), promote clean energy (SDG 7), generate decent employment and growth (SDG 8), innovate in productive processes and infrastructure (SDG 9), reduce inequality (SDG 10), build sustainable cities (SDG 11), achieve sustainable production and consumption (SDG 12), fight climate change(SDG 13), conserve marine and terrestrial ecosystems (SDGs 14 and 15), promote peace (SDG 16) and international cooperation for development (SDG 17)”.without successfully transforming rural and rural sectors (Trivelli and Berdegué 2019).The World Development Report regards rural transformation as an engine of growth for ‘agriculturebased countries’ (Alainet al.2008).Diversification of agriculture and inclusiveness of rural transformation would give smallholder households access to sources of income beyond agriculture (de Janvry and Sadoulet 2020).

Not all countries have experienced rural transformation successfully.In the preindustrial era, China had kept ahead of many nations for thousands of years in farming.However, the rural transformation did not happen in China until industrialisation began only recently.According to Hou (2008), in China, an increase in population density only resulted in a dynastic cycle, that is, periodic dynasty changes associated with political and economic chaos (Chuet al.1994), rather than reallocating excessive labour force to trigger industrialisation.This occurred because labour productivity had not increased in line with the increase in population density but ‘stagnated’, and the handicraft industry (the basis of a transition to industrial growth) was never disconnected from agriculture (Elvin 1973).The underlying reason for this stagnation could have been institutional barriers, such as the lack of property rights protection and the burden of the land tax regime.

Another example is Indonesia’s ‘agricultural involution’in the 17th and 18th centuries.Most of Java was ‘crowded with post-traditional wet-rice peasant villages: large,dense, vague, dispirited communities - the raw material of rural, non-industrialised mass society.’ (Geertz 1963).The main reason for the lack of progress was that the Javanese agricultural economy was closed.Given that external demand for agricultural products was limited and there was a lack of technological progress, the internal increase in population density led to the intensification of agriculture rather than structural change.Output per unit of the area increased, but not per capita.

In these two cases, rural transformation failed,associated with a lack of technological advances and no industrialisation.Even worse than occurred in these two cases, for example, some other countries could fall into a Malthusian Trap (Malthus 1798), in which the population,as it grew, could not feed itself.

In the modern era, we also witness several failures of rural transformation in developing countries, though the degrees vary.For example, in sub-Saharan Africa,the increase in per capita GDP from 2000 to 2014 did not reduce poverty, and agricultural productivity growth remains low (Dercon and Gollin 2014; Wuyts and Kilama 2016; Barrettet al.2017).Although the share of agriculture in GDP and employment declined, the off-farm labour has not moved towards manufacturing but to nontradable services (IFAD 2016; Rodrik 2016).

These experiences challenge policymakers to reflect on rural transformation in practice and principles.Not all the changes in a rural area should be taken for granted as examples of thriving rural transformation.Success cannot be achieved without proper policy facilitation based on a better understanding of its nature.However,the relevant policy portfolios are diverse, depending on the issues in focus.For example, one concept of rural transformation originates from the technological change induced by relative price change and influences the rural sector’s competitive advantages and economy (Acemoglu 2002, 2003), which expands production (Cochrane 1993)and captures economies of scale.This process, however,is likely to be confronted with market failures, and policy intervention is needed to address the distortion.A political economy framework, however, emphasises who controls and benefits from rural transformation and who would be the losers.In this regard, the policy focuses on gender,generational, and ethnic justice rather than efficiencydriven corrections of market failures.

2.Definition

As Weisheitet al.(1995) said, “Everyone knows the term rural, but no one can define the term very precisely.” One reason is that the perspective can differ, for example, that of researchers focused on urban organisation, and the drivers of change from events in rural areas versus their counterparts who prefer to begin with rural and regional sociology (Lobao and Meyer 2001).

According to some, rural transformation must be understood in terms of the technological revolution and industrialisation (Johnston 1970; O’Brien 1977).In this regard, rural transformation and agricultural reform are preconditions of industrial development (Chang 1949), and the industrial and agrarian revolutions always go together (Lewis 1954).The ultimate goal of rural transformation is to “achieve industrialisation, not only to build an industrialised city but also to build an industrialised agriculture” (Zhong 2016).Without the Industrial Revolution and the grand technological leaps,agricultural labour productivity is unlikely to increase.The initial productivity gap drives farming intensification,production specialisation, agrifood commercialisation and rural structural diversification.The productivity gap between the rural and urban sectors results in resource(labor, capital, etc.) flows and reallocation across regions and sectors, which furthermore leads to a more fundamental structural change in the economy (Page 1996; Bartelsmanet al.2013; Gollinet al.2014; Hickset al.2017; Tombe and Zhu 2019).Trading activities can speed up the procedure by reinforcing the comparative advantages (Anderson 2009; Bustoset al.2020).Factors like institutions, policy, and investment vary across countries and can catalyse a thriving rural transformation(IFAD 2016; Huang 2018a; Huang and Shi 2021).

In this vein, understanding rural transformation requires recognising the links between rural transformation,agricultural transformation, and structural transformation (or structural change).Jayneet al.(2018) regard agricultural transformation as embedded in rural transformation, while Berdeguéet al.(2013) say that rural transformation is embedded into a broader process of structural change.Timmer (1988) also states that agricultural transformation is embedded in the more comprehensively structural and rural transformation processes.Therefore, we can illustrate the scope and boundaries of those three definitions in Fig.1.Both agricultural and rural transformation belong to a structural transformation: the reallocation of economic activities away from the rural to industry and service sectors.It involves increasing productivity growth across industries, expanding the urban economy, declining the share of agriculture in GDP, expanding domestic and international trade, and increasing specialisation and division of labour.In the long term, it leads to increased migration of people from rural areas to urban centres and urbanisation of the countryside, usually combined with a reduction in birth rates, greater participation of women in the workforce, and profound political and socio-cultural changes (FAO 2017).

Fig.1 The realm of structural transformation, rural transformation,and agricultural transformation.Source: Authors’ own summary and illustration.

We now pay more attention to both agricultural and rural transformation.The context of structural transformation is important to recognise but remains too broad for useful policy design.Agricultural transformation is a critical element of rural transformation but not quite broad enough to capture all the interactions of interest.

2.1.Agricultural transformation

The initial driving force in agricultural transformation is agricultural productivity change driven by technical innovation, economies of scale, urbanisation, or marketisation reform (Jayneet al.2018).Productivity enhancement generates higher value returns to the producers.On the one hand, more earnings from farming stimulate demand for goods and services that induce off-farm employment.On the other hand, following the development of markets and trade, farmers have a greater incentive to produce crops of higher surplus due to the principle of comparative advantage and economies of scale, which leads to specialisation within farming (Pingali 1997; Coelli and Fleming 2004).As a result, there are fewer farmers with larger farms.Furthermore, the increase in labour productivity drives labour force migration intrasectorally and inter-regionally, moving from farming to the manufacturing and service sector and from farms to cities(McMillanet al.2014).This process of migration boosts the prosperity of industry in the urban area.

Timmer (1988) says that in agricultural transformation,“the share of agriculture in a country’s labour force and total output declines in both cross-section and time-series samples as incomes per capita increase.” He uses the degree of agricultural diversification (i.e., value-added per agricultural worker) to measure this transformation.This definition is quantitative and makes agricultural transformation accountable.The economic meaning behind the measurement is “the process by which an agrifood system transforms over time from being subsistence-oriented and farm-centred into one that is more commercialised, productive, and off-farm centred” (Jayneet al.2018).Reardon and Timmer (2014) further point out that the agrifood system transformation is an integrated system that includes five key components - agricultural transformation, supply chain and retail revolution, dietary patterns change, urbanisation, and the integrated factor markets.Given that the components can interact and reinforce each other, the transformation potentially can be very rapid and complicated.Some research also adopts“agrarian transformation” to describe the transformation occurring in rural areas (Quinn 1997).In this branch of literature, agricultural transformation is more about “farming”and associated supply chain issues (Timmer 2014).

The FAO (2017) definition of agricultural transformation considers both cause and effect.It describes the agricultural transformation as “a process that involves a shift from mainly subsistence farming to commercial, highly diversified production systems” and says that agricultural transformation is driven at the individual farm level to the macroeconomic level, demonstrated by the specialisation of farming and diversification of the economy.It says, “The process favours specialisation, which allows economies of scale through the application of advanced technologies and modern delivery systems for both inputs and outputs;this, in turn, promotes tighter integration of a more diversified farming sector with the rest of the economy and with international markets.”

2.2.Rural transformation

Rural transformation covers more than agricultural transformation, though it captures all aspects of it (Fig.1).Huang and Shi (2021) say that rural transformation involves a combination of agricultural transformation plus the growth of the non-farm employment of rural labour.It includes the change occurring within the agricultural sector and employment shifts between farm and off-farm activities (Huang 2018a).This definition is straightforward to examine in research.Other researchers consider even more dimensions.For instance, Berdeguéet al.(2013)point out that rural transformation is about rural societies changing, not disappearing.It involves reorganising the rural community while moving from dispersed villages to towns and cities, leading to economic and cultural agglomerations, infrastructure improvements, and greater accessibility to services.Jayneet al.(2018) consider the interactions between rural and urban spaces, agriculture’s role changes, and broadening investment opportunities beyond farming.Rural transformation involves “greater complexity” and includes “changes in the composition of economic life and social organisation”, according to Koppel and Zurick (1988).

The 2016 IFAD Rural Development Report (IFAD 2016)defines rural transformation as “involving rising agricultural productivity, increasing commercialisation and marketable surpluses, and diversification of production patterns and livelihoods.It also involves expanded decent off-farm employment and entrepreneurial opportunities, better rural coverage and access to services and infrastructure,greater access to relevant policy processes, and greater capacity to influence.All of this leads to broad-based rural (and wider) growth and better managed, more sustainable rural landscapes” (p.23).

In investigating rural transformation stages, Huang(2018b, 2020) and Huanget al.(2020b) review China’s rural transformation in the past four decades and identify that rural transformation may include four significant stages.In the first stage, agricultural production is primarily on staple food.In the second stage, agricultural diversification and commercialisation emerge.In the third stage, with the continuing increase in agricultural specialisation, rural non-farm employment grows,inducing urbanisation and mechanisation of farming.In the last stage, rural transformation produces highvalue, sustainable agriculture and integrated urbanrural development.Huang’s typology of the stages of rural transformation takes the views of production and considers the consequences of farming changes.He also found that China’s rural transformation accelerated during the past two decades.

Having reviewed the definitions of rural transformation and associated parts, we explore the keywords researchers used to describe rural transformation and summarise the defining framework when investigating the transformation-facilitated policies.Fig.2 presents the word cloud of rural transformation definitions.

Fig.2. The word cloud of rural transformation definitions in the literature.

3.Measurement

Quantifying rural transformation is a cornerstone for measuring progress, assessing the performance of rural policy, and conducting cross-regional comparisons.The mainstream method to measure rural transformation is building up an index or some indices to indicate its state.The index number can be continuous or discrete,depending on the indicator selection and variable availability of the data sources.Alternatively, interviews or surveys can collect the data from the secondary public dataset.

Cloke (1977) provides an index to measure the state of rurality.Cloke and Edwards (1986) replicate this method and extend it by updating some variables.Cloke’s work leads a branch of literature on measuring rural transformation, though focusing more on rurality than rural-urban differentials.Cloke’s index is a linear combination of sixteen demographic variables that come from the Department of the Environment at England and Wales district-level census data.The index construction is based on a principal components analysis, which has become the major method for rurality measurement(Nelsonet al.2021).The following studies extend the Cloke’s index in two directions.The first attempt includes more variables to indicate rurality than demographics census; the other attempt includes more dimensions(index) to systematically multi-index assessment(Harrington and O’Donoghue 1998).The Cloke’s index is a way to observe “the relative position of the rurality of individual area at one point in time in relation to other localities” (Blundenet al.1998).However, the nature of rurality changes over time, say, the districts’ administrative boundaries change, or the census definition and terms change, but the Cloke’s index is focussed on providing a snapshot of rural transformation at one point in time in order to conduct cross-sectional comparisons, rather than recording the changing pattern over time.Cloke’s method also has to be regularly updated to account for the influence of new measures (Beynonet al.2016).

Wanget al.(2013) provide a method for an inter-temporal comparison of rural transformation.Mathematically,it denotes the state of rural transformation by a vector containing all measuring indicators at every point in time.Then, it demonstrates the change of rural transformation by computing the orthogonality relations between every pair of vectors.If the angular distance of the two vectors is zero, there is no structural change at all.If the angular distance of the two vectors equals one, there is a ‘complete’structural change.The contribution of this paper is mainly in computing procedures, and while it is insightful in that respect, it does not provide principles for the assessment of rural transformation.It still largely relies on the consistency of the variable selection and census data.Moreover, as the outcome of the computation is a pure number that is unit-free, it would not provide any feasible policy implication beyond a pure “degree” comparison.

Other researchers attempt to include more dimensions so that the rural transformation index can be seen as a composition of many layers of a rural-urban system.A notable example is the use of the “Population-Land-Industry” (PLI) index to measure rural transformation by building up a three-dimensional system (Kim and Yang 2016; Liet al.2017, 2018; Yanget al.2018).This method, however, still falls within the category of the single index approach, but it attempts to add more variables.It is worth mentioning that the papers that directly add variables to the extended version of the Cloke’s index make the computation procedure more data-driven.However, the PLI approach is built from theories instead of data; thus, it links the rural transformation index with broader discussions.

The PLI approach includes two steps of analysis.(1) In the first step, the degree of rural transformation is computed by three indicators: population, land,and industry.The underlying assumption is that rural transformation would result in population change,land-use change, and industrialisation; thus, we can use those three factors to measure the outcome of rural transformation.In mathematical terms, the rural transformation index is a linear combination of population,land, and industry factors.(2) In the second step,the rural transformation index derived from the first step is regressed against many influencing factors to identify which are related.This method encompasses perspectives beyond pure demographic change to link the rural transformation assessment with more policy issues like land transition or regional planning.The other merit of this three-dimensional approach is that, unlike the Cloke’s index, it provides a spatial and temporal analysis.Rural transformation patterns can be visualised over time,allowing for the analysis speed and shifts between stages.

In another direction, researchers have developed an indicator system containing more than one rural transformation index for each dimension.In this approach, more than one indicator can also demonstrate each dimension.For example, Longet al.(2011) classify rural transformation into three levels - rural development,transformation, and urban-rural coordination.At each level, they select several indicators.The weights of indicators are determined by semi-structured interviews with experts and government officials.The main limitation of this method is that the dimension (levels) selection is flexible and lacks a common framework.For example,Longet al.(2009) classify rural transformation into a farming industry-dominated rural development type(FIT), an industry-dominated rural development type(IDT), a rural development type focusing on business,tourism, and services industries (BTT), and balanced rural development type (BDT).An Indian case study adopts the same analytical framework (Ohlan 2016).

The tradition of agricultural economics provides another perspective.Huang and Shi (2021) provide a measurement using the share of high-value agriculture and the share of rural labour employment in non-farm sectors; that is, they capture the two major elements of rural transformation in economics - agricultural transformation and non-farm employment change.However, it may be too limited and omits other aspects of rural transformation.

Another difference among these various papers emerges from the discussion of the results of calculations after the formation of the indices, particularly those on how to define the “degree” of rural transformation in crossregional comparisons.For instance, Longet al.(2009)use the sum of the indicator’s sample mean and standard deviation as a benchmark to evaluate a particular indicator, while Huang and Shi (2021) use an indicator’s average value as a benchmark.

The methods noted above are all quantitative-based approaches.However, they would not be feasible in all rural areas, especially in developing countries,where the data are not usually available to measure quantitatively.In this case, Pudiantiet al.(2019) provide a qualitative approach to measure the rural transformation of Yogyakarta’s rural village in Indonesia.They conduct interviews and analyse the data using the quadrant model to classify and analyse the rural transformation levels.

4.lndicators

We distinguish here between the dimensions of rural transformation just discussed and the indicators relevant to those dimensions.The indicators can be directly retrieved from public census data or fieldwork and survey data.In some circumstances, researchers may also build up the indicators to reflect some aspects of rural transformation, such as productivity, diversification,or commercialisation, given that no data can directly indicate the situation.The indicators that demonstrate the consequence of rural transformation are, such as poverty reduction, income growth, and off-farm employment.These indicators usually cannot be directly affected by public policy, and they must be empirically investigable and measurable for research.Other indicators, such as investment, market integration, price, or terms of trade,are driving forces of rural transformation and are primarily influenced by policies and economic activities.Reviewing all the indicators in the literature is unnecessary, and that is indeed impossible.Instead, we summarise thinking on constructing key indicators.

Labour productivityQuantifying labour productivity is usually the first step in investigating the transforming practice, and the indicator is essential for policy decisionmaking in many cases.One common method in the economics literature is to use the agricultural value-added per worker (IFAD 2016; Huang and Shi 2021).Another indicator is to measure the difference in productivity between the rural agriculture sector and the urban nonagricultural sector by using the added values.This gap reveals the disparities between rural and urban income(Schwoobet al.2019).The accuracy of the value-added approach is based on the monetary values of products in a circumstance without price distortions.In some historical literature, when price data are unavailable,historians measure how many non-agricultural persons can be sustained by an agricultural person to represent agricultural labour productivity (Peng 2004).

CommercialisationOchienget al.(2020) use the proportion of the value of agricultural products sold to the market to measure the degree of commercialisation.A percentage that is over 50% is ‘commercially-oriented’.

DiversificationThe Herfindahl-Hirschman Index (HHI)is commonly used to quantify agricultural diversification(Pope and Prescott 1980; Liet al.2016), which is positively associated with rural transformation.It is defined as

whereris for the region,tfor time,ifor agricultural product.Ais the value of the relevant product.The value of diversification indexDranges from 0 to 1; larger values denote a higher degree of agricultural diversification.

lnclusivenessManagi and Kumar (2018) propose an Inclusive Wealth Index (IWI) to consider growth inclusiveness.IWI adds natural capital and human capital into the accounting framework in addition to physical capital.Natural capital includes mineral resources and energy, forests, agricultural land, and fishery, and human capital refers to education and health conditions.It is reasonable to use IWI’s measure of the change in wealth to assess the sustainability of growth.IFAD (2016)also adopts this framework to assess inclusive rural transformation.However, IWI alone cannot reflect all the aspects of rural transformation, such as gender equality.

SustainabilityThe sustainability concepts in the rural transformation bring about the interconnectedness of economic, societal, and ecological changes.Although there is no unambiguous approach to define the measures of aspects of sustainability, we can still learn some ways of thinking about this from the literature.One perspective is the community capital-based approach, which measures the sustainability from the ecological indicators,including the ecosystems, soil, groundwater, surface water, air, minerals, and landscape (Loweryet al.2020).the UN Agenda 21 provides many tools as sustainability indicators (Ramos 2019).Although the scope broadly includes urban and industry, most indicators are still applicable for rural transformation.On the side of governance, expert-driven tools are used mainly by policy practitioners and researchers based on different cases(Ferrariniet al.2001; Bell and Morse 2008).

We summarise other indicators in the literature and list examples in Table 1.

Table 1 Indicators adopted by literature1)

5.Results and discussion

5.1.A more comprehensive indicator system is needed

Policy-oriented research relies on a comprehensive and precise indicator system to evaluate the rural transformation performance.Reviewing the current literature above, we found that researchers normally use a single indicator to monitor rural transformation.This approach can simplify the complexity to some extent.However, one indicator cannot assess more aspects of rural transformation than it can measure.Therefore, such kind of research usually delivers one-sided opinions that cannot be adopted by policymakers.Future research demands a more comprehensive indicator system by using a synthetical approach to evaluate all rounds of rural transformation.

The PLI approach attempts to establish an index that is comprised of several indicators.However, The PLI index method has several drawbacks.First, the variable selection of PLI dimensions is subjective and depends on the researchers’ understanding of the theory and data availability.For example, Yanget al.(2018) chose the population urbanisation rate to demonstrate the population dimension, while Liet al.(2018) chose the nonagricultural employment proportion of agricultural labour to demonstrate the population dimension.Results are difficult to compare across studies, even though they use the same framework.Second, the researcher chooses the weights of PLI factors subjectively, which reduces the robustness of the measurement when the three factors play different roles in different stages across countries.More importantly, the PLI measurement ignores the causality between factors in the transformation.Neither PLI dimensions nor the second step’s influencing factors can tell policymakers the driving forces or effects.The papers on this method also omit some significant indicators relating to economic growth, poverty reduction,and social advances, etc.

5.2.Rural industrialisation and rural-urban integration

An example from China illustrates the point: Offfarm employment has been increasing rapidly, and the functions of labour markets are consistent with an economy transitioning from an agrarian to an industrial economy (Zhanget al.2018).Findlayet al.(1994)find that China’s rural enterprises were the fastestgrowing sector between 1978 and to early 1990s, which created significant changes in the structure of output and employment.Kwanet al.(2018) develop a threesector model of the rural agricultural sector, rural nonagricultural sector, and urban non-agricultural sectors to study China since the 1980s.They find that the labour mobility from agriculture to the rural non-agricultural sector contributes more than that to the urban sector in terms of total factor productivity growth.It indicates the significance of rural industrialisation with the accumulation of physical and human capital in structural change.You and Sarantis (2013) use the time series model to test China’s data for the period from 1980 to 2010.They confirm that the rural transformation is one of the major engines of China’s economic growth and significantly contributes to changes in total factor productivity across all regimes.On the other hand,they find that rural transformation’s contribution to total factor productivity growth declined while that of technical progress increased.They attribute this to the process of growth shifting to a more technology-driven model so that continuous investment in education and research and development would be essential for the next stage of development.

The findings of the impact of urbanisation on poverty reduction are mixed.Using cross-country panel datasets,Imaiet al.(2017) point out that urbanisation does not essentially lead to poverty reduction.Instead, in a few cases, population growth in megacities would increase poverty rather than decrease it.The development of the rural non-agricultural sector would reduce poverty in some cases, but the magnitude is generally much smaller than that of the rural agricultural sector.Therefore, they emphasise the value of policy support for rural agricultural and nonagricultural sectors, such as favouring rural infrastructure investment policies to reduce transaction costs.

5.3.The drivers of rural transformation: lnstitution,policy, and investment (lPls)

Huang and Shi (2021) demonstrate the effectiveness of institutions, policy, and investment (IPIs) in promoting rural transformation.(1) Institutions refer here to the laws and regulations governing the use of land, labour, and capital in agriculture and the processes by which those laws and regulations are designed; (2) Policy settings (e.g., tax,subsidies, marketing policies, trade policies, and prices of agricultural inputs and outputs) directly affect the stage and speed of the transition through the incentives they create and their implications for globalisation, for instance;(3) A range of investments is likely to be relevant, ranging from technology, education, and infrastructure.

Although some researchers provide several empirical studies on the effect of IPIs, they fail to examine the impact of IPIs on the stage, speed, and outcome of rural transformation and neglect the interactive mechanism between them.Policymakers and practitioners need to understand it sufficiently to respond to the changes taking place in rural areas.

5.4.The inclusiveness of rural transformation

The current research focuses more on the gender balance issue.Rosenzweig (1993), based on Indian evidence,finds that rural transformation does not necessarily lead to great equality by sex in the intra-household distribution of resources despite the apparent normality of equality in intra-household resources.Donget al.(2004) conclude that China’s rural transformation is not gender-neutral but biased toward men, particularly in the newly private township and village enterprises.Mehrotra and Parida(2017) use micro and macro data to find a U-shaped curve of India’s female labour force participation rate.Rural transformation pushed many females out of agriculture,but they are unlikely to be employed in manufacturing sectors because of their low education and skills due to other cultural constraints.Trapagaet al.(2019) conclude that accompaniment, empathy, and training must create social micro-enterprises that foster transformation and social value and contribute to local development.

Heckertet al.(2020) conduct a cross-nation analysis of 36 low- and middle-income countries between 2010 and 2016.They find that higher levels of rural transformation are associated with a higher likelihood of land ownership for young men rather than young women.Rural transformation is associated with more chances to be employed for young rural men rather than women.

Those studies show gender imbalance in rural transformation, commonly occurring in many developing countries, particularly some Asian countries.However,to what extent the loss of gender diversity (gender imbalance) influences rural transformation in productivity,efficiency, speed, stage, and welfare is still unanswered.More importantly, does gender diversity lead to a smoother transformation at the community level,enhancing the overall performance of rural governance?In policymaking, what elements of inclusiveness can be measured and considered is another blank area that needs to be filled?

5.5.Market creation and missing services of the value chain

In a Turkish case study, Keyder and Yenal (2011) found that deepening the commodification of agriculture had significantly enhanced farmers’ living standards.It also brought about more types of agribusinesses in the countryside, such as tourism and options for seasonal employment.However, market fluctuations could reduce the welfare of smaller producers as they face higher risk and insecurity.

On the other hand, there is a less optimistic case from the Philippines where free-market and trade reform outcomes differed from those of the other countries in Southeast Asia, though they carried out the reform much earlier.Borras (2007) found that the free market did not lead to a structural change -reducing poverty or inequality - in the Philippines.Borras’ assessment is that the creation of market reform reinforced the existing structure dominated by the elites.

5.6.Rural transformation and international trade

Anderson and Martin (2009) also find that in China and Southeast Asian countries, the levels of non-agricultural protection significantly declined from 1980 to 2004,increasing the agricultural sector’s competitiveness.

Furthermore, withdrawing resources towards the comparative advantage of agriculture, Bustoset al.(2020)provide a model to show how productivity advances can generate an expansion of the capital-intensive sector,such as manufacturing, by increasing savings and,thereby, the supply of capital for engineering development.In this case, additional capital would be reallocated to the urban sectors, and the financial flows would influence the speed of such transformation.Garnaut and Anderson(1980) summarise the structural change patterns of the northeast and southeast countries.They concluded that their development models all follow the processes of dynamic comparative advantage, that is, from the natural resource-intensive sector to the labour-intensive sector and then to the capital-intensive sector.

On the other hand, Gelan (2002), using a rural-urban computable general equilibrium (CGE) model, points out that the effectiveness of international free trade in promoting a rural transformation depends critically on the synchronism of the trade reform and the domestic factor market reform.If the urban wage is fixed (i.e.,the labour market reform lags), trade reform adversely affects economic growth due to the contraction of urban regions.On the contrary, rural and urban regions expand in GDP if urban wages become flexible.This research is a reminder of the importance of an appropriate sequence of policy instruments.

Strojny (2019), using a vector autoregression model,shows that in the Baltic States’ agricultural transformation,the terms of trade are the most exogenous factor, and the gross value added is the most endogenous variable.However, whether this conclusion is applicable to other larger agrarian economies with significant transitioning policies, say, some Asian countries like China, is still to be investigated.

5.7.Rural transformation and sustainability

The sustainability of rural transformation comes to the fore in many developing countries under global efforts to mitigate greenhouse gas emissions.It distinguishes between today’s development trajectory and the Western world’s experience.The rapid structural change in rural areas may cause a tremendous environmental impact on the one hand.On the other hand, adopting new energy technology and greening business models may accelerate rural transformation, as the new technology is generally skills and capital-biased.The governments may use some sustainability-favouring policies to achieve an environmental-friendly rural transformation.For example,from 2013 to 2016, China used a photovoltaic poverty alleviation policy to increase rural people’s per capita disposable income by around 7-8% while undertaking the massive deployment of solar farms in rural areas (Zhanget al.2020).Gong (2020) finds that rural households change their energy usage as their income increases in China.A subsidy policy on clean heating energy can significantly induce the adoption of electric heating technology.

D’Alessandro (2011) shows that environmental shocks can increase peasants’ vulnerability to dropping wages and even to famine in the rural transformation.Lettaet al.(2018) find that climate change (temperature shocks) will slow the income convergence between the more affluent and poorer rural households in Tanzania.Mulwa and Visser (2020) point out that farms can increase resilience against climate change by diversifying production in northern Namibia.On the other hand,the farmers’ ability to adapt to climatic hazards such as droughts, floods, and severe storms may vary according to their location and socioeconomic status.Huanget al.(2020a) find that the older and poorer farmers with smaller land and living in mountainous areas are more vulnerable to climate change.

Some researchers focus on the impact of green energy development on rural transformation.Leknes and Modalsli (2020) provide a Norway case at the turn of the 20th century (1891-1920).They find that hydropower municipalities experienced faster structural transformation and displayed higher occupational mobility.Tagliapietraet al.(2020) find that promoting rural households’ access to electricity can increase overall labour participation.Moreover, it decreases agricultural employment by around 7% but increases non-agricultural employment by about 15%.Thus,electrification helped transform the Nigerian economy away from agricultural activities.

Barbier (2020) shows that the renewable energy technology diffusion and adoption in rural regions may be influenced by some barriers, such as the lack of longterm financing schemes, poor private sector participation,inadequate institutional structure, lack of coordination between local and national governments, and the weak purchasing power of rural communities.Therefore, this situation calls for policy facilitation to remove the market frictions.

OECD (2020) points out that most of the potential synergies on this linkage are underutilised, so they call for building institutional capacity and developing policy instruments to manage the environmental and energy transition in rural areas.Although we have learned from the existing evidence, it still leaves a large research space to explore.Future research may focus on the reciprocal relationship between rural transformation and climate change, providing a more comprehensive analytical framework, a valuable set of policy tools, and a more robust causality explanation in principle.

6.Conclusion

The rural transformation includes but is not limited to the agricultural transformation.It covers all the changes in the agrifood system - from production to consumptio.- and includes changes in rural society, landscape,and governance.Rural transformation is the pathway to development and prosperity in the countryside.For many developing countries, it is also paramount to realise industrialisation, now broadly defined as manufacturing and services.This research subject has attracted many economists, sociologists, and historians to contribute experience and knowledge.However, achieving the success of rural transformation needs a further in-depth understanding of the nature and consequences of the transforming process.

This paper reviews the rural transformation literature concerning its definition and measurement, the latter concerning its dimensions and indicators.We differentiate the scope and boundaries of rural transformation and two other related concepts - agricultural transformation and structural transformation.We conclude that the rural transformation is a part of the structural transformation of the entire economy, but it captures all aspects of agricultural transformation extending beyond the agrifood system.Although the studies from different perspectives provide different definitions of rural transformation, we can draw on the common elements of the definitions.That is, the rural transformation, we find, should consider the changes in productivity, rurality, inclusiveness, and sustainability.It changes the way of production, the landscape of rural areas, and the functions of rural space,farming activities, and the agricultural sector.Through the dynamic co-evolution with other sectors of the economy,an inclusive rural transformation will facilitate the sustainable and inclusive growth of developing countries.

The current efforts to measure rural transformation establish numerical indices using factor analysis.Although they have shed light on the assessment, the current measurements are still insufficient for policy decision-making.A more scientific measurement requires objectivity, feasibility, accountability, comprehensiveness,and comparability.It should rely on data to reveal the pattern and connect the theories to reveal the causeeffect mechanisms.The measurement methods should provide more intellectual insights and evidence on making a good policy and not be confined to the data.A reflection on the current measures would be essential in developing a new method.

We review the indicators of various dimensions of rural transformation by highlighting several key ones.We suggest that more indicators be explored for future studies.The future research agenda would include two main tasks: establishing a database to measure rural transformation and advancing the methods to construct more complicated indicators such as inclusiveness or sustainability.We further call for future research to pay particular attention to the synthetical indicator system approach, IPI-related research, gender inclusiveness issues, the missing service of the value chain, and the role of international trade.

Acknowledgements

This research is funded by the Australian Centre for International Agricultural Research (ACIAR; ADP-2017-024).

Declaration of competing interest

The authors declare that they have no conflict of interest.