Introduction

Over the last half century, the world has witnessed some dramatic shifts in food consumption patterns (Kearney 2010). Diets have evolved, shifted and continue to change for a variety of reasons (Msangi and Rosegrant 2011). Rising incomes, increases in food production, demographic expansion accompanied by migration from the countryside to urban areas, as well as new communication, transportation, food production and cooking technology, female participation in the workforce and a renewed focus on health and nutrition are but a few of the many factors driving these developments (Wilkinson and Rocha 2009; Kumar 2017).

Changes in food consumption patterns have perhaps been even more noteworthy in developing countries, particularly over the last 25 years and particularly in Asia (Timmer et al. 2010; Reardon et al. 2012; Scott and Suarez 2012a; Timmer 2014; Kumar and Joshi 2016a; Vengatesh et al. 2016). With accelerated economic growth, continued albeit declining population expansion and growing urbanization, shifts in eating habits have become more pronounced (Pingali 2006; Radhakrishna 2006; Scott and Suarez 2012b). With rising awareness of environmental degradation and the potential impact of the advent of climate change, concerns over the prospects for food consumption in Asian food systems have become that much more common (Timmer 2017).

In that context, efforts have focused on modelling future demand for those food crops that have long formed the basis of the global diet—cereals such as rice, wheat and maize—particularly in Asia (Alexandratos and Bruinsma 2012; FAO 2016; OECD/FAO 2016). However, the potential contribution of other food commodities also merits systematic analysis. This paper focuses on future scenarios for potato consumption and utilization in India to 2030.

Our interest is in India because with over 1.2 billion people, it will become the world’s most populous country before 2030 (United Nations 2015; IndiaTimes 2017). Hence, future food consumption patterns in India merit particular attention at both the national and global level.

Our interest in potato in India reflects in part the remarkable rise in potato consumption over the last half century. Per capita consumption rose from a meagre 3.5 kg person−1 in 1962 to over 24 kg person−1 in 2013 (Fig. 1), with recent estimates of availability capita−1 at the state level well above (> 70 kg year−1 in West Bengal) and below (< 1 kg year−1 in Tamil Nadu) that national average (NCCD 2015).Footnote 1 Moreover, contrary to previously anticipated declines in consumption due to a rise in incomes (Walker et al. 1999), nearly half of that increase occurred since 2003 along with rising incomes (Trading Economics 2017; Rana and Anwer 2018). Factors driving these developments include the:

  • desire by consumers to diversify their eating habits beyond simply greater consumption of cereals as their incomes continued to increase;

  • neutral taste and gastronomic versatility of the tuber;

  • potato’s nutritional attributes as a source of vitamins and minerals, particularly when consumed with the skin intact (Woolfe 1987);

  • strong vegetarian tradition in Indian diets; and, the

  • relatively modest levels of per capita consumption of potato currently in relation to the levels prevailing in other developing countries in Asia (Scott and Suarez 2012a).

Fig. 1
figure 1

Evolution of annual production and per capita consumption of potato in India, 1961–2013. Source: FAO (2017)

Full size image

Partly in response to these emerging trends, potato consumption and utilization in the developing countries of Asia, including India, have been the focus of a small number of earlier studies (Anonymous 1995; CPRI 1997, 2011; Scott et al. 2000). However, the historical databases that many of these studies drew upon are now outdated with their time horizons already well in the past (2000) or within relatively close proximity (2020). In that regard, it is noteworthy that the most optimistic estimate of total demand for potato in India in 2020—43.3 million tonnes—had already been surpassed reaching over 50 million tonnes in 2011 (Scott et al. 2000, FAO 2017). Moreover, FAO’s most recent published estimates of future demand for potato in developing countries in 2050 do not include country-specific figures for potato in India (Alexandratos and Bruinsma 2012). Furthermore, the small number of recent studies that have generated estimates of future scenarios for potato in India has been focused only on potato and only on yields (Jaggard et al. 2010; Kumar et al. 2011, 2015; Dua et al. 2015) or provided only limited information on the methodology utilized or assumptions employed to arrive at the array of estimates of consumption and utilization presented (CPRI 2011, 2015). Other studies of future food demand in India have focused largely on cereals, pulses and other commodities, but not potato (Kumar and Joshi 2016b).

In that context, this paper presents estimates of potato demand in India to the year 2030 according to high, moderate and low demand growth scenarios. In that regard, the following section presents the basic model utilized to generate the aforementioned estimates including the specification of the economic, demographic, climatic, technological, and policy-related parameters utilized in each of the scenarios. The paper then reviews the key results. This includes interpreting the growth rates in total food demand for potato, seed, and other uses that resulted in the absolute levels calculated for each of those variables in each of the scenarios and the corresponding estimates for per capita consumption and trade. The paper then addresses the implications of these estimates. These include policies to help facilitate their realization as well as specific measures to capitalize on the potential they represent in terms of improved marketing for growers and consumption and nutrition for consumers in India. It also reviews various opportunities for industry highlighting those with the greatest potential to combine commercial success with benefits for producers, consumers and the environment as well, before mentioning some topics for future research.

The IMPACT Model

Future scenarios for potato demand in India generated for this paper utilized the 2015 version of the International Model for Policy Analysis of Agricultural Commodities and Trade (IMPACT) (Rosegrant and the IMPACT Development Team 2012; Robinson et al. 2015). IMPACT is a multi-period, agricultural partial equilibrium economic model, linked with a set of crop, climate, and water models. The integrated foresight modelling framework built around IMPACT has been used extensively to examine various issues related to world agriculture, including commodity-specific supply and demand at various geographical scales over different time periods (e.g. Scott and Kleinwechter 2017) and also in analyses related to climate change impacts on agriculture and global food security (e.g. Nelson et al. 2010, 2018; Springmann et al. 2017).

This version of IMPACT includes 62 agricultural commodities and distinguishes 159 geopolitical regions. Total demand for a given commodity, potato in this case, is the sum of individual demand components, namely household food demand, feed demand and intermediate demand for other uses like seed or industrial use. For every year simulated by IMPACT, food demand QF for commodity i in country c is given by the following isoelastic function:

$$ {QF}_{i,c}={FInt}_{i,c}\times \Delta Pop\times {\left(\Delta GDP\right)}^{\varepsilon}\times {\left({\Delta P}_{i,c}\right)}^{\delta}\times \prod \limits_{j\ne i}{\left({\Delta P}_{j,c}\right)}^{\delta } $$
(1)

where FInt is the intercept (i.e. food demand at the initial year of the simulations, as reported in FAO (2017). The remaining components of the equation act as shifters from FInt:

  • ΔPop is population growth;

  • ΔGDP is the change in per capita income;

  • ε is the income elasticity of demand for commodity i;

  • ΔP is the price change of commodity i and any other competing commodity j;

  • δ is the price elasticity of demand;

  • (ΔPi, c)δ represents own-price response of demand and \( \prod \limits_{j\ne i}{\left({\Delta P}_{j,c}\right)}^{\delta } \) represents cross-price responses of demand.

The specification for intermediate demand QI is similar to Eq. (1), and also depends on population, per capita GDP and price effects. By way of contrast, demand for feed QL of commodity i in country c is determined by feed prices and the total feed requirements of the livestock sector in country c. It is also given by an isoelastic function, specified as follows:

$$ {QL}_{i,c}=\sum \limits_v\left({SL}_{v,c}\times {Req}_{i,v,c}\right)\prod \limits_K{\left({\Delta P}_{K,c}\right)}^{\xi } $$
(2)

where

  • v denotes the livestock activities.

  • SL is the production volume of livestock activity v.

  • Req is the i feeding requirement for livestock activity v.

  • K is the set of all commodities that can be used as feed, which includes i and competing commodities j.

  • ξ is the price elasticity of demand for each feed commodity in K.

Although the 2015 version of IMPACT includes estimates of consumption, utilization, trade, prices, and agricultural production up to 2050, the current study emphasizes the presentation and interpretation of calculations of future demand scenarios with a time horizon of 2010 to 2030. In that regard, the end point represents a year far enough into the future to justify alternative estimates about an uncertain outcome. It thereby allows for the on-going effects of income changes, population growth and climate change to manifest themselves while capitalizing on the growing body of literature analysing historical trends for potato consumption and utilization in India over the last five decades from which previous estimates of future potato demand did not have the benefit. Furthermore, by focusing on 2030, the results of this study leave time for plans to be made and implemented in relation to projected scenarios as well as to provide a benchmark to build on in future modelling efforts focused on subsequent decades. While calculating estimates for 2010–2030, we do so acknowledging that these calculations are part of a long-term, continuous process and, in the case of climate change, before much greater changes in climate patterns are considered likely to occur (Kumar et al. 2015). We also consider that a much longer period for projections raises the prospect of more known and unknown uncertainties (Scrieciu et al. 2013), with the associated estimates becoming more problematic to incorporate into the modelling exercise thereby undermining the credibility of the results (Rosen and Guenther 2015).

Key Components and Scenario-Specific Parameters

In this paper, we define three scenarios for potato household food demand by adjusting the income elasticity of demand ε in Eq. (1) according to different future change pathways that draw on research performed under the auspices of the Intergovernmental Panel on Climate Change (IPCC). Specifically, the scenarios are based on combinations of Shared Socioeconomic Pathways (SSPs) and Representative Concentration Pathways (RCPs), both of which have been developed by the IPCC for the purpose of carrying out foresight analysis related to climate change, while also accounting for possible socio-economic changes like population and economic growth.

SSPs represent alternative trends in socio-economic and natural systems and incorporate different assumptions related to challenges for mitigation and adaptation to climate change. These assumptions are expressed in terms of different future growth rates in income, population, and urbanization for large world regions (O’Neil et al. 2014). On the other hand, RCPs represent different climate change pathways, incorporating assumptions about radiative forcing (i.e. the positive change in Earth’s surface-troposphere energy budget that is responsible for climate change) from concentrations of greenhouse gases and other forcing agents (Van Vuuren et al. 2011). The radiative forcing assumptions posited by RCPs are translated into daily, or monthly, climate projections at various spatial resolutions with global climate models, most formally known as Earth System Models (ESMs).

Although this paper focuses on future food demand in India, the foresight analysis performed and the findings presented herein are part of a larger set of estimates that include food supply as well, under additional assumptions regarding agricultural technology and crop yields.Footnote 2 For the overall study, we combined three different SSPs with three different RCPs to link a set of socio-economic assumptions to a specific climate change pathways and associated assumptions regarding technological change in food production to those for food consumption and utilization. Following Petsakos et al. (2018) and Wiebe et al. (2015), the selected combinations were SSP1–RCP4.5, SSP2–RCP6.0, and SSP3–RCP8.5 (Table 1). The three SSPs correspond to pathways of high, medium and low mitigation and adaptation challenges respectively (O’Neil et al. 2014), while the number accompanying the RCPs expresses the corresponding maximum radiative forcing by 2100 (in watts m−2), with higher numbers indicating a more severe climate change pathway.

Table 1 Specification of simulation scenarios according to different socio-economic, climate and technology assumptions
Full size table

More specifically, SSP1–RCP4.5 is the most optimistic of the three scenarios, positing high economic growth and low population growth rates (Table 2) combined with less demanding climatic conditions and more favourable assumptions regarding the evolution of technological change (Table 1). The income elasticity of demand for potato is also assumed to be highest among the three scenarios.Footnote 3 As such for this study, the SSP1–RCP4.5 combination represents the high demand growth scenario.

Table 2 Gross domestic product (GDP) and population growth rates (% year−1) in India under different scenarios
Full size table

SPP2–RCP6.0 represents the moderate demand growth scenario. It assumes solid, albeit more moderate, demographic, and economic growth rates (Table 2) along with more challenging climatic conditions than SSP1–RCP4.5. Scenario SPP2–RCP6.0 is also linked to intermediate assumptions regarding the growth rate for technological change (Table 1).

The SSP3–RCP8.5 is the most pessimistic scenario with relatively high population growth rates, low economic growth, and more adverse climatic conditions (Tables 1 and 2). These assumptions, combined with a more conservative outlook for technological change and lower income elasticities of demand, constitute the low demand growth scenario.

The quantification of the different RCPs in each of the above scenarios in terms of daily climate projections included four ESMs integrated into IMPACT. These are the Model for Interdisciplinary Research on Climate (MIROC-ESM, Watanabe et al. 2011), ESM from the Pierre Simon Laplace Institute (IPSL-CM5A, Dufresne et al. 2013), the Geophysical Fluid Dynamic Laboratory ESM (GFDL-ESM2M, Dunne et al. 2012) and the Norwegian ESM (NorESM1-M, Bentsen et al. 2013). Each of these models incorporates different assumptions regarding the modelling of Earth’s climate and therefore interprets a given RCP accordingly. The daily climate projections produced by the previous four ESMs were then used as data inputs for the crop growth model DSSAT (Decision Support System for Agrotechnology Transfer) (Jones et al. 2003), incorporated into the IMPACT modelling suite, in order to estimate the impact of climate on yields of potato and other crops, including rice and wheat.Footnote 4

The SSP–RCP combinations selected for this study are consistent foresight scenarios for climate change research that assume similar mitigation costs (Van Vuuren et al. 2014). By also using an ensemble of ESMs to interpret each RCP, rather than a single model, the three scenarios, along with the additional supply and demand assumptions they incorporate, are able to capture a wide range of possible future changes in socio-economic and natural systems and can therefore provide a better description of India’s food systems in the future.

All three scenarios assume the continuation of a set of benign policies toward the potato sector including:

  • an absence of recurrent price controls or obligatory public procurement programmes (Rahman et al. 2019) for potato that might otherwise dampen producer/trader enthusiasm for greater production and marketing (Fuglie et al. 2000; Scott 2002; Reardon et al. 2012);

  • support for the massive expansion in cold storage infrastructure to meet the capacity requirements associated with run-up in the volumes of harvested potatoes for sale as food or use as planting material (Dahiya and Sharma 1994; NCCD 2015; Anonymous 2017); and,

  • investment in potato research and development to foster technological innovation in potato production and utilization (CPRI 2011, 2015).

Results

Moderate Demand Growth Scenario

According to the moderate demand growth scenario, total demand for potato in India will grow 3.11% year−1 on average during the period 2010–2030 (Table 3). While this annual average compound growth rate (ACR) may appear high, it actually represents a relatively conservative outlook for the future growth of potato demand. Actual total potato demand in India grew at nearly 7% year−1 during the decade 2001–2003 to 2011–2013 (Fig. 2). CPRI (2015) anticipates a much more ambitious ACR of 6% year−1 for potato demand up to 2050.

Table 3 Average annual growth rates (% year−1) for food, other uses, and total demand for potato in India 2010–2030, according to different scenarios
Full size table
Fig. 2
figure 2

Total demand for potato in India and ACRs for total demand in India, 1961–2013.a Source: FAO (2017) and calculations for this study. aWhere each square represents the average compound growth rate (ACR) over a moving 10-year period with each of the points during each of the 10 years being a 3-year average, e.g. 1971–1973 vs 1961–1963, 1972–1974 vs 1962–1964 and where the growth rate is calculated using the following expression: Y = b0eb1t, i.e. ln(Y) = ln(b0) + b1t; where Y = total demand; ln = natural log and ACR = (antilog (b1) − 1) × 100 and where each diamond represents total demand for potato (000,000 t) in India

Full size image

The moderate demand growth scenario consequently estimates that total potato demand in India will nearly double to reach 68.9 million tonnes by 2030 (Table 4). This increase reflects a combination of:

  • India’s massive population touching 1.5 billion in 2030 (IndiaTimes 2017);

  • growing urbanization (Sankhe et al. 2010)—an estimated 40% of India’s population or nearly 600 million people will be living in urban areas by 2030—and with that an increased demand for food away from home including restaurants as well as in the form of snacks and convenience foods (Pingali 2006; Scott and Suarez 2011; CPRI 2015);

  • the relatively modest levels of per capita consumption of potato (22 kg year−1) prevailing in the country at the outset of the 2010–2030 period (FAO 2017); and,

  • the relatively strong income elasticity of demand for potato—in particular in relation to staple foods such as rice or wheat (Appendix Table 11).

Table 4 Total demand (000 t) for potato as food and for other uses in India in 2010 and estimates to 2030, according to different scenarios
Full size table

The moderate demand growth scenario further estimates the ACR for food demand will average 2.84% year−1 and that of demand for “other uses” of potato to reach 3.81% year−1 (Table 3). The ACR for food demand includes both fresh and processed potato products (e.g. potato crisps, flour) for direct human consumption. It constitutes the growth rate for over 70% of total demand when comparing absolute totals (Table 4) and reflects the potato’s role in the average diet in India as most commonly that of a complementary vegetable with at least three noteworthy exceptions.

Firstly, at the peak of harvest in the major potato-producing states in the Indo-Gangetic plain, potatoes can become a seasonally more attractive food for all consumers. This is particularly true for the very poor and/or migrant labourers when as a flood of supplies appear, their meagre incomes preclude purchase of other food crops that are less readily available (Scott and Suarez 2011). Production figures in relation to cold storage capacity and utilization (see below) suggest that a disproportionate share—roughly 30%—of annual consumption takes place during and immediately after the harvest months that vary by state but tend to be concentrated in February and March (NCCD 2015). In a similar vein, potato can also serve as a substitute for other foods, i.e. cereals during fasts for religious festivals.

Secondly, early potatoes in India—harvested in November and December in selected locations (e.g. the Punjab)—also receive premium prices for their superior taste and the social status their scarcity value conveys. These tubers are more like a high-priced, seasonal vegetable than a relatively cheap complementary or staple food. In parts of India, for example Meghalaya and West Bengal, the so-called native potatoes represent varieties introduced long ago and through negative selection produce small tubers that are associated with particular gastronomic or storage traits (CIP 2009). These tubers can also fetch a higher price for their culinary appeal and exotic commercial status.

Thirdly, French fries in large metropolitan areas such as Mumbai or New Delhi are a relatively expensive food item eaten by the more affluent, younger generation for their taste and convenience as well as their association with a certain Western life style. Such a consumption pattern for potato is commonly found in other large, cosmopolitan urban areas in other parts of developing Asia (Guenthner 2001, 2010; Watson 2006; Curtis et al. 2007; Scott and Suarez 2012b).

These estimates for future consumption are also grounded in the more general trends in eating habits in India previously enumerated and their relevance for potato in particular including the:

  • strong vegetarian tradition in Indian diets (Reardon and Minten 2011);

  • as of yet very minor (< 5%) percentage of potato consumption derived from processed food products, e.g. chips and French fries (CPRI 2015);

  • rapid spread of fast foods elsewhere in developing Asia that has unrealized potential in India (Pingali 2006; Scott and Suarez 2012b); and,

  • prospects for improvements in post-production value chain activities thereby making potato more competitive in terms of cost and quality control (Reardon et al. 2012).

The growth rate estimated for “other uses” as understood in this study is a composite figure covering, according to FAO, demand for (1) seed and (2) “waste”, e.g. weight loss in cold storage or transit from farm to consumer and spoilage developing during packing and handling. No potatoes serve for animal feed in India as in other Asian countries (Scott and Suarez 2012a; FAO 2017) nor are there statistics indicating any processing of potatoes for industrial use. Given these considerations, the estimated growth rate for other uses according to the baseline scenario would appear to be particularly high for several reasons.

In India, the quantity in terms of weight of tuber seed or its equivalent used per hectare has declined over time as cut seed became a common practice and alternative types of planting material emerged, in particular mini-tubers (Scott and Suarez 2011; Fig. 3). Consequently, seed use as a share of total use declined from 17% in the early 1960s to 7% in 2011–2013 (Table 5).

Fig. 3
figure 3

Total seed use for potato in India and ACRs for total seed use in India, 1961–2013.a Source: FAO (2017) and calculations for this study. aWhere each square represents the average compound growth rate (ACR) over a moving 10-year period with each of the points during each of the 10 years being a 3-year average, e.g. 1971–1973 vs 1961–1963, 1972–1974 vs 1962–1964 and where the growth rate is calculated using the following expression: Y = b0eb1t, i.e. ln(Y) = ln(b0) + b1t; where Y = seed; ln = natural log and ACR = (antilog (b1) − 1) × 100 and where each diamond represents total seed potato demand (000,000 t) in India

Full size image
Table 5 Food balance sheets for potato in India, 1961–2013
Full size table

The opposite has occurred with losses or waste. Their share of total demand rose from 17% in the early 1960s to some 24% in 2013 (Table 5) with both percentages having little, if any, empirical justification. The sharp rise in the absolute totals calculated as waste since 2007 (Fig. 4) seems particularly suspect. Earlier studies that attempted to quantify losses from waste tended to focus on storage with less attention given to wholesaling activities forward through to final human consumption. They estimated that losses during cold storage—where most potatoes are kept after harvest—were 5% of the volumes stored (Fuglie et al. 2000). Moreover, recent field research on value chain transactions aside from storage involving potato growers, cold store operators, wholesalers, retailers and consumers found that losses were only 1% (Reardon et al. 2012). In addition, estimates of waste and losses of 24% of total available domestic supply are much higher than comparable percentages for other developing countries in Asia—estimates for China are 6% according to FAO data (Scott and Suarez 2012b). The weight of the evidence therefore suggests that the estimated magnitude of losses from other uses is exaggerated. Rather, it would appear that 10–15% or more of total demand currently assigned to the losses sub-category are and will continue to be devoted to direct human consumption instead.

Fig. 4
figure 4

Total losses for potato in India and ACRs for total losses in India, 1961–2013.a Source: FAO (2017) and calculations for this study. aWhere each square represents the average compound growth rate (ACR) over a moving 10-year period with each of the points during each of the 10 years being a 3-year average, e.g. 1971–1973 vs 1961–1963, 1972–1974 vs 1962–1964 and where the growth rate is calculated using the following expression: Y = b0eb1t, i.e. ln(Y) = ln(b0) + b1t; where Y = waste; ln = natural log and ACR = (antilog (b1) − 1) × 100 and where each diamond represents total potato waste (000,000 t) in India

Full size image

According to the moderate demand growth scenario, per capita potato consumption in India will reach 31.4 kg year−1 in 2030 (Table 6). While this represents a 50% increase in average per capita in-take during the two decades, recent estimates point to the potential for greater consumption. Examples include those locations where per capita availability remains relatively low even with sharp increases in output (e.g. 27.6 kg year−1 in Gujarat) or where improvements in transportation and the cold storage chain can facilitate consumption increases despite relatively low production levels (NCCD 2015). In broader terms, these estimates of an increase in per capita consumption represent simply the most recent manifestation of the growing importance of potato in the diet in India that emerged over the last half century (Table 7) due to a variety of factors (Scott and Suarez 2011). These include the:

  • desire by consumers to diversify their eating habits beyond simply greater consumption of cereals as their incomes increase.

  • neutral taste and gastronomic versatility of the tuber,

  • tuber’s nutritional attributes as a source of vitamins and minerals, particularly when consumed with the skins intact (Woolfe 1987),

  • massive expansion of cold storage facilities to over 27 million tonnes that enables 70% of yearly consumption to take place after the annual harvest largely concentrated during February and March (Anonymous 2017), and,

  • persistence of vegetarian diets among Indian households as a gastronomic tradition grounded in national culture (Reardon and Minten 2011).

Table 6 Annual average food demand (kg capita−1) for potato, wheat and rice in India in 2010 and estimates to 2030, according to different scenarios
Full size table
Table 7 Annual average consumption of select food commodities in India, 1961–2013
Full size table

In the moderate demand growth scenario, net trade is estimated to be a meagre 181,000 t—actually slightly less than observed potato exports in 2010 (FAO 2017)—but with considerable dispersion above and below the mean value (Table 8). As such, this estimate for net trade is in line with a half century of statistics showing that for all years but one, exports remained less than 300,000 t year−1 even as production grew remarkably (Dahiya and Sharma 1994; FAO).

Table 8 Potato net trade (000 t) in India in 2010 and estimates to 2030
Full size table

More fundamentally, these figures reflect a series of underlying factors discouraging exports of potatoes and potato products. These include (i) uncertainty about available supplies and export prices in relation to ever-growing domestic demand; (ii) limited infrastructure in the prevailing export cold chain to handle a bulky, semi-perishable commodity like fresh potato (NCCD 2015); and (iii) competition from established exporters in Europe or North America that ship to the same potential markets in the Gulf States and Southeast Asia (Scott and Suarez 2011). As a result, the moderate demand growth scenario suggests that in the continuous interplay of supply and demand, any movement toward a significant deviation from the estimated net trade figure would most likely be dampened by countervailing forces conducive to the domestic market, absorbing what otherwise might have been exported. In other words, “exports” would take place from those states in India with higher per capita availability to markets in those neighbouring states with lower per capita supply (NCCD 2015).

High Demand Growth Scenario

Based on a more rapid growth scenario, as represented by the SSP1–RCP4.5 combination and the related technology and demand assumptions, total demand for potato in India grows at a faster rate—3.54% year−1 versus 3.11% year−1—than in the moderate demand growth scenario (Table 3). The associated absolute demand estimates include 9% higher total demand, 75 million tonnes, driven by a sharply higher ACR for food demand—3.38% year−1 versus 2.84% year−1, resulting in a 17% higher per capita consumption of potato, 36.7 kg year−1 versus 31.4 kg year−1, than that estimated in the moderate demand growth scenario (Tables 3, 4 and 6).

This set of demand estimates partly reflects greater optimism about the prospects for future consumption of processed potato products made possible by, among other things, greater female participation in the workforce and the associated demand for easier to prepare meals and snacks. It would also include more potato consumption in processed form away from home (e.g. schools, colleges) and eating out (restaurants, quick service establishments). As India becomes a more urban, less predominantly rural country in the years ahead (Sankhe et al. 2010), the high demand growth scenario envisions that such non-traditional eating habits may well spur added potato consumption more than in the case of the moderate demand growth scenario. In effect, it anticipates that faster economic growth (Tables 1 and 2) will facilitate a stronger preference for greater convenience associated with newly emerging eating habits and continued diversification of diets away from the cereals (Table 7).

As a response to the more rapid growth in population and economic output, the high demand growth scenario envisions an increase in imports—1.3 million tonnes (Table 8), or < 2% of total demand, to meet domestic potato utilization requirements. However, as the quantity of imports is a relatively meagre percentage of domestic demand, it may ultimately be covered by squeezing more tubers for direct human consumption out of the commodity value chain for potato. In effect, as higher prices resulting from stronger demand will contribute to such things as reducing waste from what appears to be an unrealistic absolute total (Table 4) for the reasons previously mentioned, the overall impact may well leave little need for imports.

Low Demand Growth Scenario

The low demand growth scenario is based on a more pessimistic set of assumptions regarding economic and population growth (Table 2), compounded by more adverse changes in climatic conditions, expressed by the SSP3–RCP8.5 combination (Table 1). It also posits slower technology growth and lower income elasticities of demand for potato (Appendix Table 11).

Slower economic growth helps cut the ACR for food demand for potato to 2.36% year−1 versus 2.84% year−1 in the moderate demand growth scenario (Table 3). The lower ACR results in lower per capita consumption of potato—27.1 kg year−1 versus 31.4 kg year−1 in the moderate demand growth scenario—as consumers cut back their in-take of starchy staples such as rice and wheat by a lesser amount (Table 6).

Total demand for potato consequently reaches an estimated 63 million tonnes or some 5.8 million tonnes (8%) less than the 68.85 million tonnes estimated according to the moderate demand growth scenario (Table 4). The net effect is a rise in imports estimated to reach 2.9 million tonnes by 2030 (Table 8), or 5% of total potato demand, as more challenging environmental conditions dampen further increases in potato production and productivity (Table 9).

Table 9 Area, yield and production of potato in India in 2010 and estimates to 2030, according to different scenarios
Full size table

Discussion

Under all three scenarios, India will see a 20–30 million tonnes increase in potato demand by 2030 enabling noteworthy increases in per capita potato consumption. Given that situation, among the foremost opportunities for private investment are those related to the cold storage industry. Some 60–70% of the potatoes produced in India are utilized after the harvest has taken place (Reardon et al. 2012). Cold storage is the principal modality for keeping the tubers.

These demand estimates indicate the need for 16 to 24 million tonnes more of cold storage capacity—beyond an estimated 27 million tonnes already available as of 2017 (Table 10). They also point to the need for upgrades or replacement of ageing current capacity as part of a major effort to improve energy efficiency in such units and the quality of the tubers subsequently made available for human consumption or use as planting material (NCCD 2015). A related opportunity concerns new technology for energy saving, if not energy substituting (e.g. solar) in the years ahead to reduce dependence on the country’s chronically deficient energy grid and the tendency for power cuts to undermine more effective cold storage management. Various recent commentators have noted the considerable unrealized potential for renewable energy in India (OECD 2017) and the potential for supply chain improvements in response to carbon emissions in India and elsewhere in the decades ahead (Carbon Trust-BSR 2017). The massive and expanding cold storage sector would appear then to be a prime candidate for the combination of policies governing trade and investment necessary to facilitate innovation in the form of renewable energy in this instance.

Table 10 Production, storage available and capacity (%) for storing potatoes during 1979–1980, 1989–1990, 2009 and 2016 for selected states
Full size table

A related opportunity concerns the geographic distribution of cold storage capacity and its use. In the age of the Internet and with the growing use of mobile phones by potato farmers in India (Reardon et al. 2012), taking full advantage of India’s noteworthy software capabilities might result in innovative public–private commercial partnerships. Specifically, more might be done to generate continuous quantitative estimates of the relationship between available cold storage capacity and use in relation to annual potato production in the high potato production states (and districts) of India: Uttar Pradesh, West Bengal, Bihar and Gujarat (Rana and Anwer 2018). Such data could facilitate more informed planning of private investments for the much needed future expansion of cold storage capacity rather than the apparent prevailing oversupply in some locations (underutilized existing capacity) (Reardon et al. 2012) and an undersupply in others (with the run up in production in certain states) (Table 10). Such a scenario seems particularly true given that government operating licences traditionally are required to open such facilities as well as facilitate the purchase and importing of the necessary equipment to make them operational (Dahiya and Sharma 1994; Fuglie et al. 2000; Scott and Suarez 2011; NCCD 2015; Anonymous 2017). It might also catalyse producers and traders to make more effective use of the cold storage capacity in the future.

Related opportunities linked to increased potato consumption involve further development of potato processing and processed potato products (Pandey et al. 2006; Pandey et al. 2009) as well as emerging niche markets for different types (e.g. organic) of fresh tubers. The latter would include improvements in packaging of fresh potatoes such as in pre-weighed, sorted 5–10 kg packets, grading and sales by variety and/or geographic origin as has emerged in other developing countries (Devaux et al. 2016). As urbanization accelerates in the coming decade, this may well offer opportunities for supermarkets and other actors in food retailing in India (Reardon and Minten 2011; Sharma and Bhardwaj 2015) to capitalize on opportunities related to increased consumption of potatoes all the way from improving farm-to-market linkages to fresh product promotion and carry out meals that include the tuber in different prepared forms, as has proven highly successful in other developing countries (Lau 2008; Devaux et al. 2016). A related issue involves nutritious ways to prepare potatoes (e.g. baked rather than simply fried or boiled), a topic of growing interest as consumers in both industrialized and developing countries focus more on eating habits as a way of improving their health and that of their children (Priestly 2006; AAFC 2007) and overcoming some of the misconceptions often associated with potato consumption in developing countries (Woolfe 1987; Scott 2011).

As regards potato processing, it should be noted that although future population growth will mean increased urbanization, by 2030, roughly one billion people in India will still be living in the countryside. Under such circumstances, opportunities for building on prior research on village-level processing into such products as solar dried chips or rustic flour (Nave and Scott 1992; Scott et al. 1993) would appear likely to receive renewed interest. In that context, further applied innovation capitalizing on local capabilities with small-scale machinery (Tewari 2017) and the access to advances available through the Internet might well enable such processing activity to expand. For that to happen, however, will require overcoming the previously documented challenges associated with the economic, as opposed to technical feasibility of such processing operations, issues related to the minimum scale to achieve financial sustainability and the quality control and hygiene standards for such products produced for direct human consumption under rustic conditions (Scott and Suarez 2011).

In a similar spirit, any discussion about processed potato products in India that touches on the potential for starch (Rana 2011) would do well to recall the major restructuring of the potato starch industry that took place in Europe with the expansion of the European Union. Less efficient producers in places like Poland were driven out of the market by very large-scale, specialized processors in Western Europe (Haase and Haverkort 2006). The outcome left idle considerable excess capacity posed to re-start operations should prices for potato starch rebound in the future.

Lastly, every effort should be made to continue the habit of eating potatoes with their skins intact so as to optimize the nutritional benefits and avoid the potential adverse impact from overeating unhealthy foods (Jacobs and Richtel 2016; Anand 2017). The potential for adopting, then introducing currently available improved varieties with specific nutritional traits, such as high vitamin A and E contents, also merits particular attention (Chitchumroonchokchai et al. 2017) as do efforts to encourage greater consumption of fresh potato so as to optimize the in-take of the tuber’s nutritional attributes (Woolfe 1987).

Conclusion

While much of the literature on future food consumption in India in the decades ahead focuses on cereals, this paper presents estimates for potato demand to the year 2030 according to three different scenarios. All three scenarios envision a substantial increase in consumption in both absolute and per capita terms with very modest, if any, increases in foreign trade. These results highlight the potato’s growing importance in Indian diets as the king of all vegetables, as food consumption patterns continue to evolve while maintaining their traditional roots.

The implications of these estimates suggest a series of opportunities made more attractive due to increased demand for potato in India in the years ahead. These include initiatives based on the most likely increase in demand for expansion and improvements in cold storage hardware and software. They also point to growing consumption of processed potato products both in rural as well urban areas, the emergence of different niche markets for fresh potatoes intended to cater to particular tastes and preferences and the potential to make more effective use of the potato’s nutritional attributes to target dietary deficiencies identified by previous research.

Topics for future research that merit the highest priority would include a better understanding of trade patterns in the domestic market. There is a great deal of good quality information about potato production in India but very little, if any, statistics on internal trade flows across states for potato marketing. Simple monitoring of cold storage capacity and utilization at the district level—akin to the district-wise estimates of production, area and yields—albeit on a more continuous as opposed to once-a-year basis would be a first step in that direction. A related topic would be research to foster an energy innovation breakthrough in cold storage industry. Breeding and varietal trials that not only incorporated post-harvest traits like retention of vitamin and mineral content in storage as well as the diffusion of those planting materials with good agronomic traits but also enhanced quantities of vitamins A and E would be highly desirable for their impact potential. Finally, further advances as regards estimates of future demand for potato in India would be a given as this study is but one point-in-time contribution toward that overall effort.