Does subjective well‐being matter when assessing the impacts of irrigation infrastructure? Empirical evidence from Madhya Pradesh, India

Investments to rehabilitate irrigation infrastructure aim enhance climate resilience and improve farmers' livelihoods in the face of hydrological risks. Studies seeking assess impacts from investments primarily focus on yield income-related poverty measures. This approach provides an incomplete picture distributional implications investments, such as impact rehabilitated subjective well-being farmers. Drawing data 707 farmers Madhya Pradesh, India, we explore whether are associated with changes among smallholders. Our findings show a positive relationship between access surface improvements. Results reveal that areas canals experience improvements which 20%–30% higher comparison without canals. When results disaggregated, find larger effect for deficient-rainfall compared average- or high-rainfall areas. The integration contextual factors related into future assessments India can knowledge multidimensional benefits canal diverse farming groups facing differing degrees risk. Les investissements pour remettre en état les infrastructures d'irrigation visent à renforcer la résilience climatique et améliorer moyens de subsistance des agriculteurs aux risques hydrologiques. études visant évaluer dans l'irrigation se concentrent principalement sur mesures pauvreté liées au rendement revenu. Cette approche fournit une image incomplète distributives investissements, telles que l'impact réhabilitées le bien-être subjectif agriculteurs. En nous appuyant données du Inde, examinons si réhabiliter sont associés changements petits exploitants. Nos résultats montrent relation entre ayant accès réhabilitée améliorations bien-être. révèlent zones avec canaux réhabilités connaissent leur subjectif, qui 20 30% plus élevés sans réhabilités. Lorsque désagrégés, constatons un effet positif important pluviométrie insuffisante par rapport moyenne ou élevée. L'intégration facteurs contextuels liés futures évaluations Inde peut connaissance avantages multidimensionnels parmi divers groupes d'agriculteurs confrontés différents degrés risque hydrologique. Global agriculture systems under pressure increase food production feed growing populations while simultaneously increasing climatic shocks. In context these challenges, water supply farmers, minimizing risks enhancing (Cook & Bakker, 2012; Swamee Chahar, 2015; Siddiqi, Wescoat, Muhammad, 2018). over course past decades, state governments, financed by World Bank other donors, have channelled public towards rehabilitating degraded designed performance standards (Ward, Dinar, Darghouth, 2006; Burton al., 2011). By investing improving ageing infrastructure, has attempted agricultural climate-induced variability (Food Agriculture Organization United Nations (FAO), Although system help lift constraints productivity (Suhardiman, Clement, Bharati, Wichelns, 2017), there many confounding negatively influence promised investments. Multiple studies exposed not evenly distributed all members community, existing social, political, institutional barriers limit how vulnerable poorer communities benefit (Barker Molle, 2004; Shah, 2005; Berkoff, 2013). As result, had variable often weakly understood (Von Carnap, 2017). partly be attributed types indicators considered traditional infrastructure. To date, evaluations tend outcomes intensification, crop yields, income (Narayanamoorthy, 2001; Shah Singh, 2002). However, highlighted limitations purely economic income-based welfare (Dasgupta Mäler, Lokshin Ravallion, 2005). Assessments gains unable provide insights linkages socio-economic households (Hussain Hanjra, Huang Scholars argue some cases, incorporate well-being1 more nuanced information perception individual's personal happiness satisfaction (Stiglitz, Sen, Fitoussi, 2009; Rogers Alkire Seth, Mendola Simtowe, 2015). Subjective analysis is defined exercise understanding assets, policies, environment individual life (Guardiola, González-Gómez, Grajales, For most part, literature assesses does study (Jana, Palanisami, Manna, 2018; Passarelli addition, research Indian focused empirical investigations association increased availability through Instead, responses respect consumption goods (Linssen, Van Kempen, Kraaykamp, 2011), privilege deprivation water, electricity, toilet facilities (Mishra Shukla, 2016), household aquifer sustainability watersheds (Ward importance natural capital satisfying rural indigenous societies (Zorondo-Rodríguez others. Thus, role changing been any studies, far aware. Due lives millions across need understand this paper explores key question: Can assessing farmer before/after with/without better inform development support climate-resilient India? address query, smallholder rainfall using sample five districts where 10-year project systems. structured follows: Section 2 describes methodology, including details site collected, our estimation strategy determining 3 presents results. 4 concludes discussion wider policy implications. Pradesh government initiated Water Sector Restructuring Project (MPWSRP) 2005, programme Bank, was multiple river basins deliver timely effective services (World maximizing its usage 495,000 ha (hectares) potential culturable command (CCA)2 31 6 out 10 (Chambal, Sindh, Betwa, Ken, Tons, Wainganga) (Figure 1). Source: Biswas, S. (2017), Resources Department, Government cropping intensity, efficiency canals, promote summer (kharif) crops (e.g., rice) winter (rabi) wheat gram) (Sinha Upon closure 2015, made total investment USD$ 440 million 202 minor, 21 medium, 5 major schemes 228 projects 2018).3 MPWSRP rehabilitation included repairing dams embankments, addressing seepage drainage issues, boundary stones, sluice radial gates lining old concrete reduce distribution claimed reached goal irrigating improved also created trained 400 user associations (WUA). Over half (61%) 193,000 farms area small average landholding size less than 1 Most respondent poorest society mainly coming caste Out within basins, received no funding MPWSRP. conducted survey WUA did not. late cross-sectional led corresponding author team enumerators Land Management Institute (WALMI). survey, were selected (treatment cohort) receive (control (Table Farmers both control treatment farmland district respondents subsets population We excluded sampling urban (such Bhopal Gwalior districts, large populations) (see Tables 2–4 detailed section 1.3 supplementary file). elicited about four individual, household, farm-level domains after each verbally provided informed consent.4 selection individuals participation based multistage purposive random techniques (Fowler, Detailed specifications final samples file. No (n = 430) Mean (SD) Rehabilitation 277) gave us economic, social vulnerability, irrigation, agricultural, preferences self-reported during interviews present (at time 2015) questions requesting recall their (prior investments).5 Next, variability, well-being, incorporated annual period years. ascertain sub-basin patterns, on-farm availability, Spatial at scale high, grouped according three following Sinha al. (2018): low rainfall, high rainfall. line previous identifies environmental conditions (elevation, soil quality) significant component poverty, interactions logit model (Iiyama Kim, Mohanty, Subramanian, 2016). Details collection 1.2 core relies district-fixed effects test measure controlling exogenous variables district-specific indicators, management practices take note farmer's response question constructing dichotomous dependent representing well-being. Specifications logistic regression well series robustness checks different stratifications validity identification 1.4 full (i.e., farmers), 39% farm plots (277 farmers) 61% 430 respondents). characteristics respondents. Table 2, observe significantly likely live average-rainfall equally low-rainfall districts. cohort Vidisha, Neemuch, Ujjain. From sampled male, number female only 3% group 1% group.6 members, own Kisan credit card, house, “semi-permanent” house; “other backward caste” (OBC) (a socio-economically deprived India; Borooah 2014); radio, motorcycle, automobile; obtain poultry/meat production, business, services; land; use hybrid seeds, automatic seed drills, harvesters 4). flood methods upgraded rented house “other” ownership, temporary “general category, bicycle 3). 4, difference terms main source income, 99% production; head, middle, tail ends respective areas; below (BPL) card ownership. additional controls ownership assets indicates asset poor economically vice versa (Filmer Pritchett, 2001). one could owning physical implies better-off households, potentially imply feelings others Harttgen, Klasen, Vollmer (2013) indices correlate poorly overestimate do rely but include indicator being located head (ensuring first rights water), adopting modern technologies sprinkler drip systems), and/or advanced may contribute outputs thus (Ogundari Bolarinwa, 2019). summary statistics probability (20%) illustrate Figure 2. figure average, although experienced time, improvement group. Respondents likely, therefore, (or rehabilitation) group, hand, (27%) deteriorated remained same error bars follow 95% confidence interval. MPWSRP, determine influences sample, report 5. column reports estimates basic specification explanatory “rehabilitation” set sex, years education, income. make contribution treatment. second column, consider type, owned, location, preferred methods, adopted. columns indicate coefficients highly coefficient regressor relatively unaltered (after controls). Full (1) (2) further interpret computed marginal (AME) 6). that, keeping fixed, reporting 29% greater without, coinciding 1, 6, when added, 28% who grow horticultural improvement, engage dairy 5). statistics, however, horticulture whereas 30%–33% Lastly, semi-permanent communication correlation gains. illustrates specific mobile phone) evidence gender, family noted above, very few female. location plot appear nor drip/sprinkler except ridge furrow method. ensure results, alternative specifications. check reformulation still generates consistent align presented 6. First, it argued degree hydroclimatic zone intra-annual average) exert stronger “investment effect.” (2018) little yields implying despite sensitive fluctuations patterns. incomes directly, scarcity stemming reduced conjecture, run dummy zones. interaction term rehabilitation. term, moderates 7. Influence Rainfall Overall, 7 independent zones, remains stable, positive, significant. predictive margins comparing 0 (no MPWSRP) (investment 8). Predictive (standard error) Based margins, compare relative margin (column 8) 8), constant. constant, rehabilitated. see areas, 27% 8% available, lower-than-average times zone. These against climatological rainfall), consequential levels, strongest significant, larger, (in addition queries around ownership) general, OBC, scheduled caste, tribe), they BPL (yes no), no). variables, ways levels. possible nonlinear varying measures, changes. 5–8 2.1.2 file along explanation file) allow conclude increases likelihood even owns membership Last, “recall horizon” nonrehabilitation cohorts running Ujjain 2006 2007 overall 201 equal groups. highest practices. AME 21% experiencing upgraded. reassuring confirm limitation design periods unequal sizes treated highlights reduction size, aligning gives model's indeed due tables, 2.1 consequences complex, multidimensional, understood. Traditional exclude considering India. paper, explored assessment, ceteris paribus, result stable limitations. econometric strategy. reveals deficient tank aims agriculture, stress rainfall-deficient Pradesh. signal enable cope sense satisfaction, employment sample. findings, lowest subcategories smallholders matters populations, caveat gendered accrue mostly male insight Past states consideration Kumar, Building analyses study, element should order granular impacted interventions completed. climate-related principally particularly agriculture-dependent regions, recognize multidimensional. While important, identified complementary guiding irrigation. At present, process new construction throughout state.7 Yet presently donors designing, implementing, monitoring large-scale programmes previously developing (a) comprehensive respond infrastructure-related contexts (b) learning “degree changed” integrated resource guides exploration what manner scholars, officials develop current approaches advance strategies states. available upon reasonable request. Data S1. Supporting Information Please note: publisher responsible content functionality supporting supplied authors. Any (other missing content) directed article.