The researchers used a mixed method of data collection. They purposively selected 100 farmers who cultivated both ginger and turmeric from identified villages in Sikkim. Their stated aim was to assess economic viability, technical efficiency and associated risk during the period from 2019–2020 to 2023–2024, as well as analyze constraints faced during transitions in the state. Because farmers were purposively rather than randomly selected, the results describe the sampled farms and do not establish that the same changes occurred across all organic ginger and turmeric farms in Sikkim.

According to the abstract, Data Envelopment Analysis was used to measure relative technical and scale efficiencies. The mean pure technical efficiency score for organic ginger increased from 59.00% in 2019–2020 to 81.00% in 2023–2024. Organic turmeric recorded a rise from 49.00% to 74.00% over the same period. These are relative efficiency scores produced by the analysis rather than direct measures of yields or productivity alone. Turmeric was comparatively more vulnerable to risk, as indicated by its higher income variability and lower efficiency score.

A Minimization of Total Absolute Deviations model also indicated a significant reduction in income variability and production risk for both crops over the study period. The authors interpret the results as supporting examination of organic cultivation benefits over the long term and beyond yield. The evidence available here is limited to the abstract of an accepted article published online in advance of the final version, which the publisher says remains subject to further edits. The abstract does not provide the underlying observations, model specifications or uncertainty estimates, and the observational comparisons do not establish that the measured changes were caused by organic farming practices alone.

Subject of Research: Economic viability, technical efficiency and modeled production risk of organic ginger and turmeric farming in Sikkim

Article Title: Economic viability, technical efficiency, and risk assessment of organic ginger and turmeric farming in Sikkim

Article References: Roy, B. C., Mondal, B., Roy, A., & Bhattacharyya, S. (2026). Economic viability, technical efficiency, and risk assessment of organic ginger and turmeric farming in Sikkim. Discover Sustainability. https://doi.org/10.1007/s43621-026-04908-w

Image Credits: AI Generated

DOI: 10.1007/s43621-026-04908-w

Keywords: organic farming, ginger cultivation, turmeric cultivation, Sikkim, farm economics, technical efficiency, production risk, income variability, data envelopment analysis, MOTAD, Economic, viability

The researchers, writing in Frontiers in Sustainable Food Systems, combined crop-specific DEA models with a pooled analysis, Tobit regression, fractional-logit robustness tests and non-parametric comparisons across the valley’s three geographical zones. Apple, pear and cherry production were analyzed separately because their underlying technologies may differ. The pooled model placed all 418 farms against a common frontier and generated the efficiency scores used in the second-stage analysis of household, institutional and regional factors. This design also allowed the team to distinguish overall technical efficiency from pure technical efficiency, which measures performance independently of production scale, and scale efficiency, which measures how closely a farm approaches the most productive scale.

DEA is a non-parametric method that constructs an empirical production frontier from observed combinations of inputs and outputs. Farms closer to that frontier are considered more efficient relative to their peers. The study used total horticultural output, measured in kilograms, as the single output. Eight inputs entered the models: landholding in marlas; urea, diammonium phosphate and potassium fertilizer in kilograms; vermicompost in kilograms; family and hired labour in man-days; plant-protection sprays in litres; and the number of horticultural assets. The input-oriented specification was chosen because growers have more discretion over the quantities and combinations of inputs they employ than over output, which is also affected by weather, biological conditions and markets.

Under the constant-returns-to-scale model, the mean score was highest for apple growers, followed by pear and cherry growers. More than half of apple farms, 51.08%, scored at least 0.80, compared with 28.07% of pear farms and 28.82% of cherry farms. Fully efficient farms—those scoring 1—accounted for 8.19% of apple growers, 6.38% of pear growers and 5.90% of cherry growers. Cherry production had the lowest average score and the widest dispersion, indicating especially uneven performance among its farms. These figures describe relative efficiency within the sample, not absolute productivity or the maximum output physically possible on each orchard.

The variable-returns-to-scale results were markedly higher: 92.05% of apple growers, 92.60% of pear growers and 92.01% of cherry growers were classified as efficient when farms were benchmarked against others of comparable scale. Scale-efficiency averages were substantially lower, at 0.8181 for apple, 0.7218 for pear and 0.6771 for cherry, and only 8.19%, 6.38% and 5.90%, respectively, reached full scale efficiency. In other words, most growers appeared to manage their available inputs effectively within their own operating size, yet many operated below or beyond the scale that would have produced the strongest frontier performance. The findings do not mean that every farm should be enlarged: scale efficiency identifies a distance from the sample’s productive scale, not the crop or size that would be optimal for an individual farm.

Farm size and production also differed sharply among crops. Apple growers reported average output of 8,488.63 kg from an average holding of 136.17 marlas, while pear growers averaged 1,643.44 kg from 36.63 marlas and cherry growers 206.15 kg from 29.27 marlas. Large standard deviations in output and input use showed substantial heterogeneity among farms. Some households cultivated more than one fruit, so the crop-specific samples included 415 apple growers, 392 pear growers and 288 cherry growers even though the household survey covered 418 farms. Sampling was multi-stage: all 10 districts were placed into North, Central and South Kashmir; two horticulture-intensive blocks were selected in each district; and commercially active households with relevant production experience were then purposively selected.

In the pooled constant-returns model, the mean farm-level score was 0.5823, with a standard deviation of 0.1623 and a range from 0.3050 to 1.0000. The lower pooled mean reflects the greater heterogeneity of a common frontier encompassing different crops, farm sizes and production systems. Tobit regression, which accommodates efficiency scores bounded between zero and one, showed that the explanatory variables were jointly significant. Family size had a positive coefficient of 0.0422, education of the household head 0.2196, artificial irrigation 0.1504 and access to institutional credit 0.0727, all with p-values below 0.01. Cold-storage availability was also positive, with a coefficient of 0.1041, but was significant only at the 10% level.

Regional location was associated with the pooled efficiency measure even after the regression accounted for household and institutional variables. Relative to Central Kashmir, coefficients were 0.1734 for North Kashmir and 0.2197 for South Kashmir, both significant at p<0.01. A fractional-logit model broadly reproduced the direction of the main relationships: family size, education, irrigation, credit and regional location remained positive, while cold storage became significant at the 5% level. Economic status was not significant in the Tobit model but negative and significant at p=0.008 in the fractional-logit specification. Market accessibility was negative in both models, significant at 10% in Tobit and 5% in fractional logit. These are conditional associations rather than evidence that any factor caused changes in efficiency.

Non-parametric tests supported the presence of geographic variation. A Kruskal–Wallis test rejected the null hypothesis that efficiency distributions were identical across the three zones at p=0.015. Pairwise Mann–Whitney comparisons with adjusted p-values found a significant difference between South and North Kashmir at adjusted p=0.018. The adjusted Central–North comparison was not significant at the conventional 5% threshold, while South–Central had an adjusted p-value of 1.000. Together with the regression results, this points to location-specific production conditions, infrastructure or services as possible explanations, although the cross-sectional design cannot isolate which underlying features produced the differences.

The study offers evidence that efficient horticulture in the Kashmir Valley may depend as much on production scale and supporting institutions as on better allocation of fertilizer, labour and other inputs. Its results associate technical efficiency with larger families, literacy among household heads, irrigation, credit, cold storage and location, while also showing that proximity to a market alone does not correspond to higher measured efficiency. Several limitations constrain generalization: the data were cross-sectional, so they do not track changes over time; they came from one valley and three fruit crops; and purposively selected commercial growers may differ from households outside the sample. The authors also note that future work could add environmental indicators, panel data and comparisons with other horticultural regions. The present results identify measurable associations and the relative importance of scale, but do not demonstrate that a specific intervention would raise output, farm income or environmental performance.

Subject of Research: Farm-level technical efficiency among apple, pear and cherry growers in Kashmir Valley

Article Title: Farm-level technical efficiency in temperate horticulture: evidence from apple, pear, and cherry growers

Article References: Sajjad, S., Dar, K. H., Ganie, A. U. H., Yousif, G. M., Salim, M., & Othman, M. S. (2026). Farm-level technical efficiency in temperate horticulture: evidence from apple, pear, and cherry growers. Frontiers in Sustainable Food Systems, 10, Article 1928386. https://doi.org/10.3389/fsufs.2026.1928386

Image Credits: AI Generated

DOI: 10.3389/fsufs.2026.1928386

Keywords: technical efficiency, Data Envelopment Analysis, Kashmir Valley, apple cultivation, pear cultivation, cherry cultivation, scale efficiency, rural credit, irrigation, sustainable horticulture, Farm-level, technical

The researchers built a balanced firm-level panel from the CMIE Prowess database. Companies with incomplete financial information were removed, leaving 647 textile businesses observed in all three selected years. The sample covered 11 National Industrial Classification sub-sectors and included market leaders as well as companies described as struggling in the market. The three observation years were spaced across a decade to capture changes in the sector’s efficiency. The design does not represent every year in that interval, and the absence of informal producers means the findings apply to the formal firms included in the database rather than to the entire Indian textile industry. Regional effects were also not assessed because regional data were unavailable.

Technical efficiency was estimated with input-oriented Charnes, Cooper and Rhodes DEA under constant returns to scale and Banker, Charnes and Cooper DEA under variable returns to scale. The constant-returns model produced overall technical-efficiency scores, while the variable-returns model separated pure technical efficiency from scale efficiency. In this framework, a score of one denotes a firm on the estimated efficient frontier. A firm with a pure technical-efficiency score of one but an overall score below one was classified as locally efficient, indicating that its apparent shortfall was associated with scale rather than management of the input-output mix. DEA estimates a piecewise production frontier from observed input-output data without imposing a specific functional form on the technology.

All three years of observations were placed against one global frontier rather than separate annual frontiers. This common-reference approach was intended to make changes over time comparable, because conventional year-by-year DEA estimates can combine changes in a firm’s performance with shifts in the technology frontier itself. The analysis used three inputs: raw materials, labour and gross fixed assets converted into a capital-stock measure. Monetary values were deflated using relevant wholesale-price indices for net sales and raw materials and a national capital-stock deflator. Net sales served as the output variable. The sample of 647 firms was large relative to the number of inputs and outputs, reducing, though not eliminating, the possibility that the estimated frontier would be distorted by a small number of observations.

The bootstrap procedure produced both original and bias-corrected DEA estimates. According to the study’s test, the calculated bias measure was below one in each selected year for overall and pure technical efficiency, so the authors ruled out the bias-corrected estimates and interpreted the original CCR and BCC results. The descriptive results showed mean pure technical-efficiency scores of only 0.115 to 0.129 across the three periods, a very large distance from the frontier. Mean scale-efficiency scores were much higher, around 0.75 in the earlier observations and 0.723 in 2020–21. The smaller change in scale efficiency, contrasted with the persistently very low pure technical efficiency, points toward shortcomings in how firms used resources rather than a simple mismatch in production scale. The study’s interpretation is that weak managerial practices, inadequate input utilization and limited technology adoption were central contributors to the observed inefficiency, although the analysis did not directly measure each management practice or technology adoption.

Distributions by enterprise size showed no convincing upward shift over the decade. Large, medium, micro and small firms were all heavily concentrated in the lowest technical-efficiency range, and few firms approached a perfect-efficiency score. The authors suggested that large organizations may face bureaucratic complexity, regulatory burdens or difficulty in using their capacities fully, whereas smaller enterprises may be constrained by limited access to capital and technology. These interpretations remain plausible explanations rather than effects demonstrated by the statistical design, which measured associations among firms rather than experimentally assigning constraints or management practices. The analysis also found technical and scale inefficiencies across all enterprise sizes, indicating that size alone did not guarantee efficient performance.

Sub-sector results were similarly uneven. Manufacture of made-up textile articles other than apparel, wearing apparel other than fur apparel, and knitted and crocheted apparel performed better than many categories. Around 12% to 13% of firms manufacturing carpets and rugs registered scores above 0.9 in the earlier analysis, although that group’s efficiency range declined in 2020–21. Cordage, rope, twine and netting, knitted and crocheted fabrics, man-made fibres and other textiles were among the weaker or stagnant categories. The authors associated many 2020–21 declines with the effects of COVID-19, but the panel includes only one observation from that period and therefore cannot separate pandemic effects from other concurrent changes in demand, operations or the wider economy. The results also indicated that efficiency was not uniformly distributed within sub-sectors, with substantial variation among firms in the same classification.

DEA slack analysis identified resources that could theoretically be reduced without changing output. Excess labour slack appeared in about 39% of firms, with the tendency toward labour abundance increasing alongside actual workforce size. Capital-stock slack was more widespread, appearing in approximately 68% of firms and indicating substantial underuse of physical capacity in much of the sample. Most inefficiency was described as radial, meaning inputs could be reduced proportionately to reach the frontier, although a distinct subset showed a non-optimal input mix. These are model-based slack measurements rather than direct observations of unemployed machinery or idle workers, and their interpretation depends on the estimated production frontier and the quality of the reported inputs. The results therefore identify potential input-reduction opportunities within the model rather than directly documenting how much capacity could be removed in practice.

In the second stage, Tobit regression modeled the bounded efficiency scores, with zero and one representing the limits, while ordinary least squares served as a robustness check. Relative to the reference category of preparation and spinning of textile fibres, seven industrial classifications had significantly higher efficiency estimates: weaving of textiles, knitted and crocheted fabrics, carpets and rugs, other textiles, wearing apparel, knitted and crocheted apparel, and man-made fibres. Carpets and rugs had the largest positive industry effect. Firm size and the capital-to-labour ratio were also positively and significantly associated with efficiency, whereas firm age, ownership type and year effects were statistically insignificant. Coefficient signs, magnitudes and significance were broadly consistent between Tobit and OLS, and all variance inflation factors were below five. The analysis also used the common-frontier DEA estimates to examine how input-oriented efficiency related to slack in labour and capital, rather than treating slack as a direct measure of unobserved managerial quality.

Overall, the analysis points to industry characteristics, firm size and capital intensity as the clearest correlates of technical efficiency, rather than organizational maturity or public-versus-private status. The authors interpret the results as support for better management, workforce training, digital manufacturing, process automation and more productive use of installed capital. They also call for improved quality control, stronger marketing, support for small and medium-sized enterprises, domestic production of man-made fibres, research into high-value textile products, and streamlined regulation. These are policy implications derived from the observed efficiency gaps, not measured outcomes from a policy trial. The study’s main limitations are its exclusion of the large informal textile sector, lack of regional variables, dependence on three annual snapshots and reliance on secondary company data. The researchers note that future work should examine performance after the COVID-19 pandemic and the effects of major policy changes, including simplified GST filing, reduced textile-industry rates and the production-linked incentive scheme launched in 2017. Until such evidence is available, the results describe persistent measured gaps in the formal firms studied rather than demonstrating how a particular policy or managerial change will affect the industry.

Subject of Research: Firm-level technical efficiency and its determinants in India’s textile industry, 2011–12 to 2020–21.

Article Title: Efficiency and its determinants in India’s textile firms using two stage bootstrap DEA and tobit model

Article References: Pawar, S., & Sunny, D. (2026). Efficiency and its determinants in India’s textile firms using two stage bootstrap DEA and tobit model. Discover Global Society, 4(1), Article 280. https://doi.org/10.1007/s44282-026-00616-y

Image Credits: AI Generated

DOI: 10.1007/s44282-026-00616-y

Keywords: Indian textiles, technical efficiency, data envelopment analysis, bootstrap DEA, Tobit regression, capital intensity, labour slack, firm productivity, industrial classification, input utilization, Efficiency, determinants

The research team, led by Ain ul Momina of King Edward Medical University in Lahore and including collaborators from the University of the Punjab, the University of Cambridge, and Charité Universitätsmedizin Berlin, focused on family planning service delivery. This is not a niche concern. Pakistan has long struggled with a contraceptive prevalence rate that remains stubbornly below the levels needed to meet demand, and the national and provincial governments have invested heavily in programs designed to accelerate access to modern contraception. One such initiative, the Delivering Accelerated Family Planning in Pakistan project, known as DAFPAK, has supported facilities in expanding services. Yet whether that additional support translates into genuinely better performance has been difficult to measure with conventional indicators alone, which is where the study’s methodological approach becomes important.

The technique at the heart of the study is data envelopment analysis, or DEA, a non-parametric operations research method that has become a workhorse of health systems research worldwide. Rather than comparing each facility to a fixed benchmark or an average, DEA constructs a frontier of best performers, the so-called efficient frontier, and measures every other facility by its distance from that frontier. Each clinic, hospital, or health unit is treated as a decision-making unit that consumes inputs, in this case contraceptive stock, staff availability, and facility size, and produces outputs in the form of services delivered. A facility that achieves the maximum output from its given inputs is rated as fully efficient, with a score of one hundred percent, while a facility producing less than it could with the same resources receives a proportionally lower score. The method’s appeal lies in its ability to handle multiple inputs and outputs simultaneously without requiring the researcher to specify in advance how each factor should be weighted.

The study’s design was deliberately two-stage. In the first stage, DEA was applied to data collected from one hundred public health facilities managed by the Primary and Secondary Health Department of Punjab. Half of these facilities belonged to the DAFPAK cohort and half did not, allowing the researchers to examine whether participation in the accelerated family planning program was associated with different efficiency profiles. Data were gathered through computer-assisted personal interviews and drawn from electronic medical records, a choice that streamlined collection and, because the records system already incorporates patient permission at registration, meant that separate consent procedures were not required. Ethics approval came from the Departmental Doctoral Program Ethical Committee of Public Health at the University of the Punjab’s Institute of Social and Cultural Studies.

The headline result is that the average technical efficiency score across the one hundred facilities stood at seventy-eight percent. In practical terms, this suggests that the typical facility could, in principle, deliver roughly the same level of family planning services with about twenty-two percent fewer inputs, or conversely, expand its output by a similar margin without consuming any additional resources. In a province where every rupee and every trained health worker counts, that gap represents a substantial hidden reserve of capacity. The researchers emphasize that DEA does not merely rank facilities; it identifies specific peers against which underperforming units can be compared, offering managers a concrete picture of what better performance looks like in comparable settings rather than an abstract target.

Beneath the average, however, the picture becomes more uneven. The study found significant variations across facility types, with Basic Health Units and Rural Health Centers showing particularly pronounced resource optimization gaps. These are the front line of Punjab’s rural health system, the smallest and most numerous institutions, often staffed by a handful of workers and serving dispersed populations. The finding that they lag in efficiency is doubly consequential. First, it means that the resources allocated to rural primary care are not stretching as far as they could. Second, because these facilities are often the only accessible point of care for rural women seeking contraception, inefficiency at this level translates directly into unmet need among the very populations that family planning programs most aim to reach.

The second stage of the analysis tackled a question that first-stage DEA alone cannot answer: why do some facilities perform better than others? Efficiency scores are influenced not only by how a facility manages its own resources but also by its surrounding environment, factors beyond the control of any individual manager. To disentangle these effects, the researchers employed the Simar-Wilson regression approach, a statistical method specifically developed for the truncated nature of efficiency scores produced by DEA. Ordinary regression techniques are ill-suited to such scores, which are bounded between zero and one, and the Simar-Wilson procedure corrects for this, yielding more reliable estimates of how environmental variables shape performance.

One environmental factor stood out. The analysis revealed a negative association between the number of other healthcare providers in a facility’s catchment area and that facility’s efficiency. In other words, clinics operating in areas crowded with alternative providers, whether public, private, or non-governmental, tended to score lower on technical efficiency than clinics serving areas with fewer competing sources of care. The interpretation offered by the researchers is instructive: service provision needs to be aligned with local demand. When multiple providers cluster in the same catchment, each may draw a smaller share of patients, leaving facilities with staff and supplies that are underutilized relative to their capacity. This is a planning problem as much as a management one, and it points toward smarter geographic allocation of resources rather than simply adding more facilities or more stock in areas that are already well served.

The study’s implications extend well beyond Punjab. Resource-constrained health systems around the world face the same fundamental challenge of deciding where to direct scarce funds, and the two-stage DEA framework offers a template for answering that question with data rather than intuition. For Pakistan specifically, the findings arrive at a moment when family planning remains a critical lever for maternal health, child survival, and broader development goals. The authors conclude that DEA effectively identifies inefficiencies and can inform resource optimization in public health facilities, and they argue that their results offer valuable insights for improving family planning services and healthcare delivery in settings where every input must count. The research received no specific grant from any funding agency, and the underlying data came from facilities’ own electronic records, giving the analysis a grounded, operational character.

What makes the study compelling is its refusal to accept averages as the end of the story. A seventy-eight percent efficiency score sounds respectable until one recognizes that it conceals a distribution in which rural Basic Health Units and Rural Health Centers fall well short of their potential. By pairing a frontier-based efficiency measurement with a statistically appropriate second-stage regression, the researchers have converted a routine performance audit into a diagnostic tool that tells policymakers not only where the waste is but also what surrounds it. Aligning service provision with local demand, as the negative association with provider density suggests, may prove as important as any internal reform. For the millions of women in Punjab whose access to contraception depends on the performance of a small rural clinic, that insight could make the difference between a health system that merely functions and one that truly delivers.

Subject of Research: Technical efficiency of public health facilities delivering family planning services in Punjab, Pakistan, assessed using two-stage data envelopment analysis

Article Title: Technical efficiency of public health facilities in Punjab, Pakistan: application of a two-stage data envelopment analysis

Article References: Momina, A. U., Ahmad, T., Zakar, R., Shahzad, S., Ahmad, F., Mahmood, S., & Fischer, F. (2026). Technical efficiency of public health facilities in Punjab, Pakistan: application of a two-stage data envelopment analysis. BMC Health Services Research, 26(1), Article 1397. https://doi.org/10.1186/s12913-026-15847-7

Image Credits: AI Generated

DOI: 10.1186/s12913-026-15847-7

Keywords: data envelopment analysis, technical efficiency, public health facilities, family planning, Punjab, Pakistan, contraceptive prevalence, DAFPAK, Simar-Wilson regression, health services research, Basic Health Units, resource optimization

The stakes are high. Maize is the cheapest source of calories among Ethiopia’s major cereals, dominating household diets and providing roughly twice as many calories per dollar as other staples. Yet Ethiopian farmers harvest only about 30 percent of the crop’s potential yield, held back by limited access to improved technologies, market imperfections, liquidity constraints, and technical inefficiency on the farm itself. Earlier efficiency studies have estimated efficiency gaps of 14 to 28.3 percent, meaning producers lose more than a quarter of their potential output to avoidable waste. Closing even part of that gap could translate into meaningful gains for food security across the region.

Cluster farming, introduced nationally through the agricultural commercialization cluster initiative in 2010, is designed to overcome precisely these barriers. The approach groups farmers geographically around a commodity’s production potential and encourages them to coordinate land preparation and planting, share improved seeds, fertilizers, and crop protection chemicals, and receive concentrated support from extension agents. The idea is to shift smallholders from subsistence production toward market-oriented farming by pooling resources and standardizing agronomic practice. In principle, a farmer embedded in a cluster can apply recommended input rates, time sowing within a narrow window alongside neighbors, and tap into collective bargaining power that an isolated plot owner cannot.

To test whether the strategy actually improves efficiency rather than merely raising yields, the researchers surveyed 421 randomly selected smallholder maize farmers during the 2021 production season across the Halaba, Wolaita, Gamo, and Gofa zones. Half of the sample, 206 farmers, produced maize within clusters, while 215 farmed outside them. The team used a two-stage sampling design, first selecting six districts with strong maize potential, three cluster districts and three non-cluster districts, and then randomly drawing households from kebeles within each. A structured questionnaire captured everything from household demographics and land use to input quantities, agronomic practices, and institutional variables such as credit access and extension contact.

The analytical machinery behind the study is where its technical rigor lies. The researchers estimated a stochastic frontier model using a flexible translog production function, chosen over the more restrictive Cobb-Douglas form after a likelihood ratio test decisively rejected the simpler specification. Because smallholder farmers operate under wildly varying socioeconomic and institutional conditions, the team employed a heteroscedastic frontier model, allowing the variance of inefficiency itself to depend on factors such as cluster participation, oxen ownership, income, and extension contact. Ignoring this heterogeneity, the authors note, can seriously bias efficiency estimates. Inputs and outputs were scaled by their sample means so that first-order coefficients could be read directly as elasticities evaluated at the mean.

The frontier results paint a clear picture of what drives maize output. Land is by far the most responsive input, with an estimated output elasticity of 36 percent, followed by labor at 27 percent and fertilizer at 16 percent. Oxen power and seed contributed smaller elasticities of 8 and 6 percent respectively, and the returns to scale estimate of 0.81 indicated decreasing returns overall. Notably, the second-order effect of fertilizer significantly increased returns to output, which the authors attribute to the very low baseline application of inorganic fertilizer, itself a consequence of liquidity and institutional constraints. The second-order terms for oxen and labor were negative, signaling diminishing returns as these inputs intensify.

The headline finding is the efficiency gap between the two groups. Farmers producing maize in clusters achieved an average technical efficiency of 74 percent, while non-cluster farmers reached only 60 percent. A kernel density analysis of the efficiency distributions confirmed that cluster farmers’ scores were skewed toward higher values. Because farmers who join clusters may differ systematically from those who do not, the team went a step further, applying an endogenous switching regression model that accounts for both observable and unobservable differences between participants and non-participants. The frequency of extension contact served as the instrumental variable identifying the model, since extension agents actively promote cluster participation but do not directly determine efficiency.

The counterfactual analysis sharpened the conclusion considerably. Cluster farmers would have lost 18 percent of their technical efficiency had they not participated in cluster production, the average treatment effect on the treated. Conversely, non-cluster farmers would have gained 33 percent in efficiency had they joined, the average treatment effect on the untreated. In other words, the farmers most likely to benefit from clustering are precisely those currently standing outside it. The first-stage probit model, which fit the data well, identified male household headship, oxen ownership, frequent extension contacts, greater market distance, credit access, and household income as significant drivers of cluster participation, with the market distance result suggesting that farmers far from markets see cooperation as a way to cut transaction costs.

The study’s authors argue that the findings carry direct policy weight for a region where land fragmentation is more severe than elsewhere in Ethiopia. They recommend that policymakers and development organizations promote clustering in crop production while addressing the socioeconomic factors that gate participation, including gender gaps in access to extension advice and liquidity constraints that limit input purchase. Strengthening the improved seed system, modernizing extension through digital communication technologies, and promoting climate-smart practices are flagged as complementary priorities. The authors also emphasize that collective action among farmers, government, the private sector, and donors will be essential to scale the model.

For a country where agriculture anchors employment and output but productivity remains stubbornly suboptimal, the message from southern Ethiopia is unusually concrete: farming together measurably reduces the waste that has long eroded smallholder potential. With maize efficiency gaps of this magnitude, the difference between a 60 percent and a 74 percent efficient harvest is not an academic abstraction but the equivalent of several additional weeks of food for a farming household. If the cluster approach can be extended to the non-participants who stand to gain the most, the study suggests, Ethiopia’s staple grain system may be able to squeeze far more nutrition and income from the same scarce land.

Subject of Research: The effect of maize cluster farming on the technical efficiency of smallholder farmers in southern Ethiopia

Article Title: Effect of maize cluster farming on smallholder farmers’ technical efficiency: evidence from Southern Ethiopia

Article References: Fola, M., Tsegaye, G., Zawde, S., & Matsalo, M. (2025). Effect of maize cluster farming on smallholder farmers’ technical efficiency: evidence from Southern Ethiopia. BMC Agriculture, 1(1), Article 4. https://doi.org/10.1186/s44399-025-00006-w

Image Credits: AI Generated

DOI: 10.1186/s44399-025-00006-w

Keywords: cluster farming, maize, technical efficiency, smallholder farmers, Ethiopia, stochastic frontier analysis, endogenous switching regression, agricultural policy, land fragmentation, extension services, food security, agricultural economics

The package, called PEAXAI, short for Probabilistic Efficiency Analysis using Explainable Artificial Intelligence, was developed by Ricardo González-Moyano, Juan Aparicio, José L. Zofío and Víctor J. España and described in the journal SoftwareX. Written in R and freely available under the GPL-3 license on CRAN and GitHub, the software implements a framework the authors introduced in a companion methodological paper. Its central idea is elegant: instead of asking whether a decision-making unit, or DMU, sits exactly on the efficient frontier, PEAXAI trains supervised classifiers to estimate the probability that a unit belongs to the efficient class. That single shift transforms efficiency analysis from a deterministic exercise into an uncertainty-aware one, allowing managers to set benchmarks at whatever confidence level they consider appropriate.

The workflow begins in familiar territory. PEAXAI first estimates a weighted additive DEA model, using the range-adjusted measure that normalizes slacks by the observed spread of each input and output. Units whose objective function evaluates to zero, meaning no input can be reduced or output expanded without worsening something else, are labeled efficient; all others are labeled not efficient. These binary labels then become the training data for machine learning classifiers, which can include neural networks, support vector machines and other algorithms implemented through the widely used caret infrastructure. Crucially, the package evaluates candidate models with stratified cross-validation in which the withheld observations are labeled by a frontier built only from the training subset, so the model’s generalization is tested against a technology it has never seen.

One of the thorniest practical problems the developers had to solve is class imbalance. In real datasets, the efficient frontier is typically sparsely populated. In the package’s own illustrative application to 1,267 meat-industry firms in Spain and Portugal, drawn from the SABI balance-sheet database, DEA under variable returns to scale flagged only 16 firms, roughly 1.26 percent of the sample, as efficient. A classifier trained on such lopsided data would struggle to learn where the boundary lies. PEAXAI addresses this with an adapted version of SMOTE, the synthetic minority over-sampling technique, redesigned to respect the polyhedral geometry of the DEA technology. When efficient units are scarce, synthetic firms are generated as convex combinations lying on maximal-dimensional facets of the estimated frontier, so they remain genuinely efficient. When inefficient units are scarce, synthetic observations are placed strictly inside the technology, close to but not on the frontier.

Once the classifiers are trained, the package turns to explainable artificial intelligence to open the black box. PEAXAI offers both global and local variable-importance tools, drawing on aggregated SHAP values, permutation feature importance and sensitivity analysis for the global picture, and local SHAP values, LIME explanations and sensitivity analysis for unit-specific insight. The practical payoff is that analysts can identify which inputs and outputs most strongly drive the probability of frontier membership, both across the whole sample and for any individual firm. In the meat-industry example, where personnel expenses and fixed assets serve as proxies for labor and capital and operating income is the output, such explanations can reveal whether workforce costs or capital stock dominate the efficiency classification, information that conventional DEA simply does not provide.

The most striking feature of the package may be its counterfactual machinery. Rather than reporting an abstract inefficiency score, PEAXAI quantifies inefficiency as the minimum adjustment a firm must make to cross the probabilistic efficiency boundary. Using the directional distance function, a framework that allows simultaneous input reduction and output expansion along a user-chosen direction, the software searches for the smallest change that pushes a firm’s predicted efficiency probability above a chosen threshold. The direction itself can be weighted by the data-driven variable-importance results, so the improvement path reflects the factors that actually matter to the classifier. Managers can thus ask a concrete question: how much would we need to cut costs or grow revenue to have, say, a 90 percent chance of being classified efficient?

That question has direct economic weight. Because the directional distance function permits improvements on both the input and output sides, the resulting targets translate into simultaneous cost savings and revenue gains rather than indiscriminate austerity. The authors argue this should caution managers against reflexive cost-cutting; in their example, expanding operating income may matter as much as trimming personnel expenses or fixed assets. The package complements this with proximity-based peer selection, pairing each firm with the observed high-probability unit closest to its counterfactual target, either by plain Euclidean distance or by a distance weighted by feature importance. The result is a benchmark that is not merely statistically nearest but managerially relevant, a comparable firm whose practices could plausibly be emulated.

Ranking is also rethought. Traditional DEA rankings rest on a single deterministic score, which can be unstable in high-dimensional settings where the curse of dimensionality erodes the method’s discriminating power and inflates efficiency estimates. PEAXAI offers two ordering schemes: one based purely on predicted efficiency probability, and a hierarchical attainable ranking that prioritizes the maximum probability a unit could reach, then the magnitude of the required input reductions and output expansions, and finally the current predicted probability. This blends where a firm stands today with how far it could plausibly climb, producing a prioritization that is more informative for policy and sectoral planning than a static league table.

The developers are candid about limitations. The DEA labeling stage requires both classes to be populated, which is precisely why the synthetic balancing step exists, and identifying all facets of the estimated technology becomes the computational bottleneck for very large or high-dimensional datasets. The package also does not yet accept fuzzy or interval-valued data, and the authors stress that the predicted probability reflects classifier uncertainty about the DEA label, not the true probability of technical efficiency, so users should rely on the outputs only when discrimination and calibration metrics are satisfactory. Future work, they suggest, includes post-hoc probability calibration, contextual environmental variables, robust frontier labels from order-m and order-alpha models, and faster counterfactual algorithms. Even so, by uniting frontier estimation, machine learning, explainability and counterfactual target setting in a single reproducible R workflow, PEAXAI signals where efficiency analysis is heading: away from verdicts and toward quantified, explained and actionable probabilities.

Subject of Research: Probabilistic technical efficiency analysis combining data envelopment analysis, machine learning classifiers and explainable AI in R

Article Title: PEAXAI: A probabilistic efficiency analysis using explainable artificial intelligence in R

Article References: González-Moyano, R., Aparicio, J., Zofío, J. L., & España, V. J. (2026). PEAXAI: A probabilistic efficiency analysis using explainable artificial intelligence in R. SoftwareX, 36, Article 103089. https://doi.org/10.1016/j.softx.2026.103089

Image Credits: AI Generated

DOI: 10.1016/j.softx.2026.103089

Keywords: data envelopment analysis, machine learning, explainable AI, R package, technical efficiency, SHAP values, counterfactual explanations, directional distance function, benchmarking, class imbalance, SMOTE, open-source software

Technical efficiency is a deceptively simple idea with powerful implications. It measures the ratio of what a farmer actually produces to the maximum that could be produced with the identical bundle of inputs, a concept that traces back to Farrell’s foundational 1957 work on productive efficiency. A farmer operating at 80 percent efficiency is leaving a fifth of potential output on the table, not because of bad luck or poor soil, but because of sub-optimal decisions about how much labor to hire, how densely to plant and how to allocate scarce resources. The Rwandan study, led by Jean Marie Ntakirutimana of Kabuye Sugar Works together with researchers at Jomo Kenyatta University of Agriculture and Technology and Machakos University, quantified exactly that gap for the first time in this crop and region, where empirical evidence on the causes of low sugarcane productivity had been almost entirely absent.

The methodology is worth understanding, because it is what separates this finding from anecdote. The team used stratified random sampling across the two ecological zones suitable for sugarcane along the Nyabarongo River, drawing 202 farmers from three cooperative societies using Slovin’s formula at a 95 percent confidence level. They then estimated a Cobb-Douglas stochastic production frontier with maximum likelihood methods, a technique that splits the deviation between observed and frontier output into two components: random statistical noise, and a one-sided inefficiency term that captures genuine management shortfalls. This is a crucial advantage over non-parametric approaches like data envelopment analysis, which attribute every deviation to inefficiency and cannot easily be tested statistically. The estimated gamma parameter of 0.6297 told the researchers that roughly 63 percent of the variation in output among these farmers stemmed from differences in technical inefficiency rather than random shocks, a moderate but actionable level of controllable waste.

The input-level results contain some genuine surprises. Pesticide use emerged as the star performer: each additional unit of pesticide was associated with a 21.5 percent increase in sugarcane output, a highly significant effect attributed to effective control of pests and diseases, with chlorpyrifos and mancozeb the most commonly applied products. But three other inputs worked against the farmers. Organic compost reduced output by nearly 10 percentage points on average, likely because it releases nutrients slowly while farmers divert their best compost to vegetable crops. Labor told a similar story, with each additional man-day cutting output by about 12 percent, a pattern the authors link to an aging, inadequately trained agricultural workforce, echoing evidence from China that a graying rural labor force drags down crop productivity. Most strikingly, planting more seed cane reduced output by 7.8 percentage points, pointing to overcrowded fields where stalks compete for light, water and nutrients and end up thinner and shorter.

The socioeconomic and institutional findings are even more provocative, because several of them invert conventional development wisdom. Older farmers were more efficient, with each additional year of age reducing technical inefficiency, plausibly because age proxies decades of accumulated knowledge about local agro-ecological conditions, input timing and irrigation. Access to bank credit also cut inefficiency sharply, with a coefficient of minus 1.6433, consistent with evidence from India, Indonesia and elsewhere that timely financing lets farmers buy quality seed setts, fertilizer and irrigation equipment when they actually need them. Only 16.8 percent of the surveyed farmers had access to credit, so this lever remains largely unused. Gender, education level, household size, marital status and market distance, meanwhile, showed no significant effect at all.

Then come the counterintuitive results that should make policymakers sit up. Farmers who received extension services were less efficient than those who did not. Cooperative members were less efficient than non-members. Trained farmers were less efficient than untrained ones. Larger landholdings were associated with more inefficiency, not less. The authors do not dismiss these institutions; they diagnose them. Rwanda’s extension system, they argue, tends toward a standardized, one-size-fits-all advisory model that disseminates uniform messages poorly matched to the specific agronomy of sugarcane, a crop with a long growth cycle and complex management demands. Cooperatives can suffer from weak governance, delayed input delivery and over-reliance on collective decisions that stifle individual initiative. Training sessions and meetings often collide with critical field preparation windows, pulling farmers away from their fields at exactly the wrong moments. And beyond a certain scale, larger plots may simply exceed a household’s management capacity.

Placed in global context, Rwanda’s numbers look respectable but reveal a clear ceiling. Mean technical efficiency of 80.58 percent sits comfortably within the 70 to 85 percent range reported for sugarcane producers in Brazil and Kenya, and well above India’s national figure of 66 percent or Ethiopia’s 60 to 70 percent. Pakistan’s Faisalabad smallholders exceed 90 percent on technical efficiency but collapse to 28 percent on allocative efficiency, hampered by irrigation costs and saline groundwater. Brazil, the global benchmark, reaches 85 to 90 percent thanks to mechanization, advanced agronomy and strong extension systems, while Thailand’s roughly 85 percent reflects government-backed technology adoption and widespread integrated pest management. Rwanda applies moderate pesticide levels but lacks structured IPM, and its reliance on manual labor and organic composting, however sustainable in intent, contrasts sharply with the mechanized systems of the most efficient producers.

The regional breakdown adds another layer of insight. Farmers in the Eastern Province achieved a mean efficiency of 83.16 percent, while those in Kigali City averaged just 78.00 percent, a gap the authors attribute to differences in resource access, infrastructure and the effectiveness of support services. Across the pooled sample, nearly 60 percent of farmers operated in the 81 to 90 percent efficiency band and another 35 percent in the 71 to 80 percent band, with no farmer below 42 percent. The inefficiency gap of 19.42 percent overall, 22 percent in Kigali City and 16.84 percent in the Eastern Province represents the concrete productivity prize on offer. The surveyed farmers averaged 6,261.9 kilograms of cane on plots averaging just 1.22 hectares, confirming that this is an overwhelmingly smallholder system where marginal gains per farm translate into meaningful national totals.

The policy prescriptions that flow from this analysis are refreshingly specific. Rather than simply pumping more money into extension or cooperatives, the authors argue for improving the quality and crop-specificity of advisory services, strengthening cooperative governance, pairing credit expansion with financial literacy training, and promoting optimal planting densities so that farmers stop treating seed cane quantity as a proxy for ambition. They also call for encouraging innovation and youth participation in a sector where the average farmer is 55 years old with fewer than seven years of sugarcane experience. Given that Africa’s sugar output is projected to grow 36 percent by 2030 and developing countries will account for 79 percent of global growth, Rwanda has a genuine opportunity to convert efficiency gains into import substitution and rural income growth. The study’s core message is quietly radical: before Rwanda plants another hectare, it should fix the fifth of its existing harvest that poor management is quietly throwing away.

Subject of Research: Technical efficiency and its determinants in smallholder sugarcane production in Rwanda

Article Title: Optimizing sugarcane production in Rwanda: a stochastic frontier analysis of smallholder farmers

Article References: Ntakirutimana, J. M., Otieno, G. O., Ngigi, M., & Majiwa, E. (2025). Optimizing sugarcane production in Rwanda: a stochastic frontier analysis of smallholder farmers. BMC Agriculture, 1(1), Article 22. https://doi.org/10.1186/s44399-025-00021-x

Image Credits: AI Generated

DOI: 10.1186/s44399-025-00021-x

Keywords: sugarcane, Rwanda, technical efficiency, stochastic frontier analysis, smallholder farmers, agricultural productivity, extension services, cooperatives, agricultural credit, pesticide use, organic compost, sugar imports

The stakes could hardly be higher. Sustainable Development Goal 3.1 calls on countries to reduce maternal mortality to fewer than 70 deaths per 100,000 live births by 2030, while Goal 3.2 targets neonatal mortality below 12 deaths per 1,000 live births. Across the 46 countries examined, Africa remains far from both benchmarks. Maternal mortality averaged a staggering 354.2 deaths per 100,000 live births, roughly five times the global target, and neonatal mortality stood at 23.5 per 1,000 live births, nearly double the agreed ceiling. With the 2030 deadline approaching, the study offers a stark quantitative diagnosis of why progress has stalled, and it points to a factor that international donors and national governments alike have often overlooked in their push to increase funding.

To measure efficiency, the researchers turned to a method originally developed for evaluating the productivity of firms and public agencies. Data envelopment analysis constructs a mathematical frontier representing best practice: each country is treated as a decision-making unit that consumes inputs, such as health spending, doctors, nurses and midwives, and hospital beds, to produce outputs, including reductions in maternal, neonatal, infant, and stillbirth mortality alongside gains in service coverage. Countries whose combination of inputs and outcomes places them on this frontier are deemed technically efficient; those falling inside it are wasting resources relative to what their peers demonstrate is achievable. The team ran the analysis twice, once under constant returns to scale and once under variable returns to scale, the latter allowing for the fact that small countries may face different scaling dynamics than large ones.

The results were sobering. Only 12 of the 46 countries, or roughly one in four, sat on the efficiency frontier, meaning 74 percent of the continent’s health systems perform below best-practice levels. Mean efficiency under variable returns to scale came out at 0.849, a figure with a striking practical interpretation: inefficient systems could theoretically achieve their current outcomes while using approximately 15 percent fewer inputs. Put differently, hundreds of millions of dollars in health spending across the continent are effectively being absorbed by structural waste rather than translated into saved lives. Three small nations emerged as the continent’s benchmark performers: Eritrea, Seychelles, and São Tomé and Príncipe. Their appearance at the top of the efficiency rankings is itself instructive, since none of them ranks among Africa’s biggest health spenders, underscoring the study’s central message that resourcefulness matters more than resources.

Having quantified the efficiency gap, the researchers then asked what explains it. This is where the second stage of their two-stage design comes in. Using Tobit regression, a statistical model suited to dependent variables that are bounded between zero and one, as efficiency scores are, they tested whether a set of financial, coverage, governance, and health-system factors predicted a country’s efficiency score. The inputs and outputs drawn from World Health Organization data spanning 2005 to 2021 included current health expenditure as a share of gross domestic product, per capita health spending, external donor financing, the density of doctors, nurses, midwives, and hospital beds, the proportion of births attended by skilled personnel, vaccination card visibility, and the composite coverage index, a summary measure of reproductive, maternal, newborn, and child health interventions.

The regression results challenge a comfortable assumption that has shaped global health policy for two decades. Current health expenditure, the composite coverage index, and per capita health spending were all negatively associated with efficiency scores. In plain terms, countries that spend more, and countries that achieve higher coverage of maternal and child health services, are not automatically better at converting those resources into mortality reductions. Higher spending can coexist with, and may even mask, deep inefficiencies in how funds are allocated and managed. The finding does not imply that investment is unnecessary; rather, it demonstrates that investment without accountability and optimization yields diminishing returns. A system can report high coverage of antenatal visits or skilled birth attendance while still failing to prevent the deaths that those services are supposed to avert, if quality, timing, and targeting are poor.

The preferred Tobit model displayed remarkably strong explanatory power, with a pseudo-R-squared of 0.8943, indicating that the variables the researchers examined account for the vast majority of the variation in efficiency across the continent. This is unusual in health-systems research, where cross-country models often explain only a fraction of the variance. The strength of the result suggests that the framework captures something real about how African health systems function, and it gives the findings practical weight for governments designing reforms. The study’s period, covering the years from 2005 through 2021, also means the analysis reflects the strains of recent global shocks, including the COVID-19 pandemic, which disrupted routine maternal and child health services across much of the continent.

The authors’ conclusions extend beyond measurement to prescription. Achieving the maternal, newborn, and child health targets of the Sustainable Development Goals, they argue, requires more than increased investment. African health systems must strengthen accountability mechanisms so that funds reach frontline services, optimize resource allocation to match the epidemiological burden, close persistent gaps in the health workforce and infrastructure, and adapt practices from efficient peers. The benchmark countries identified by the analysis provide a natural laboratory for this kind of peer learning. Rather than importing models from high-income countries with fundamentally different resource environments, policymakers can study how countries facing similar constraints manage to extract more health gain from each dollar spent.

What makes the study distinctive within the crowded field of Sustainable Development Goal monitoring is its shift of focus from outcomes to conversion. Most assessments of progress toward Goals 3.1 and 3.2 simply track whether mortality ratios are falling and by how much. That approach implicitly rewards wealthy countries and penalizes poor ones, regardless of how well each manages what it has. By framing each health system as a production process and measuring its distance from the empirical frontier, data envelopment analysis offers a fairer and more actionable benchmark. It tells a finance ministry not merely that it is lagging, but by roughly how many percentage points its inputs could be trimmed, or equivalently how much additional output it should expect, if it operated as efficiently as the best performers under comparable conditions.

The study arrives at a critical juncture. With five years left on the Sustainable Development Goal calendar, many African countries are widely acknowledged to be off track for the maternal and newborn survival targets, and fresh financing alone is unlikely to close the gap in time. The evidence assembled by Er-Rays and Ait Lemqeddem suggests that the continent’s fastest route to progress runs through efficiency: reclaiming the roughly 15 percent of wasted capacity embedded in underperforming systems would represent an enormous, largely untapped resource at a moment when every percentage point of mortality reduction counts. Whether governments and their partners act on that evidence, by treating efficiency, equity, and measurable health gains as the organizing principles of health-sector reform, may well determine whether the promises made in 2015 are kept by 2030.

Subject of Research: Efficiency of maternal, newborn, and child health services in achieving Sustainable Development Goals 3.1 and 3.2 across African countries

Article Title: Efficiency of maternal, newborn, and child services in achieving Sustainable Development Goals 3.1 and 3.2 in Africa: a two-stage data envelopment analysis

Article References: Er-Rays, Y., & Lemqeddem, H. A. (2026). Efficiency of maternal, newborn, and child services in achieving Sustainable Development Goals 3.1 and 3.2 in Africa: a two-stage data envelopment analysis. BMC Health Services Research. https://doi.org/10.1186/s12913-026-15692-8

Image Credits: AI Generated

DOI: 10.1186/s12913-026-15692-8

Keywords: maternal health, newborn health, child health, Sustainable Development Goals, Africa, health systems, data envelopment analysis, Tobit regression, health economics, technical efficiency, health financing, neonatal mortality

The research, published in the journal Discover Agriculture, examined the sources of total factor productivity (TFP) and profit growth in Beninese cotton production between 2000 and 2018. A team led by Idelphonse O. Saliou of the University of Abomey-Calavi assembled an impressive dataset: 482 cotton producers surveyed across three major agroecological zones—the central region around Savalou, the northern district of Banikoara, and the western Atacora region around Cobly. To qualify for the study, producers had to be at least 40 years old with a minimum of 19 consecutive years of farming experience, ensuring that each participant could reconstruct nearly two decades of production history.

Reconstructing nineteen years of farm records in a setting where most smallholders keep no formal accounts is a formidable challenge. The researchers used retrospective recall techniques, dividing the study period into four sub-periods aligned with successive presidential regimes—a practical aid to memory in a country where agricultural policy shifts with political transitions. Producers identified years of high and low performance within each sub-period, then supplied production details for the remaining years. Where available, farmers supplemented their recollections with accounting books, input purchase invoices, and labor contracts. The final estimation sample included 5,577 observations, reduced from a theoretical maximum of 9,158 by recall limitations, missing variables, and strict data-cleaning procedures that the authors argue should not introduce systematic bias under standard missing-data assumptions.

Methodologically, the study employed a parametric translog cost frontier approach, a flexible second-order approximation of the true cost function that imposes few prior restrictions on the underlying technology. The framework, following the decomposition methods of Kumbhakar and colleagues, separates productivity growth into three distinct components: technical efficiency change, which measures how close farmers operate to the best-practice frontier; technological change, which captures shifts in that frontier itself; and scale effects, which reflect economies or diseconomies of expanding production. Input prices for fertilizer, labor, and a Laspeyres index of other inputs—including seeds, insecticides, herbicides, and animal or mechanical traction—were normalized by land prices to satisfy linear homogeneity constraints. Crucially, the model accounted for unobserved heterogeneity among farms, distinguishing persistent inefficiency rooted in structural conditions from time-varying inefficiency that fluctuates year to year.

The headline result is modest but meaningful: TFP in Beninese cotton production grew by an average of 1.26 percent per year between 2000 and 2018. Decomposition reveals that this growth was powered almost entirely by technological progress, which advanced at 2.81 percent annually, complemented by a modest gain in technical efficiency of 0.24 percent per year. The technological momentum reflects concrete changes in the field: continuous varietal improvement programs that guarantee quality seed to producers, the gradual replacement of hand tools with animal traction and mechanical power, and intensified use of mineral fertilizers and pesticides for weed control and plant health. The technical efficiency gains, meanwhile, are credited in part to Benin’s dense extension network—nearly all cotton farmers belong to Village Cotton Producers Cooperatives, and extension agents are evaluated on the basis of cotton production performance.

But there is a troubling counterweight. The scale component exerted a negative effect of 1.83 percent per year on TFP, indicating that Beninese cotton farms are operating under decreasing returns to scale. When farmers increased the use of all inputs, output rose proportionally less, driving up unit costs for each additional kilogram of cotton harvested. In plain terms, farms are growing beyond their most efficient size, and unexploited economies of scale represent a pool of unrealized productivity. The authors note similar findings in Chinese agriculture after reforms, and contrast them with European and Finnish dairy farms where scale effects contributed positively to productivity—evidence that lower average costs could be achieved by producing at more optimal scales.

The study also traced how productivity translated, or failed to translate, into the bottom line. Farm profits grew by an average of 1.65 percent per year over the period, but the sources of that growth were largely external rather than internal. Rising cotton prices, which climbed at 3.03 percent annually, were the dominant driver, aided by Benin’s price stabilization mechanism that guarantees a minimum income for producers and shields them from world market volatility. Output quantity growth of 0.89 percent per year and the TFP gains also helped. Working against these gains, input prices rose at 2.91 percent per year, steadily eroding profitability—a pattern the authors note mirrors Kumbhakar and Lien’s findings in Norwegian dairy farming. Profit growth was strongest between 2011 and 2015, driven by favorable price movements, while the earliest sub-period saw profits squeezed by input cost inflation.

One finding stands out for its starkness: the estimated overall cost efficiency of Beninese cotton farms is only about 23.5 percent, combining persistent efficiency of 28 percent with time-varying efficiency of 83.8 percent. Actual production costs remain substantially above the minimum attainable frontier. The authors caution that this does not simply reflect poor management. Rather, the persistent inefficiency component likely captures structural and systemic constraints beyond individual farmers’ control: poor rural infrastructure, high transportation and input transaction costs, imperfect access to mechanization services, climate variability, credit market imperfections, and institutional rigidities. Similar low-efficiency findings across West Africa support this interpretation, with prior research showing that institutional environments—particularly access to credit, inputs, and marketing channels—significantly shape producer performance.

The researchers confirmed both of their formal hypotheses: technical efficiency gains contributed positively to TFP growth, and rising cotton prices positively influenced farm profits. Yet they are candid about the study’s limitations. Retrospective data collection risks recall bias and measurement error; the sample of experienced, older producers may overestimate efficiency; the cost frontier framework does not address potential endogeneity, including the possibility that government-set cotton prices are not truly exogenous; and the assumption of full allocative efficiency—that farmers use optimal input combinations—may not hold in practice. Future work using profit frontier models could test how sensitive these conclusions are to those assumptions.

For policymakers, the recommendations are clear. The authors call for strengthening producer capacities through training, promoting technological innovations such as mechanization and pest-resistant varieties, and implementing incentive-compatible price policies that support optimal input use. The stagnation of TFP growth in the final sub-period of the study, driven by losses in technical efficiency and negative scale effects, serves as a warning that past gains are not guaranteed to persist. With nearly three million livelihoods tethered to the cotton plant, Benin’s experience offers a broader lesson for agricultural development across West Africa: productivity growth is possible even under structural constraints, but converting it into durable farmer prosperity requires tackling the institutions, infrastructure, and input markets that determine whether efficiency gains reach the farm gate.

Subject of Research: Total factor productivity and profit growth in cotton production in Benin, West Africa

Article Title: Total factor productivity and profit growth in cotton production in Benin, West Africa

Article References: Saliou, I. O., Zannou, A., Honlonkou, A. N., & Aoudji, A. K. N. (2026). Total factor productivity and profit growth in cotton production in Benin, West Africa. Discover Agriculture, 4(1), Article 285. https://doi.org/10.1007/s44279-026-00749-3

Image Credits: AI Generated

DOI: 10.1007/s44279-026-00749-3

Keywords: Benin, cotton production, total factor productivity, technical efficiency, profit growth, cost frontier, agricultural economics, West Africa, technological change, economies of scale, farm profitability, agricultural policy