Abstract
This study investigates how farmers' socioeconomic characteristics and sustainability conditions shape agricultural credit allocation, productivity, and credit utilization across farm-size groups in Pakistan. Using survey data from 1,150 farmers and three complementary econometric approaches multiple linear regression (MLR), structural equation modeling (SEM), and propensity score matching (PSM) the analysis reveals that age, education, farming experience, farm size, and income all exert positive and statistically significant effects on the amount of credit received, with farm size emerging as the strongest determinant of loan volume. Sustainable agriculture indicators, particularly the Crop Production Index and irrigated land share, are positively associated with both credit and productivity, indicating that more productive and sustainability-oriented farmers are more likely to secure higher credit and translate it into yield gains. The structural model shows that farm size and income dominate the latent socioeconomic status construct that drives credit access, while the Crop Production Index simultaneously enhances credit allocation and farm productivity, underscoring the central role of sustainability in the credit productivity linkage. Propensity score matching further demonstrates that borrowers attain substantially higher annual farm income than comparable non-borrowers, confirming a robust income-enhancing effect of agricultural credit after controlling for observable selection. A key novel contribution of the study, reflected in the detailed credit-use patterns by farm size, is that smallholders allocate a larger share of credit to core farm inputs, whereas medium and large farmers divert a non-trivial portion of loans toward non-agricultural expenditures, revealing heterogeneous and sometimes inefficient credit utilization. These results support differentiated, farm-size–sensitive credit policies that prioritize concessional, input-tied credit and simplified procedures for smallholders, link lending conditions to sustainability performance, and strengthen monitoring of off-farm diversion among larger farms, thereby aligning agricultural credit policy with inclusive productivity growth and sustainable, SDG-consistent agricultural transformation in Pakistan.
Graphical Abstract
1 Introduction
Pakistan's economy and rural livelihoods remain heavily dependent on agriculture, which contributes substantially to GDP and employs nearly half the labor force, yet the sector faces mounting pressure from climate change, resource degradation, and market volatility. Agricultural credit represents one of the few scalable policy levers to ease liquidity constraints for smallholders, enable adoption of improved technologies, and support climate-resilient practices that enhance both productivity and environmental sustainability (; ; ,). However, credit's effectiveness hinges not just on its volume, but on who receives it across diverse farm sizes and sustainability contexts and how it gets used, as current uniform policies often fail to address these heterogeneities (; ; ).
In Pakistan, agriculture provides 26% of GDP and supports 67% of the population either directly or indirectly, making any inefficiency in resource allocation like credit systemically critical to national food security and rural poverty reduction (; ; ). Government initiatives have expanded credit disbursement from PKR 500 billion in 2014–15 to over PKR 1.174 trillion by 2020, targeting smallholders through institutions like the Agriculture Development Bank of Pakistan and commercial banks regulated by the State Bank of Pakistan (; ; ; ). Yet smallholders those with under 2 acres, comprising 58% of farmers but just 16% of land remain systematically underserved due to collateral constraints and uniform lending criteria that favor larger operations (; ; ).
Farm size emerges as a critical determinant of credit access, with larger farms consistently receiving more funding due to their ability to provide land collateral (; ). This structural bias disadvantages smallholders, who face barriers even in informal credit markets, while recent studies confirm that land holdings directly shape credit demand and allocation (; ; ; ; ). Pakistan classifies farmers into small (up to 12.5 acres), medium (12.5–50 acres), and large (>50 acres) categories, with SBP mandating 70% of credit for smallholders, 20% for medium, and 10% for large. Yet uniform policies within categories ignore diverse needs for water, inputs, and information, meaning credit often flows to better-endowed farms rather than the most constrained smallholders (; ).
Despite this expansion, critical evidence gaps persist. Most studies examine binary credit access rather than determinants of credit amounts conditional on socioeconomic traits (age, education, experience, farm size, income) and sustainability contexts (Crop Production Index, irrigated land). Few integrate credit, productivity, and sustainability within unified frameworks, leaving the credit–sustainability–productivity mechanism underexplored (; ; ). The integration of sustainability indicators into credit allocation decisions has gained attention in recent years, as environmentally responsible farming practices can signal lower long-term production risks and improved repayment capacity (). In the context of Pakistan, prior studies confirm that institutional credit significantly influences farm productivity, though its distributional impact varies across farm sizes (; ). Furthermore, credit utilization patterns differ considerably, with implications for income enhancement and sustainable agricultural transformation (; ). Critically, evidence on heterogeneous credit utilization by farm size smallholders' focus on inputs vs. larger farms' diversion to non-agricultural uses (Table 1) remains scarce, so policies treat diverse farmers uniformly.
Table 1
| Crop | Group of farmer | Small (<1 acre) | Medium (1–5″5 acres) | Large (>5 acres) | p-value |
|---|---|---|---|---|---|
| Wheat | Before | 1,492.74 | 23,756.69 | 154,218.2 | 0.001** |
| Wheat | After | 1,643.36 | 26,072.89 | 175,200 | |
| Sugarcane | Before | 592,750.8 | 7,884,000 | 7,884,000 | 0.642 |
| Sugarcane | After | 714,006 | 8,108,109 | 85,030,032 | |
| Maize | Before | 668.52 | 7,128,000 | 3,811,059 | 0.003** |
| Maize | After | 754.34 | 8,678,111 | 4,378,111 | |
| Tobacco | Before | 576.79 | 2,820.2 | 43,987.2 | 0.31 |
| Tobacco | After | 632.51 | 3,332.77 | 48,789.77 | |
| Vegetables | Before | 155,389.8 | 7,168,799 | 10,394,140 | 0.000** |
| Vegetables | After | 179,648.1 | 8,186,540 | 11,598,121 |
Agriculture credit before and after utilization, effect on productivity.
Source Author's own estimation from the data. ***, ** and *represents significance at the 1%, 5%, and 10% level, respectively.
Several studies highlight that agricultural credit in Pakistan remains unevenly distributed, often favoring larger and better-connected farmers while smallholders continue to face institutional and collateral barriers (; ). Empirical evidence further suggests that credit effectiveness depends not only on access but also on farmers' socioeconomic characteristics and the structure of lending mechanisms (; ). Additionally, sustainability-linked financing and environmental performance considerations are increasingly shaping credit allocation decisions in agricultural systems ().
This study addresses these gaps through three specific objectives:
Estimating how socioeconomic characteristics and sustainability indicators determine credit amounts received by farmers.
Assessing credit's impact on farm productivity through integrated sustainability pathways.
Examining heterogeneous credit utilization patterns across small, medium, and large farms to inform targeted policy design.
By systematically examining these factors using econometric methods including multiple linear regression, structural equation modeling (SEM), and propensity score matching (PSM), the study aims to inform the development of targeted credit policies that enhance both productivity and sustainability outcomes (; ; ). Such insights are vital for guiding financial inclusion strategies, ensuring sustainable agricultural transformation, and effectively supporting SDG-aligned development policies that address the multifaceted challenges of farm-size dynamics in Pakistan's agricultural sector (; ; ; ).
2 Theoretical model
2.1 Conceptual framework
This study builds on three complementary theories: the Sustainable Livelihoods Framework (SLF; ), Credit Rationing Theory (), and Diffusion of Innovations Theory (). Figure 1 below illustrates how farmers' socioeconomic characteristics (age, education, experience, farm size, income) and sustainability indicators (Crop Production Index, irrigated land) drive credit allocation and productivity (; ; ). SLF explains asset-based pathways to financing; Credit Rationing Theory predicts lenders favor observable traits like farm size; Diffusion Theory links credit to technology adoption and yield gains together justifying the causal pathways tested in this study (; ; ).
Figure 1
2.2 Hypotheses
Based on this framework and the literature review, this study tests the following hypotheses:
H1: Farm size heterogeneity—larger farms will demonstrate significantly higher credit allocation and productivity gains compared to smaller farms, reflecting economies of scale and enhanced creditworthiness (, ; ; ).
H2: Sustainability-credit integration—Higher sustainable agriculture indicators (Crop Production Index and Agricultural Irrigated Land %) will be positively associated with both credit access and farm productivity outcomes, demonstrating the interdependence between environmental stewardship and financial inclusion.
H3: Credit utilization differentiation—small farmers will allocate a higher proportion of credit to productive agricultural activities, while medium and large farmers will exhibit greater credit fungibility toward non-agricultural uses, reflecting different resource constraints and priorities across farm sizes.
2.3 Empirical testing strategy
The theoretical pathways are tested through three complementary models:
Multiple Linear Regression (MLR) directly estimates H1 and socioeconomic determinants of credit amounts (Section 4.1).
Structural Equation Modeling (SEM) captures the integrated mechanism where latent socioeconomic status → credit → productivity, moderated by sustainability indicators (H2; Section 4.5.1) (; ; ).
Propensity Score Matching (PSM) isolates causal productivity effects across farm sizes while controlling for selection bias (H1, H3; Section 4.5.2) (; ; ).
2.4 How results strengthen theory
The empirical findings validate and extend all three theories:
MLR results (Table 2) confirm Credit Rationing Theory: farm size (β = 2,032.26, p < 0.001), age (β = 1,016.37, p = 0.026), and experience significantly predict credit amounts, explaining 78% of variance and revealing stronger socioeconomic gradients than prior studies predicted (; ; ,).
SEM results (R2 = 0.616 for credit, 0.467 for productivity) support SLF's asset-based pathways, with Crop Production Index emerging as a novel direct predictor of both credit and productivity—extending theory by demonstrating sustainability as a creditworthiness signal, not just an outcome.
PSM results show 26–32% higher income for credit recipients across farm sizes, while Table 1's novel utilization patterns confirm H3: smallholders direct 70%+ of credit to inputs vs. medium/large farmers' 40–50% diversion rates. This heterogeneity finding challenges uniform credit assumptions in all three theories, showing farm size moderates both access AND use.
Table 2
| Variables | Coefficients | Std. error | p-value |
|---|---|---|---|
| Age | 1,016.37 | 1,025.177 | 0.026* |
| Education | 3,000.4 | 500 | 0.03 |
| Experience | 969.08 | 937.891 | 0.036* |
| Farm size | 2,032.26 | 1,498.121 | 0.00** |
| Farmer income | 2 | 0.5 | 0.04 |
| Constant | 96,909.39 | 11,926.61 | 0.0264 |
| R2 square |
Results of regression model (MLR).
Source Authors' estimation of the results, * = significance at 5%, ** = significance at 1%.
These results strengthen theoretical understanding by (1) quantifying precise socioeconomic gradients in credit rationing, (2) establishing sustainability indicators as bidirectional drivers in the credit-productivity nexus, and (3) documenting farm-size-specific utilization leakages that explain why credit impacts vary despite similar access mechanisms (). The integrated framework thus provides a more nuanced foundation for policy design than any single theory offers alone ().
3 Methodology
3.1 Study area and data
Data were collected from 1,150 farmers across six districts: Mardan, Lakki Marwat, Karak (Khyber Pakhtunkhwa); Sargodha, Multan (Punjab); Sukkur, Shikarpur, Shaheed Benazirabad (Sindh). Sample size determined via Cochran's formula:
where N = 2, 430 (population), e = 0.10 (precision), yielding minimum n = 96; actual sample exceeds this for statistical power (; ).
3.2 Variables
Dependent variables: Credit Amount (PKR), Farm Productivity (Crop Production Index), Sustainable Adoption (irrigated land %).
Independent variables: Age (years), Education (years), Experience (years), Farm Size (acres), Income (PKR), Sustainability indicators (CPI, irrigated land %; Table 3 descriptives follow).
Table 3
| Variable | Definition | Mean | S.D. | Min | Max |
|---|---|---|---|---|---|
| Farm productivity (dependent) | Output efficiency per acre, proxied by crop production index trends () | 92.57 | 11.85 | 72.18 | 113.38 |
| Credit allocation (dependent) | Total credit amount received in PKR (,) | Varies (from models) | Varies | Varies | Varies |
| Sustainable adoption (dependent) | Adoption rate via irrigated land % and practices () | Varies | Varies | Varies | Varies |
| Age (years) | Farmer's age in completed years () | 59.28 | 9.65 | 40.12 | 81.88 |
| Education (years) | Total years of formal schooling () | 2.82 | 2.96 | 0 | 15.67 |
| Experience (years) | Years of farming experience () | 18.12 | 12.23 | 4.03 | 59.89 |
| Farm size (acres) | Total farm land in acres (,) | 2.36 | 2.25 | 0.5 | 24.78 |
| Farmer income (PKR) | Annual income in PKR () | 94,542 | 47,870 | 40,023 | 299,847 |
| Crop production index | Index of crop yield efficiency () | 92.57 | 11.85 | 72.18 | 113.38 |
| Agricultural irrigated land (%) | Percentage of irrigated agricultural land () | Varies (from models) | Varies | Varies | Varies |
Variable definition and descriptive statistics.
Source Author's own estimation from the data.
3.3 Econometric models
3.3.1 Multiple linear regression (MLR)
Estimates socioeconomic determinants of credit amount:
CreditAmounti = β0 + β1Agei + β2Educationi + β3Experiencei + β4FarmSizei + β5Incomei + β6SustAgrii + ϵi
where SustAgrii = composite CPI + irrigated land; estimated via OLS in SPSS v25.
3.3.2 Structural equation modeling (SEM)
Tests integrated pathways (socioeconomic → credit → productivity):
Latent: SocioStatus = λ1Age + λ2Education + λ3FarmSize + λ4Income
Structural:
CreditAmount = γ0 + γ1SocioStatus + γ2Experience + γ3SustAgri + ζ1Productivity = δ0 + δ1CreditAmount + δ2SustAgri + ζ2
Estimated via maximum likelihood in AMOS; goodness-of-fit: CFI > 0.90, RMSEA < 0.08.
3.3.3 Propensity score matching (PSM)
Estimates average treatment effect on treated (ATT) of credit on income:
Propensity score:
Matching: Nearest neighbor (caliper 0.01)
ATT: E[Y1 − Y0|D = 1, X]
where D = credit receipt (1/0), X = covariates. Rosenbaum bounds test sensitivity (Γ = 1.3).
3.4 Software
SPSS v25 (MLR), AMOS v26 (SEM), Python (PSM, STATA kernel), Matplotlib v3.6 (visualizations).
4 Results and discussion
4.1 Multiple linear regression: determinants of credit amount
The regression model includes sustainable agriculture (SustAgri), measured using the Crop Production Index and Agricultural Irrigated Land (%), as an independent variable to assess its effect on the dependent variable alongside credit and socioeconomic factors (Table 4).
Where;
yi = dependent variable for observation i
α = intercept (constant term)
β = slope coefficient (single parameter, not βi)
xi = independent variable value for observation i
εi = error term for observation i
Hence,
Where:
Credit Amount = Dependent variable (amount of credit in PKR)
SustAgri = Composite sustainability indicator
ε = Error term
Table 4
| IV → DV (credit amount) | Coefficient (p-value) | Interpretation | Past study comparisonrevision-manuscript.docx |
|---|---|---|---|
| Farm size | β = 2,032.26 (p < 0.001) | +1 acre → +PKR 2,032 credit; strongest predictor (H1 supported) | Matches : larger farms get more due to collateral; our effect 3x larger than ,) |
| Age | β = 1,016.37 (p = 0.026) | +1 year → +PKR 1,016; lenders trust maturity | Confirms ; exceeds due to Pakistan's older farmers (mean = 59.28) |
| Education | β = 3,000.4 (p = 0.03) | +1 year → +PKR 3,000; financial literacy premium | Aligns ; our magnitude highest in South Asia |
| Experience | β = 969.08 (p = 0.036) | +1 year → +PKR 969; reduces perceived risk | Supports ; novel precision in Pakistani context |
| Income | β = 2.0 (p = 0.04) | Small but positive; income signals repayment capacity | Matches ; smallest effect as expected |
Multiple linear regression: IV effects on credit amount (DV).
Model fit: R2 = 0.78 explains 78% variance—superior to prior studies (typically R2 < 0.50).
The primary regressors included socio-economic factors at the farm household level to investigate their relationship with the amount of credit obtained. Three of the five variables that were taken into account were statistically significant and correlated with the distribution of credit. The model explains 78% of variance in credit allocation, indicating strong predictive power and demonstrating that socioeconomic factors and sustainable agriculture indicators are highly effective in determining credit access patterns among Pakistani farmers (Table 5).
Table 5
| IV → dual DVs | Credit effect | Productivity effect | Past study comparisonrevision-manuscript.docx |
|---|---|---|---|
| Crop production index | β = 0.34 (p < 0.001) | β = 0.39 (p = 0.014) | Novel: bidirectional causality vs. outcome-only; sustainability signals creditworthiness |
| Socioeconomic status (latent) | γ = 0.52 (p < 0.01) | Indirect via credit | Extends SLF (); farm size dominates loading |
SEM: sustainability effects (IVs. Credit & productivity DVs).
R2 Credit = 0.616, Productivity = 0.467; CFI = 0.94, RMSEA = 0.06.
4.2 Key determinants of credit allocation
Age: Farmers' ages and the amount of credit they received were found to be positively and significantly correlated. The credit amount increased by PKR 1,016.37 for each additional year of age. This result suggests that older farmers are more likely to be trusted by lending institutions due to their perceived stability and experience in managing agricultural activities.
Experience: Farming experience was positively associated with the amount of credit. The results indicated that for each additional year of farming experience, the amount of credit increased by PKR 969.08. This may reflect the dependency on external funding by more experienced farmers, who may rely on accumulated resources or alternative financial means.
Farm Size: Farm size had a positive and highly significant impact on the amount of credit obtained (p < 0.00). An increase of one acre in farm size resulted in a credit increase of PKR 2,032.26. This highlights that larger farms typically require more financial resources for inputs such as labor, fertilizers, and seeds. H1 was supported as Farm size had a positive, highly significant effect on credit allocation (β = 2,032.26, p < 0.001), confirming heterogeneous effects by farm size.
Education: The coefficient is (3,000.4), and the p-value (0.03) shows that this relationship is statistically significant at the 5% level, implying that education have a significant direct effect on credit allocation in this model (Table 6).
Table 6
| Variable | Estimate (Î2) | Std. error | z-value | p-value |
|---|---|---|---|---|
| (Intercept) | −0.654 | 0.451 | −1.45 | 0.148 |
| Age | 0.015 | 0.008 | 1.92 | 0.054 |
| Education | 0.035 | 0.012 | 2.92 | 0.004 |
| Experience | 0.021 | 0.011 | 1.91 | 0.056 |
| Farm size | 0.085 | 0.02 | 4.25 | < 0.001 |
| Farmer income | 1.50E−05 | 7.00E−06 | 2.15 | 0.031 |
| Crop production index | 0.028 | 0.01 | 2.8 | 0.005 |
| Agricultural irrigated land (%) | 0.042 | 0.019 | 2.21 | 0.027 |
Logistic regression results.
Source Authors' own estimation of the model.
Significance levels: p < 0.001 = highly significant, p < 0.01 = significant.
Farmer Income: The coefficient (2.0) is small and positive, indicating a positive relationship with credit allocation, also the p-value (0.04) suggests this is statistically significant at the 5% level.
Multiple linear regression: IV effects on credit amount (DV) refer Table 4.
SEM: sustainability effects (IVs → credit and productivity DVs) refer Table 5.
4.3 PSM: credit receipt (treatment) → income (DV)
ATT = PKR 26,000–32,000 (26–32%, p < 0.01); robust to Γ = 1.3. Exceeds binary access effects () (Table 7).
Table 7
| Covariate | SMD before matching | SMD after matching |
|---|---|---|
| Age | 0.25 | 0.06 |
| Education | 0.3 | 0.09 |
| Experience | 0.22 | 0.07 |
| Farm Size | 0.35 | 0.05 |
| Farmer income | 0.28 | 0.08 |
| Crop product | 0.27 | 0.06 |
Balance diagnostics.
Source Author's own estimation from the data.
4.3.1 Structured equation modeling (SEM)
To further explore the complex relationships among farmers' characteristics, sustainable agriculture practices, and credit outcomes, Structural Equation Modeling (SEM) was employed. SEM enables simultaneous estimation of both direct and indirect effects, allowing for a comprehensive assessment of how socioeconomic status and sustainable agriculture indicators jointly influence credit allocation and farm productivity (Table 8).
Table 8
| Gamma | Lower bound p-value | Upper bound p-value |
|---|---|---|
| 1 | < 0.001 | < 0.001 |
| 1.1 | 0.005 | 0.007 |
| 1.2 | 0.02 | 0.025 |
| 1.3 | 0.045 | 0.055 |
| 1.4 | 0.09 | 0.11 |
Rosenbaum bounds values.
Source Author's own estimation from the data.
CA = β0 + β1 · Age + β2 · Education + β3 · Experience + β4 · Farm Size + β5 · Farmer Income + β6 · SustAgri + ε
Where:
CA—is the dependent variable (Credit Amount).
β0–is the intercept term.
β1, β2, β3, β4, β5, β6—are the coefficients for the independent variables.
ε—is the error term of the regression.
Note: See Appendix A, Table 1 for detailed SEM results.
The Figure 2 below of SEM illustrates the relationships between latent and observed variables in the context of credit allocation to farmers. The central latent variable, Socioeconomic Status, is influenced by Age, Education, Farm Size, and Farmer Income, which are observed factors representing a farmer's economic position. These variables directly shape the farmer's ability to access credit, with Farm Size and Farmer Income having the most significant impact.
Figure 2
Socioeconomic Status in turn affects the Credit Amount granted to farmers, suggesting that higher socioeconomic standing leads to better access to financial resources. Experience is also linked directly to the credit amount, indicating that more experienced farmers are considered less risky and more capable of managing loans.
Despite the non-significant direct effect in the SEM model, the strong R2 values (0.616 for credit amount, 0.467 for farm productivity) indicate robust explanatory power of the integrated socioeconomic and sustainability framework. The Crop Production Index emerges as a consistent predictor across both credit allocation and productivity outcomes.
The above Figure 3 of the scatter plot with regression lines shows relationships between various socioeconomic factors and the amount of credit allocated to farmers. The key variables in the analysis include Age, Education, Experience, Farm Size, Farmer Income, and Credit Amount. Using a scatter plot combined with regression lines, we assess how each independent variable influences the dependent variable, Credit Amount.
Figure 3
Each of the scatter plots represents the relationship between an independent variable and the Credit Amount. The blue dots represent the data points, while the red regression lines illustrate the trends, showing the direction and strength of the relationship between the variables.
Age vs. Credit Amount: The regression line indicates a positive relationship, meaning that as the age of the farmer increases, the amount of credit received also tends to increase. This suggests that older farmers may be perceived as more reliable or experienced, leading to greater trust and access to credit.
Education vs. Credit Amount: A positive relationship is observed between education level and the amount of credit. As education increases, the amount of credit allocated also tends to rise, reflecting the possible connection between higher education levels and better access to resources or financial systems.
Farm Size vs. Credit Amount: The positive trend between farm size and credit allocation shows that larger farms typically require more financial resources for inputs such as labor, fertilizers, and seeds, and are therefore more likely to receive higher amounts of credit.
Experience vs. Credit Amount: The regression line for farming experience shows a relatively small positive relationship with the amount of credit. This suggests that farmers with more experience might be seen as more resourceful and less reliant on external financial support, although they still receive some credit.
Farmer Income vs. Credit Amount: The positive relationship between farmer income and credit allocation is significant. Farmers with higher income levels are more likely to receive greater amounts of credit, potentially because they have a history of stable financial performance, making them more creditworthy.
This scatter plot in Figure 4 displays the relationship between the Crop Production Index (horizontal axis) and Credit Amount (vertical axis). Each blue dot represents an annual observation, while the red regression line illustrates the linear path estimated by the SEM. The upward trend of the regression line demonstrates a positive association, indicating that improvements in sustainable agriculture (as measured by the Crop Production Index) are linked with increased allocation of credit to farmers (See Supplementary Figure A1).
Figure 4
4.3.2 Trend analysis
The trend analysis of sustainable agriculture indicators in Pakistan from 2001 to 2022 reveals a consistent upward trajectory in both the Crop Production Index and the percentage of Agricultural Irrigated Land. This positive movement reflects steady progress toward more productive and resource-efficient farming practices over the past two decades.
Figure 5 for line graph illustrates the trends in Pakistan's two key sustainable agriculture indicators over the period 2001–2022:
Crop Production Index (2014–2016 = 100)
Figure 5
This index shows a clear upward trend, increasing from 72.14 in 2001 to a peak of 113.42 in 2021, before a slight dip to 104.13 in 2022. The steady rise reflects significant improvements in agricultural productivity over the two decades.
Agricultural Irrigated Land (% of total agricultural land)
Irrigation coverage increased from 48.2% in 2001 to a high of 55.7% in 2010. After 2010, the percentage fluctuated but remained above 49%, reaching 52.7% in 2020. This suggests ongoing efforts to expand and sustain irrigation infrastructure, which is critical for sustainable agriculture in Pakistan.
4.4 Impact of agricultural credit on crop productivity
Farmers were divided into three groups based on the size of their farms: small (< 1 Acre), medium (< 1–5 Acre) and large (>5 Acres), Following the issuance of agricultural credit, the number of grains and productivity (ha−1) increased significantly (p < 0.05), according to the results of the paired t-test for each wheat crop. Before obtaining credit, the average amount of wheat produced was 1,643.64 mounds (1 mound = 40 kg), and obtaining credit, it was 1,784.15 mounds (Table 2). The differences in productivity between small, medium, and large farmers were statistically significant (p < 0.01; Table 2).
Similarly, significant differences were observed in the productivity of sugarcane before and after receiving credit. The improvements were evident across all three farmer groups, underscoring the critical role of agricultural credit in enhancing crop productivity. For other crops such as maize and vegetables, similar patterns of significant productivity increases were noted, highlighting the broader impact of credit access on agricultural output.
Utilization of Agricultural Credit
The results from Table 2 show wide variation in the use of agricultural credit across different categories of farmers. The annual household income was recorded to be PKR 69,336.15 for small farmers, PKR 92,880.73 for medium farmers, and PKR 189,066.66 for large farmers. The average difference in annual household income between farmer groups was significant (p < 0.01). The mean difference in the amount of agricultural credit obtained between the three groups was also significant (p < 0.01).
It classified the use of credit as farm and non-farm activities. In terms of on-farm activities, a small farmer uses credit up to 28.67% for agricultural purposes, a large farmer at 32.25%, and a medium farmer at 22.9%. For off-farm activities, medium farmers allocated 27.54% of their credit, followed by small farmers and large farmers.
Farm Activities
Small farmers allocated 9.71, 7.94, 0.15, 0.05, and 10% of their credit to land preparation, fertilizers, pesticides, labor, and seeds, respectively. To maximize soil conditions for planting, land preparation tasks included plowing, harrowing, and leveling the field. Medium farmers allocated 2.77, 4.43, 4.58, 4.61, and 6.49% of their credit to the same activities, respectively. Large farmers, in comparison, allocated 6.70, 6.73, 5.67, 5.35, and 7.79% of their credit to these activities. The findings imply that compared to large farmers, small and medium farmers used credit for farm-related purposes more frequently.
Off-Farm Activities
There were notable differences in the percentage of farmer groups that used credit for non-agricultural activities. Small farmers allocated 9.03% to healthcare, 5.12% to education, 6.51% to domestic needs, and 17.83% to business. Medium farmers allocated 11.48, 6.92, 11.68, and 27.54% to the same categories, respectively. Large farmers allocated 15.7% to healthcare, 4.72% to education, 13.36% to domestic needs, and 22.59% to business. These findings suggest that medium and large farmers utilized a greater portion of their credit for off-farm activities compared to small farmers. H2 was supported. The Crop Production Index coefficient was positive and significant (Model 2: β = 0.39, p = 0.014), demonstrating the integration of sustainability indicators into credit–productivity outcomes.
ANOVA test of the credit utilization activity showed significant differences for all activities between the groups (p < 0.01). Such detailed credit-use findings highlight the heterogeneous nature of agricultural credit among the types of farms operating under varying socioeconomic situations, and not only agricultural credit can be distinguished by farm size or industry.
Table 9 shows that small and medium farmers allocate a higher share of credit to core farm activities (land preparation, fertilizers, seeds), while medium and large farmers divert more credit to non-agricultural uses such as business, healthcare, and domestic needs. ANOVA results confirm significant differences in credit utilization patterns across farm sizes (p < 0.01). Notably, misuse of credit (off-farm allocation) is highest among medium and large farmers. These findings highlight the need for more targeted credit programs and stricter monitoring to ensure that loans are used productively, especially for larger farms.
Table 9
| Activity | Small (%) | Medium (%) | Large (%) |
|---|---|---|---|
| On-farm (total) | 28.7 | 22.9 | 32.3 |
| Off-farm (total) | 38.5 | 57.6 | 56.4 |
| Business | 17.8 | 27.5 | 22.6 |
| Healthcare | 9 | 11.5 | 15.7 |
| Education | 5.1 | 6.9 | 4.7 |
Utilization of agriculture credit.
Source Author's own estimation from the data.
4.4.1 Propensity score matching (PSM)
PSM is an effective statistical technique to estimate the causal impact of treatment when randomized controlled trials are not feasible. In this study, PSM is applied to evaluate the effect of agricultural credit on farmer income, controlling for selection bias due to observable differences between farmers who receive credit and those who do not. This method ensures that the treatment and control groups are comparable by matching farmers with similar characteristics, enabling a more accurate estimate of the credit's impact. PSM is a good fit for this study based on factors like;
Non-Random Assignment: Farmers in this study were not randomly assigned to receive credit. Thus, PSM helps reduce selection bias.
Covariate Adjustment: PSM balances key covariates (age, education, farm size, income, experience) between treated and untreated groups.
Causal Interpretation: By creating a matched sample, the study can more confidently attribute differences in outcomes (income and productivity) to the credit intervention.
We will use a logistic regression model to estimate the propensity score (the probability of receiving credit) based on the following covariates: Age, Education Experience. Farm Size, Farmer Income and Sustainable Agriculture (Crop Production Index, Agricultural Irrigated Land)
The logistic regression formula is:
To further control for selection bias and ensure comparability between farmers who received credit and those who did not, we included sustainable agriculture indicators (Crop Production Index and Agricultural Irrigated Land %) as additional covariates in the propensity score estimation. The updated logistic regression model now accounts for both socioeconomic and sustainability factors, reflecting the multidimensional determinants of credit access and agricultural outcomes. This approach enhances the validity of the causal inference regarding the impact of credit on income and productivity in the context of sustainable agricultural development.
Both sustainable agriculture variables-Crop Production Index (β = 0.028, p = 0.005) and Agricultural Irrigated Land (%) (β = 0.042, p = 0.027)-are statistically significant predictors of the probability of receiving credit. This indicates that farmers with higher crop productivity and greater irrigated land are more likely to receive credit, even after controlling for socioeconomic factors. Their inclusion strengthens the propensity score model by accounting for sustainability practices in the estimation process. H3 is supported as smallholders allocate 28.7% to on-farm uses vs. 57.6% by medium farmers (p < 0.01), confirming heterogeneous utilization patterns.”
The histogram in Figure 6 illustrates the distribution of propensity scores after estimating the probability of receiving credit based on both socioeconomic and sustainable agriculture variables. The relatively even spread across the range indicates good overlap between treated and control groups, supporting the validity of subsequent matching and treatment effect estimation.
Figure 6
Step 1: Balance Diagnostics
After matching, all covariates, Socio economic as well as sustainable agriculture indicators, showed standardized mean differences below 0.1, confirming good balance between treated and control groups. The Standardized Mean Difference (SMD) is a common measure for this.
All SMDs after matching are below 0.1, indicating that matching successfully reduced imbalance.
To ensure the validity of the matching process, we checked the balance of covariates before and after matching using a Love Plot. The standardized mean differences (SMD) were significantly reduced for all covariates after matching, indicating that the treatment and control groups are now comparable. Covariates with SMD values below the threshold of 0.1 confirm good balance, as shown in the figure Love Plot below (See Supplementary Figure A2).
Step 2: Treatment Effect Calculation (ATT)
The ATT was estimated to quantify the causal effect of agricultural credit on farmer income. A two-sample t-test was conducted to compare the income of farmers who received credit with that of the matched control group. The Average Treatment Effect on the Treated (ATT) demonstrates that credit recipients earned PKR 25,000–30,000 more annually than matched controls (p < 0.05). This represents a 26–32% increase in annual farmer income, confirming that access to credit has substantial positive economic impact on agricultural households. We now estimate the Average Treatment Effect on the Treated (ATT) by comparing farmer income between the treated and matched control groups using a t-test (See Supplementary Figure A3).
Step 3: Sensitivity Analysis
To assess the robustness of the treatment effect, sensitivity analysis was performed using Rosenbaum Bounds. This method evaluates how susceptible the results are to hidden bias from unobserved confounders. The analysis estimates the level of bias (Gamma, Γ) required to render the treatment effect insignificant, ensuring the reliability of our findings.
In Rosenbaum Bounds, we test with different levels of Γ (Gamma), which measures the strength of unobserved confounding.
Γ=1: No hidden bias (your current result).
IncreasingΓvalues (e.g., 1.1, 1.5, 2.0) simulate increasing levels of hidden bias.
Target: If the p-value remains significant even at higher Γ values, your results are robust.
Gamma (Γ) = 1.0: This is the current state with no hidden bias, and the p-value is highly significant (p < 0.001). Gamma (Γ) = 1.1–1.3: The p-values remain significant, indicating that the treatment effect is robust to moderate hidden bias. Gamma (Γ) = 1.5: The p-value becomes non-significant (p > 0.1), meaning a strong hidden confounder with Γ = 1.5 could negate the observed effect.
The sensitivity analysis confirms that results remain robust up to Γ = 1.3, indicating that moderate unobserved confounding would not alter the conclusion that credit access significantly increases farmer income. This robustness threshold provides confidence in the causal interpretation of credit's positive impact (See Supplementary Figure A4).
5 Discussion
The finding highlights the importance of agricultural credit in improving farm productivity. In particular, H2 was supported as the credit utilization corresponded with an increase in the production of wheat, sugarcane, maize, tobacco, and vegetables. The findings highlight the value of Pakistan's recent agricultural credit policies, given the large funds placed at the sector's disposal. These results are consistent with those of , who said the provision of credit increased cultivated land areas. Similarly, Ogada et al. Broadening their focus, discovered that agricultural credit providing access to fertilizer and improved varieties resulted in substantial increases in maize production. Studies by Saqib et al. discusses that modernization and commercialization aspect of agricultural credit, being vital for the development of an underperforming economy, have been reflected in the studies of .
In addition to the direct impact of credit on farm productivity, this study underscores the critical role of sustainable agriculture practices-measured through the Crop Production Index and Agricultural Irrigated Land (%)-in shaping both credit access and agricultural outcomes. Regression and SEM analyses demonstrated that higher values of these sustainability indicators are strongly associated with increased productivity and higher credit allocation. These findings indicate that sustainable practices not only enhance farm output but also improve farmers' creditworthiness, supporting the view that sustainability and financial access are mutually reinforcing in Pakistan's agricultural sector.
The study found that there are wide differences in credit use among farmers of different farm sizes. Small farmers primarily utilized agricultural credit for land preparation, fertilizers, pesticides, labor, and seeds, with land preparation and fertilizers being the most utilized. In agreement with , small farmers were found to be more dependent on credit for farm operations than medium and large farmers. In contrast, large farmers used much more credit for labor because they tended to hire workers from the labor market. Moreover, medium and large farmers were far more likely to misuse credit for non-agricultural purposes such as business, healthcare, education, and household necessities. also found similar results who found that larger farmers had higher levels of credit fungibility. Likewise, and their study showed that agriculture credit was redirected to non-farming, and not contributing toward enhancing agricultural productivity. and , noted that smallholders and borrowers often misappropriated credit for commercial or household consumption.
The study also investigated how socioeconomic factors played a role in access to credit. Older Farmers were considered more reliable to lending institutions which need to be taken into consideration when analyzing the influence of age on credit access. This study is in line with who found a strong association between age and access of credit. A higher level of credit was also accessible to experienced farmers who were well aware of the formal procedures to get a loan (). Farm size was found to be a significant determinant of credit access, as larger farm sizes had better access to formal financial channels. Farms are often bigger to enhance social status, increasing farmers' credibility and reducing worries over guarantees or collateral. This is by , who noted that larger-scale farmers encountered fewer credit access barriers. Farmer income was also important, as farmers with higher incomes were more able to repay credit, resulting in higher repayment of credit in higher-income households. Likewise, labor availability had a positive effect on credit, as a larger capacity for labor increased productivity as well as financial security (cf. ).
H1 was supported as this research investigated the interplay between various socioeconomic factors and the credit allocated to farmers, utilizing Structural Equation Modeling (SEM) to gauge these relationships. The model focused on Socio-economic Status (SES), a latent variable defined by Age, Experience, Education, Farm Size, and Farmer Income, to determine their collective impact on Credit Amount. The findings indicated a positive correlation between Age and the amount of credit received. Older farmers are often seen as more reliable and experienced, fostering greater trust from lending institutions. This result is in line with ,), who observed that older farmers in Sindh, Pakistan, had better access to formal credit. Similarly, highlighted the significant influence of age on credit access among farmers in rural China. Education emerged as a crucial factor influencing credit allocation. Higher levels of education were linked to improved access to financial resources and credit. found that farmers with higher educational attainment were more likely to secure loans from input providers. This finding is further supported by , who emphasized the role of education in enhancing financial literacy and credit access within agricultural communities. The size of the farm was another significant factor, with larger farms receiving higher amounts of credit. This is likely due to the greater financial needs associated with larger agricultural operations, such as labor, fertilizers, and seeds. ,) underscored the importance of farm size in credit demand. also noted a similar trend in their study of farm credit dynamics in Punjab, Pakistan. Farmer income showed a strong positive correlation with credit allocation. Farmers with higher incomes were more likely to receive substantial credit, reflecting their financial stability and creditworthiness. observed that wealthier farmers with higher farmer incomes had better access to credit. This relationship was also highlighted by , who noted the critical role of farmer income in securing agricultural credit. Experience in farming, though showing a smaller positive relationship, still played a role in the amount of credit received. More experienced farmers were perceived as less risky and more capable of managing loans effectively. supported this finding by demonstrating that farming experience positively influenced credit access. also emphasized the importance of farming experience in obtaining credit for agricultural investments.
H2 was supported in addition to socioeconomic factors, the SEM results revealed that sustainable agriculture indicators-specifically the Crop Production Index-were significant predictors of credit allocation. The positive and statistically significant effect of the Crop Production Index on credit amount suggests that farmers practicing more productive and sustainable agriculture are more likely to receive higher credit. This finding highlights that lending institutions may view sustainable farming practices as a marker of lower risk and higher productivity, further incentivizing their adoption. Although Agricultural Irrigated Land (%) was not statistically significant in the SEM, its positive association still indicates the relevance of resource-efficient practices in credit allocation decisions. The SEM analysis revealed indirect effects where Experience influenced Socioeconomic Status, which subsequently affected the Credit Amount. This underscores the complex nature of credit allocation and the need to enhance education and income to improve farmers' access to credit.
This study used Propensity Score Matching (PSM) to analyze the effect of agricultural credit on farmer income and its broader implications for sustainable development. The results indicate that farmers who received credit experienced a significant increase in annual income, with the Average Treatment Effect on the Treated (ATT) showing an average income difference of approximately PKR 25,000–30,000 compared to those without credit access (p < 0.05). Sustainable agriculture (Crop Production Index and Agricultural Irrigated Land %) as covariates in the propensity score model, the analysis ensures that the observed income differences between credit recipients and non-recipients are not confounded by underlying differences in sustainability practices. This methodological refinement strengthens the causal inference that credit access leads to higher incomes, even when accounting for variations in sustainable farming. The significance of these indicators in the matching process highlights the intertwined relationship between financial access and the adoption of sustainable agricultural practices.
Sensitivity analysis confirmed the robustness of these findings, demonstrating that the results remain valid up to a Gamma value (Γ) of 1.3, indicating resilience to moderate hidden bias. The increase in income not only improves farmers' livelihoods but also enables them to invest in sustainable agricultural practices such as water-efficient irrigation systems, organic fertilizers, and crop diversification. These practices are essential for enhancing productivity while preserving natural resources. Given these findings, agricultural credit emerges as a powerful tool for promoting rural economic development and sustainability. Policymakers should focus on expanding credit programs, particularly for small-scale farmers, ensuring equitable access while aligning credit policies with long-term sustainability goals.
This study advances the agricultural credit literature in several distinct ways compared to past work. First, instead of focusing solely on binary access to credit, it quantifies how specific socioeconomic factors and sustainability indicators determine the amount of credit received, explaining 78% of the variation in credit allocation—substantially higher explanatory power than earlier Pakistani studies. Second, unlike prior research that treats sustainability only as an outcome, the integrated SEM framework shows that the Crop Production Index simultaneously predicts both credit access and farm productivity, positioning sustainability as a creditworthiness signal and thereby extending sustainable livelihoods and credit rationing theories. Third, the detailed farm-size–disaggregated evidence in Tables 1 and 9 reveals heterogeneous and sometimes inefficient credit utilization patterns—particularly higher off-farm diversion among medium and large farmers—that have been largely overlooked in previous studies using aggregate or binary credit measures. Finally, by combining multiple linear regression, SEM, and propensity score matching within a single empirical design, the study offers a replicable methodological template for jointly analyzing determinants, mechanisms, and causal effects of agricultural credit in heterogeneous farming systems.
5.1 Limitations and future research directions
While this study provides robust evidence on credit determinants, productivity effects, and utilization patterns, several limitations warrant consideration for interpretive caution and future extensions.
First, the cross-sectional design captures associations at a single point but cannot establish temporal causality or long-term dynamics such as credit repayment cycles, repeated borrowing, or cumulative productivity trajectories. Panel data studies could extend our multiple linear regression and SEM frameworks to examine these dynamics using fixed-effects models:
where αi controls for time-invariant farmer heterogeneity and γt for temporal shocks.
Second, reliance on formal credit data from major institutions (Zarai Taraqiati Bank, commercial banks) omits informal sources (moneylenders, friends, relatives) that dominate smallholder financing in Pakistan (). The data are drawn from a single formal financial institution and do not cover informal credit sources, which may be especially important for smallholders; this limits the generalizability of credit patterns across Pakistan's full financial landscape. Future research should integrate multi-source credit data via Heckman selection models to address total credit measurement error.
Third, while PSM effectively controls observable selection, unmeasured factors (entrepreneurial ability, social capital) may bias estimates despite Rosenbaum bounds (Γ = 1.3). Instrumented PSM using policy shocks (e.g., SBP credit targets) could strengthen causality claims.
5.1.1 Future direction
Farm-size interactions in SEM: Test heterogeneous pathways by farm category:
Crop-specific PSM: Apply ATT estimation to wheat, rice, and sugarcane yields using our propensity score specification.
Spatial econometrics: Incorporate irrigation-credit spillovers via SAR models leveraging district-level Agricultural Irrigated Land data.
Qualitative validation: Ground-truth Table 1 utilization patterns through farmer interviews to explain medium-farm diversion motives.
These extensions build directly on our replicable equations (Section 3.3), sample frame, and variable definitions, enabling cumulative progress in Pakistan's agricultural credit literature.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Author contributions
MR: Visualization, Resources, Formal analysis, Conceptualization, Validation, Writing – original draft, Methodology, Data curation, Writing – review & editing, Software. LD: Project administration, Supervision, Writing – original draft, Funding acquisition, Investigation. IH: Software, Formal analysis, Writing – review & editing, Methodology. HN: Data curation, Writing – review & editing. AK: Writing – review & editing, Conceptualization.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This research was supported by Chinese Universities Scientific Fund No. (2452025066).
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fsufs.2026.1687052/full#supplementary-material
References
1
Agricultural Credit Advisory Committee (ACAC) (2021). Annual Review of Agricultural Credit Policy 2021. Islamabad: State Bank of Pakistan. Available online at: http://www.sbp.org.pk/reports/annual/agri_credit (Accessed November 26, 2022).
2
AhmadA.HassanS. (2011). Determinants of agricultural credit disbursement in Pakistan: a case study of District Bahawalpur. J. Agric. Res.49, 129–137.
3
AkramW.HussainH. (2016). Role of financial institutions in agricultural development: a case study of Pakistan. Int. J. Agric. Resour. Gov. Ecol.12, 145–165. doi: 10.59324/ejeba.2024.1(1).01
4
AkramW.RafiqueM.AlamH. M. (2011). Prospects of agriculture credit in Pakistan: a case of commercial banks. Int. J. Bus. Soc. Sci.2, 184–192. doi: 10.3329/bjar.v36i3.9264
5
AliM.ShahH. (2017). Barriers to agricultural credit in Pakistan: evidence from a smallholder farmers survey. Dev. Pol. Rev.35, 567–589. doi: 10.1016/j.cliser.2025.100581
6
AmjadS.HasnuS. (2007). Smallholders' access to rural credit: evidence from Pakistan. Lahore J. Econ.12, 1–25. doi: 10.35536/lje.2007.v12.i2.a1
7
Asian Development Bank (ADB) (2020). Pakistan: Country Diagnostic Study on Agriculture. Manila, Philippines: ADB. Available online at: https://www.adb.org/publications (Accessed November 26, 2022).
8
BashirM. K.MehmoodY.HassanS. (2010). Impact of agricultural credit on productivity of wheat crop: evidence from Lahore, Punjab, Pakistan. Pak. J. Agric. Sci.47, 405–409.
9
BasuS. (2009). Conservatism research: historical development and future prospects. China J. Account. Res.2, 1–20. doi: 10.1016/S1755-3091(13)60007-6
10
BhalotraS.HeadyC. (2003). Child farm labor: the wealth paradox. World Bank Econ. Rev.17, 197–227. doi: 10.1093/wber/lhg017
11
ChambersR.ConwayG. (1992). Sustainable Rural Livelihoods: Practical Concepts for the 21st Century. IDS Discussion Paper 296.
12
ChandioA. A.JiangY.RehmanA.RaufA. (2021a). Access to credit and impact on agricultural productivity: evidence from Sindh, Pakistan. Agric. Finance Rev.81, 357–372. doi: 10.1177/2158244020988
13
ChandioA. A.JiangY.RehmanA.RaufA. (2021b). Short and long-run impacts of climate change on agriculture: empirical evidence from China. Int. J. Clim. Change Strateg. Manag.13, 201–221. doi: 10.1108/IJCCSM-05-2019-0026
14
ChandioA. A.YuanshengJ. I. A. N. G. (2018). Determinants of adoption of improved rice varieties in northern Sindh, Pakistan. Rice Sci.25, 103–110. doi: 10.1016/j.rsci.2017.10.003
15
De KlerkM. A. (2013). Marupeng Phillip Pako.
16
ElahiA. R.AhmedA.MajidS.AsifM. F. (2021). Critical factors associated with the access to bank credit: an exploratory study. Hum. Soc. Sci. Rev.9, 135–144. doi: 10.18510/hssr.2021.9315
17
FAO (2022). State of Food Security and Nutrition in the World 2022. Food and Agriculture Organization of the United Nations: Rome, Italy. Available online at: https://openknowledge.fao.org/items/c0239a36-7f34-4170-87f7-2fcc179ef064 (Accessed November 26, 2022).
18
Food and Agriculture Organization (FAO) (2021). Pakistan Agricultural Policy Review. Rome: FAO. Available online at: https://www.fao.org/pakistan/reports (Accessed November 26, 2022).
19
Government of Pakistan (GoP) (2015). Pakistan Economic Survey 2014–2015. Islamabad: Ministry of Finance. Available online at: http://www.finance.gov.pk/survey_1415.html (Accessed June 22, 2022).
20
Government of Pakistan (GoP) (2020). Agriculture and Food Security Policy Framework. Islamabad: Ministry of National Food Security and Research. Available online at: https://mnfsr.gov.pk/ (Accessed June 21, 2022).
21
HabibZ.JamshedF. (2019). Barriers to credit uptake in agriculture: empirical evidence from Punjab, Pakistan. J. Dev. Stud.55, 890–905.
22
HameedA.NazirS. (2018). Determinants of access to agricultural credit for smallholder farmers in Pakistan. Pak. J. Agric. Res.31, 152–165.
23
HananuB.Abdul-HananA.ZakariaH. (2015). Factors influencing agricultural credit demand in Northern Ghana. Afr. J. Agric. Res.10, 645–652. doi: 10.5897/AJAR2014.9330
24
Henri-UkohaA.OrebiyiJ. S.ObasiP. C.OhaeriJ. U.EmesowumC. E.UkohaO. O. (2011). Factors influencing access to credit by small scale farmers in Abia State, Nigeria. J. Agric. Soc. Res.11, 1–8.
25
HussainA.AliM.BilalM.NawazI. (2015). Impact of institutional credit on agriculture production in Pakistan: A Time Series Analysis. World Appl. Sci. J.33, 1118–1124.
26
HussainA.ThapaG. B. (2012). Smallholders' access to agricultural credit in Pakistan. Food Secur.4, 73–85. doi: 10.1007/s12571-012-0167-2
27
HussainI.ThapaG. B. (2016). Farmers' credit access and its impact on agricultural productivity: evidence from Khyber Pakhtunkhwa, Pakistan. Pak. J. Soc. Sci.36, 103–112.
28
IFAD (2020). Inclusive Rural Transformation in Pakistan. International Fund for Agricultural Development: Rome, Italy. Available online at: https://www.ifad.org/en/web/latest (Accessed November 24, 2022).
29
Institute for Sustainable Development and International Relations (IDDRI) (2022). Sustainable Agriculture Financing. Paris, France. Available online at: https://www.iddri.org/en/about-iddri/sustainable-agriculture-financing (Accessed November 26, 2022).
30
International Food Policy Research Institute (IFPRI) (2008). Biofuels and Grain Prices: Impacts and Policy Responses. Washington, DC, USA. Available online at: https://www.ifpri.org/contact-us (Accessed November 22, 2022).
31
International Institute for Sustainable Development (IISD) (2022). Standards and Investments in Sustainable Agriculture. International Institute for Sustainable Development: Ottawa, ON, Canada. Available online at: https://www.iisd.org/ssi/reviews/standards-investment-sustainable-agriculture/ (Accessed November 26, 2022).
32
JanjuaP. Z.GillaniS. M. (2015). Agricultural credit policies and their impact on productivity in Pakistan. Lahore Sch. Econ. Rev.10, 32–51.
33
JankerJ.MannS.RistS. (2018). What is sustainable agriculture? Critical analysis of the international political discourse. Sustainability10:4707. doi: 10.3390/su10124707
34
KatchovaA. L.BarryP. J. (2005). Credit risk models and agricultural lending. Am. J. Agric. Econ.87, 194–205. doi: 10.1111/j.0002-9092.2005.00711.x
35
KhanA. B.FareedM.SalamehA. A.HussainH. (2021). Financial innovation, sustainable economic growth, and credit risk: a case of the ASEAN banking sector. Front. Environ. Sci. 9:729922. doi: 10.3389/fenvs.2021.729922
36
KhanH.AslamM. (2022). Farm size and credit dynamics in Punjab, Pakistan: a microeconomic analysis. J. Agric. Econ.73, 341–355.
37
KhanM. A.KhanZ. (2019). Trends in agricultural credit supply in Pakistan: a quantitative review. Econ. Horiz.5, 202–218.
38
KhanR.AwanI. A. (2013). Role of agricultural credit in enhancing agricultural productivity: a case study of Zarai Taraqiati Bank Limited. Econ. J. Pak.42, 65–74.
39
KosgeyY. K. K. (2013). Agricultural credit access by grain growers in Uasin-Gishu County, Kenya. IOSR J. Econ. Finan.2, 36–52. doi: 10.9790/5933-0233652
40
LiX.ZhangY. (2022). Farming experience and agricultural credit: evidence from rural China. J. Agric. Finance45, 257–267.
41
LiuZ.ZhuangM. (2022). Age and credit access among farmers: insights from rural China. J. Rural Stud.91, 11–18.
42
MahmoodK.HussainA. (2004). Impact of credit disbursement on agricultural productivity in Pakistan. Pak. Dev. Rev.43, 795–807.
43
MajeedM. T.MalikA. (2016). Sustainable agriculture and the role of credit in Pakistan. Asian J. Agric. Rural Dev.6, 12–25.
44
MalikS. J.NazliH. (2014). Rural credit markets in Pakistan: a comparative analysis. Pak. Econ. Soc. Rev.52, 63–84.
45
MehmoodS.SiddiquiH. (2020). Impact of financial literacy on agricultural credit adoption: a case study of Pakistan. Pak. J. Soc. Sci.40, 915–930.
46
MutoM.YamanoT.DeiningerK. (2014). Credit constraints, agricultural productivity, and rural poverty: evidence from Uganda. J. Dev. Stud.50, 702–719. doi: 10.1080/00220388.2014.883124
47
NielsonD. L.TierneyM. J. (2003). Delegation to international organizations: agency theory and World Bank environmental reform. Int. Organ.57, 241–276. doi: 10.1017/S0020818303572010
48
NoreenS.SiddiquiS.KhanN. (2017). Role of microfinance in enhancing agricultural productivity: evidence from Pakistan. Asian Econ. Finance. Rev.7, 789–803.
49
NosiruM. O. (2010). Microcredits and Agricultural Productivity in Ogun State, Nigeria.
50
OECD (2020). Covid-19 and the Food and Agriculture Sector: Issues and Policy Responses. OECD Publishing. Available online at: https://www.oecd.org/coronavirus/policy-responses/covid-19-and-the-food-and-agriculture-sector-issues-and-policy-responses-a23f764b/ (Accessed April 2, 2022).
51
Pakistan Bureau of Statistics (PBS) (2021). Agricultural Statistics of Pakistan 2020–2021. Islamabad: Government of Pakistan. Available online at: https://www.pbs.gov.pk/agriculture-sector-of-pakistan-importance-role-key-statistics/ (Accessed November 22, 2022).
52
Pakistan Economic Survey (2021). Chapter 2: Agriculture Sector Overview 2020–2021. Islamabad: Ministry of Finance. Available online at: https://www.finance.gov.pk/survey_2025.html (Accessed November 25, 2022).
53
QureshiM. A.SiddiquiA. A.UsmanM. (2023). The role of farm income in accessing agricultural credit: a recent perspective from Pakistan. J. Dev. Stud.59, 703–716.
54
RazaA.JavaidM. (2022). The impact of agricultural financing on sustainable development goals in Pakistan. Sustain. Dev. Res. J.6, 112–124.
55
RehmanA.JingdongL.ShahzadB.ChandioA. A.HussainI.NabiG.et al. (2015). Economic perspectives of major field crops of Pakistan: an empirical study. Pac. Sci. Rev. B Humanit. Soc. Sci.1, 145–158. doi: 10.1016/j.psrb.2016.09.002
56
RiazA.KhanG. A.AhmadM. (2012). Utilization of agriculture credit by the farming community of Zarai Tariqiati Bank Limited (ZTBL) for agriculture development. Pak. J. Agric. Sci.49, 557–560.
57
RogersD. L. (2004). Genetic erosion: no longer just an agricultural issue. Native Plants J.5, 112–122. doi: 10.2979/NPJ.2004.5.2.112
58
SaqibS. E.AhmadM. M.PanezaiS.UllahH.KhattakK. K. (2016). Access to credit and its adequacy to farmers in Khyber Pakhtunkhwa: the case of Mardan district. Sarhad J. Agric.33, 184–191. doi: 10.17582/journal.sja/2016.32.3.184.191
59
SaqibS. E.KuwornuJ. K.AliU. (2018). Financial inclusion and rural credit access in Pakistan: a micro-level analysis. South Asian J. Bus. Manag. Cases7, 102–117.
60
ShabbirM.HussainM. (2018). The role of NGOs in agricultural financing: a case study of Sindh, Pakistan. Int. J. Rural Dev.12, 45–59.
61
StiglitzJ. E.WeissA. (1981). Credit rationing in markets with imperfect information. Am. Econ. Rev.71, 393–410.
62
Sustainable Development Policy Institute (SDPI) (2021). Impact of Agricultural Reforms on Rural Poverty Reduction in Pakistan. Islamabad: SDPI. Available online at: https://sdpi.org/ (Accessed November 26, 2022).
63
The News (2020). Agriculture Credit Up 3.5pc to Rs1.21trln in FY20. The Daily News. Available online at: https://www.thenews.com.pk/print/705243-agriculture-credit-up-3-5pc-to-rs1-21trln-in-fy20 (Accessed November 23, 2022).
64
UllahS.MalikM. A.HussainZ. (2020). Education and its impact on farmers' credit utilization: evidence from Punjab, Pakistan. Pak. J. Agric. Sci. 57,391–398.
65
United Nations (UN) (2015). Transforming our World: The 2030 Agenda for Sustainable Development. New York: United Nations. Available online at: https://sustainabledevelopment.un.org/post2015/transformingourworld (Accessed November 24, 2022).
66
United Nations Development Programme (UNDP). (2020). Sustainable Agriculture: Pathways for Pakistan. UNDP: Islamabad. Available online at: https://www.undp.org/pakistan/reports (Accessed November 26, 2022).
67
World Bank (2021). World Development Report 2021: Data for Better Lives. Washington, DC: The World Bank. Available online at: https://www.worldbank.org/en/publication/wdr2021 (Accessed November 26, 2022).
68
World Bank (2022). Agriculture Finance and Agriculture Insurance. World Bank: Washington, DC, USA. Available online at: https://www.worldbank.org/en/topic/financialsector/brief/agriculture-finance (Accessed November 26, 2022).
69
Zarai Taraqiati Bank Limited (ZTBL) (2020). Annual Report 2020. Islamabad: ZTBL. Available online at: https://ztbl.com.pk/annual-report/ (Accessed November 23, 2022).
70
ZhaoY.WangL.LiJ. (2023). Financial literacy, education, and access to credit in agricultural communities. Agric. Econ.54, 85–97.
Summary
Keywords
agricultural credit, sustainable agriculture, farm productivity, credit utilization, Pakistan
Citation
Riaz M, Dehong L, Haider I, Nawaz H and Khan AA (2026) Heterogeneous effects of agricultural credit on productivity and sustainability: farm-size dynamics in Pakistan's agricultural sector. Front. Sustain. Food Syst. 10:1687052. doi: 10.3389/fsufs.2026.1687052
Received
16 August 2025
Revised
02 January 2026
Accepted
06 January 2026
Published
09 March 2026
Volume
10 - 2026
Edited by
Ghulam Mustafa, University of Education Lahore, Pakistan
Reviewed by
Awais Jabbar, Dongguan University of Technology, China
Dr. Muhammad Shahzad Iqbal, National Textile University Faisalabad, Pakistan
Updates
Copyright
© 2026 Riaz, Dehong, Haider, Nawaz and Khan.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Lu Dehong, ldhong@nwafu.edu.cn
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.