Mismeasurement and efficiency estimates: Evidence from smallholder survey data in Africa

Smallholder agriculture in sub-Saharan Africa is commonly characterised by high levels of technical inefficiency. However, much of this characterisation relies on self-reported input and production data, which are prone to systematic measurement error. We show theoretically that non-classical measurement error introduces multiple identification challenges and sources of bias in estimating smallholders' technical inefficiency. We then empirically examine the implications of measurement error for the estimation of technical inefficiency using smallholder farm survey data from Ethiopia, Malawi, Nigeria and Tanzania. We find that measurement error in agricultural input and production data leads to a substantial upward bias in technical inefficiency estimates (by up to 85% for some farmers). Our results suggest that existing estimates of technical efficiency in sub-Saharan Africa may be severe underestimates of smallholders' actual efficiency and what is commonly attributed to farmer inefficiency may be an artefact of mismeasurement in agricultural data. Our results raise questions about the received wisdom on African smallholders' production efficiency and prior estimates of the productivity of agricultural inputs. Improving the measurement of agricultural data can improve our understanding of smallholders' production efficiencies and improve the targeting of productivity-enhancing technologies.     

Irrigation water productivity in Cambodian rice systems

Knowledge of the marginal productivity of water is crucial for decisions about its allocation between uses, which is particularly important in the context of increasing competition for water. Using primary, plot‐level panel data, this article estimates the marginal productivity of water from supplementary irrigation in lowland rice systems in Cambodia, taking into account farmer and plot heterogeneity as well as self‐selection of supplementary irrigation. Our estimates indicate a range of elasticities for rice output with respect to water inputs of between 0.057 and 0.069 for wet season production, substantially lower than previous estimates based on either aggregate or trial data. We discuss the policy implications of these results, in particular with respect to the utility of demand management policies and the challenges they pose to the decentralization of water management to Farmer Water Users Groups.     

Productivity,credit, risk,and the demand for weather index insurance in smallholder agriculture in Ethiopia

The article explores the relationship between fertilizer use and the demand for weather index insurance (WII) among smallholder farmers in Ethiopia. We examine whether fertilizer use is profitable under current smallholder production conditions, whether risk‐related factors affect fertilizer use, and we estimate the returns to inputs in the agricultural production function in the absence of insurance. We then study how these primitives of agricultural production functions relate to insurance demand. The study compares a survey‐based estimate of willingness to pay with actual uptake for the weather insurance, finding the stated and actual demand to be almost completely uncorrelated. While those with high marginal returns to inputs say they would purchase insurance, only those with low marginal returns actually do, consistent with the stated purpose of the product as input insurance. Insurance demand proves to be highly responsive to the existence and amount of randomly allocated insurance vouchers.     

Evaluating spatial and temporal variation in agricultural output elasticities in Turkey

This study evaluates spatial variation in the relationship between agricultural output and input use across Turkey. The potential impact of the national agricultural policy reform introduced in 2001 on the spatial variation in agricultural output elasticities across the country was explored. By applying a spatial production function to the province‐level data in 2000 and 2007, spatial heterogeneity in the variance of provincial total factor productivity and the input factor‐output elasticities was identified across the country. Results show that the disparities in agricultural activities and geographic conditions affected return from input factors. Empirical findings from a spatial spillover model also suggest that changes in the input factor‐output elasticities varied significantly across Turkey between 2000 and 2007, after the policy reform. Results suggest that future policy reform that recognizes regional comparative advantage through understanding the geographic heterogeneity of the agricultural sector is important for enhancing Turkey's agricultural output.     

A note on synthetic data for replication purposes in agricultural economics

Empirical studies in agricultural economics usually involve policy implications. In many cases, such studies rely on proprietary or confidential data that cannot be published along with the article, challenging the replicability and credibility of the results. To overcome this problem, the use of synthetic data—that is, data that do not contain a single unit of the original data—has been proposed. In this note, we illustrate the utility of synthetic data generation methods for replication purposes using a range of methods from agricultural production analysis. More specifically, we compare input elasticities and technical efficiency scores based on different farm-level production data between original data and synthetic data. We generate synthetic data using a non-parametric method of classification and regression trees (CART) and parametric linear regressions. We find synthetic data result in elasticities and technical efficiency distributions that are very similar to the original data, especially when generated with CART, and conclude with implications for the research community.     

ICT and agricultural productivity: evidence from cross-country data

This article carries out agricultural production function estimations, based on data for the period 1995–2000 on 81 countries, to present empirical evidence on the relationship between the adoption of information and communication technology (ICT) and agricultural productivity. It is found that new ICT has a significantly positive impact on agricultural productivity. The evidence suggests that the adoption of modern industrial inputs in agricultural production relies on the information and communication infrastructure. However, the empirical evidence from this study also suggests that new ICT could be a factor for the divergence between countries in terms of overall agricultural productivity. Not only do we find that the ICT adoption levels of the richer countries are much higher than those of the poorer countries, but also that returns from ICT in agricultural production of the richer countries are about two times higher than those of the poorer countries. A plausible explanation for the poorer countries' relatively low productivity elasticity of ICT is the lack of important complementary factors, such as a substantial base of human capital.     

Impact of de-coupling agricultural support on dairy investment and milk production volume in Finland

This study analyses the impacts of de-coupling of agricultural support from production in Finland. A dynamic agricultural sector model, which includes 17 production regions and endogenous investments and technical change, is used in the analysis. Investment in different production techniques is dependent on the relative profitability and the spread of each technique in the population of heterogeneous farms. There are relatively few large farms which use efficient production techniques in Finland. De-coupling weakens the incentive for investment in dairy production and causes a temporary but significant slow down in dairy investments and technical change. Consequently, de-coupling is likely to result in a significant drop in milk and beef production in the next 10-20 year period if no corrective measures are taken in agricultural policy in less-favoured areas such as Finland. However, a slow recovery of investment and output levels are expected in the long run.     

Human health and pesticide use in Sub‐Saharan Africa

While pesticides–such as insecticides, fungicides, and herbicides–are often promoted as inputs that increase agricultural productivity by limiting a range of pre‐harvest losses, their use may have negative human health and labor productivity implications. We explore the relationship between pesticide use and the value of crop output at the plot level and a range of human health outcomes at the household level using large‐scale, nationally representative panel survey data from four Sub‐Saharan African countries where more than 10% of main season cultivators use pesticides. We find that pesticide use is strongly correlated with increased value of harvest, but is also correlated with higher costs associated with human illness, including increased health expenditures and time lost from work due to sickness in the recent past. We take these results as suggestive that the findings of more targeted studies are indeed generalizable beyond their original, purposively chosen samples.     

Productivity and farm size in Australian agriculture: reinvestigating the returns to scale

A positive relationship between farm size and farm productivity is often considered to be largely due to increasing returns to scale in farm production. However, using farm‐level data for the Australian broadacre industry, we found that constant or mildly decreasing returns to scale is the more typical scenario. In this study, the marginal returns to various farm inputs are compared across farms with different sizes. We found that large farms achieved higher productivity by changing production technology rather than increasing scale alone. The results highlight the disparity between ‘returns to scale’ and ‘returns to size’ in the industry, suggesting that productivity improvement among smaller farms can be made through increasing their ability to access advanced technologies, rather than simply expanding their scale.     

Measuring agricultural total factor productivity in China: pattern and drivers over the period of 1978‐2016

The Chinese agricultural sector has experienced a substantial increase in total output since dramatic reforms were introduced in 1978. This paper uses the index method to measure agricultural total factor productivity (TFP) for China’s crop and livestock industries, based on the gross output model from 1978 to 2016. We construct production accounts for the industries using input‐output relationships for the 26 main agricultural commodities and commodity groups, which account for over 90 per cent of the total agricultural inputs and outputs. The results show that China’s agricultural TFP grew at a rate of approximately 2.4 per cent a year before 2009, which is comparable to the main OECD countries and is double the world average. TFP growth accounts for approximately 40 per cent of output growth, suggesting that input growth was the main driver of output growth in the past. However, average productivity growth slowed down after 2009 though it has gradually recovered since 2012. The slowdown reflects the emerging challenges to existing farm production practices in Chinese agriculture, suggesting the need for further institutional reform.     

Biodiversity Conservation and Productivity in Intensive Agricultural Systems

This paper explores the economic effects of biodiversity loss on marketable agricultural output for intensive agricultural systems, which require an increasing level of artificial capital inputs. A theoretical bio‐economic model is used to derive a hypothesis about the effect of the state of biodiversity on the optimal crop output both in the longer run and in the transitional path towards the steady‐state equilibrium. The hypothesised positive relationship between biodiversity stock and optimal levels of crop output is empirically tested using a stochastic production frontier approach, based on data from a panel of UK specialised cereal farms for the period 1989–2000. The results support the theoretical hypothesis. Increases in biodiversity can lead to a continual outward shift in the output frontier (although at a decreasing rate), controlling for the relevant set of labour and capital inputs. Agricultural transition towards biodiversity conservation may be consistent with an increase in crop output in already biodiversity‐poor modern agricultural landscapes.     

The benefits from public agricultural research in Uruguay

We use newly constructed data to model and measure agricultural productivity growth and the returns to public agricultural research conducted in Uruguay over the period 1961–2010. We pay attention specifically to the role of levy‐based funding under INIA, which was established in 1990. Our results indicate that the creation of INIA was associated with a revitalization of funding for agricultural R&D in Uruguay, which spurred sustained growth in agricultural productivity during the past two decades when productivity growth was stagnating in many other countries. The econometric results were somewhat sensitive to specification choices. The preferred model includes two other variables with common trends, a time‐trend variable and a proxy for private research impacts, as well as a variable representing the stock of public agricultural knowledge that entailed a lag distribution with a peak impact at year 24 of the 25‐year lag. It implies a marginal benefit‐cost ratio of 48.2, using a real discount rate of 5 per cent per annum and a modified internal rate of return of 24 per cent per annum. The benefit‐cost ratio varied significantly across models with different lag structures or that omitted the trend or the private research variable, but across the same models, the modified internal rate of return was very stable, ranging from 23 per cent per annum to 27 per cent per annum. These results suggest that the revitalized investment in research spending under INIA has been very profitable for Uruguay and that a greater rate of investment would have been justified.     

A Cost Function Approach to Dynamic Duality: An Application to the US Cigarette Manufacturing Industry

The dynamic duality econometric approach with the case of multiple outputs is applied to the US cigarette manufacturing industry to test for the presence of adjustment costs and quasifixed inputs with regard to stocks of capital and tobacco. Capital and tobacco stocks are found to be quasi-fixed inputs and the empirical results indicate that there are significant adjustment costs associated with adjusting these inputs. Short- and long-run own- and cross-price elasticities of factor demands are estimated for domestic and imported tobaccoes, materials, tobacco stocks, and capital. Output demand elasticities are also estimated. The two outputs, cigarettes produced for export and for the US market, are examined for equality of marginal costs. No evidence of differences in marginal costs was found. There is evidence that government restrictions on advertising have negative effects on output demand.     

Nonpecuniary Benefits to Farming: Implications for Supply Response to Decoupled Payments

We develop a household model wherein farmers allocate labor to maximize utility from leisure, consumption, and nonpecuniary benefits from farming. The model shows that farmers with decreasing marginal utility of income respond to higher decoupled payments by decreasing off-farm labor and increasing farm labor, resulting in greater agricultural output. We then estimate the difference between farm and off-farm returns to labor using data from three nationally representative farm household surveys. The finding of a large on-farm/off-farm wage differential provides compelling evidence of substantial nonpecuniary benefits from farming.     

Explaining gender differentials in agricultural production in Nigeria

This article uses data from the General Household Survey Panel 2010–2011 to analyze differences in agricultural productivity across male and female plot managers in Nigeria. The analysis utilizes the Oaxaca‐Blinder decomposition method, which allows for decomposing the unconditional gender gap into: (i) the portion caused by observable differences in the factors of production (endowment effect) and (ii) the unexplained portion caused by differences in returns to the same observed factors of production (structural effect). The analysis is conducted separately for the North and South regions, excluding the west of the country. The findings show that in the North, women produce 28% less than men after controlling for observed factors of production, while there are no significant gender differences in the South. In the decomposition results, the structural effect in the North is larger than the endowment at the mean. Although women in the North have access to less productive resources than men, the results indicate that even if given the same level of inputs, significant differences still emerge. However, for the South, the decomposition results show that the endowment effect is more important than the structural effect. Access to resources explains most of the gender gap in the South and if women are given the same level of inputs as men, the gap will be minimal. The difference in the results for the North and South suggests that policy should vary by region.     

E-commerce and supply chain resilience during COVID-19: Evidence from agricultural input e-stores in China

COVID-19 has caused major disruptions to agricultural supply chains around the world. Researchers and policy-makers are interested in identifying means to reduce the disruptive effects caused by the pandemic. We investigate the impacts of COVID-19 on the operation of e-commerce stores (in short, e-stores) specialising in agricultural inputs. The difference-in-differences method (DID) is employed to estimate the causal relationship between COVID-19 and online sales of agricultural inputs using data from 54,244 agricultural input e-stores registered in 118 prefecture-level cities across 15 provinces and hosted on two major Chinese e-commerce platforms. The results show that COVID-19 led to a substantial growth in monthly sales of agricultural input e-stores, and this growth of online sales varied across store scales and by types of agricultural inputs. In particular, e-stores selling seeds and seedlings experienced a larger growth in sales than stores selling agricultural machinery and implements, and the mid- and larger-scaled e-stores experienced more growth of sales than micro- and small-scaled e-stores. Further analysis reveals that the growth of online sales of agricultural inputs was driven mainly by an increase in the quantity of customer orders (QCO). The findings of this paper underscore the importance of e-commerce in ensuring the resilience of the agricultural supply chain during the pandemic period.     

International Supply Response

Long-run aggregate agricultural supply elasticities obtained from conventional supply functions fitted to time series data tend to be relatively inelastic in the range of 0.1 to 0.4. I argue that these estimates substantially understate the true long-run supply response in agriculture. Because of the lack of international input price data, implicit output/input price ratios are estimated from a production function assuming profit maximization. The estimation of an aggregate supply function utilizing these price ratios yields long run aggregate supply elasticities in the range of 0.90 to 1.19. These figures are substantially larger than those obtained from conventional supply functions fitted to time series data, but correspond closely to estimates reported in an earlier crosscountry study that used different price data for different points in time. The results imply that policies which distort domestic and/or world market prices of agricultural products cause greater output distortions in both the DCs and LDCs than are predicted by the small supply elasticities obtained from conventional supply estimation.     

Advanced breakeven analysis of agricultural enterprise budgets

Advanced conceptual techniques for breakeven analysis of agricultural enterprise budgets are developed and applied. Breakeven points of a single agricultural enterprise, breakeven points between enterprises, and elasticities of breakeven points between enterprises are defined and mathematically derived. Breakeven equations were reduced to computational formulas for several budget components: output price, yield, input price, input requirement, variable cost, fixed cost, and total cost. Application of the advanced breakeven techniques provide evidence that both rice and wheat production seem especially desirable for the Arkansas Delta region in light of a fluctuating economic environment.     

Productivity in Indonesian agriculture: Impacts of domestic and international research

Total factor productivity growth contributed 38% of Indonesia's agricultural output growth from the mid-1970s to the mid-2000s. This study uses time series data analysed with an error correction mechanism to examine the contribution that Indonesian publicly funded agricultural research made to this outcome, allowing for other possible determinants of productivity growth, including international agricultural research, extension, government price policy and weather. The results imply a 27% real annual rate of return from a marginal increase in Indonesian agricultural research expenditure. Indonesia's public agricultural research explains virtually all of its agricultural total factor productivity growth between 1975 and 2006.     

Impacts of prices and transactions costs on input usage in a liberalizing economy: evidence from Tanzanian coffee growers

Despite improvements in production incentives, agricultural output in Africa remained sluggish through the 1990s. Low use of purchased inputs may be part of the cause of persistently low productivity in African agriculture. This article analyzes the roles of relative prices and transactions costs in explaining low use of chemical inputs among Tanzanian coffee growers. A sample selection model indicates that output prices exert great influence on input purchases and that both fixed and variable transactions costs affect input use decisions. Travel costs in input and output markets have distinct effects on input usage, implying distinct avenues for interventions to promote more intensive use of agricultural inputs.