Next Article in Journal
Effects of Chemical Fertilizer Combined with Organic Fertilizer Application on Soil Properties, Citrus Growth Physiology, and Yield
Next Article in Special Issue
Cereal-Legume Value Chain Analysis: A Case of Smallholder Production in Selected Areas of Malawi
Previous Article in Journal
Nutritional Quality and In Vitro Rumen Fermentation Characteristics of Silage Prepared with Lucerne, Sweet Maize Stalk, and Their Mixtures
Previous Article in Special Issue
Assessing Food Insecurity and Its Drivers among Smallholder Farming Households in Rural Oyo State, Nigeria: The HFIAS Approach
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

R&D Innovation Adoption, Climatic Sensitivity, and Absorptive Ability Contribution for Agriculture TFP Growth in Pakistan

1
Department of Economics, Arid Agriculture University Rawalpindi, Rawalpindi 46000, Pakistan
2
Dean Faculty of Social Sciences, Arid Agriculture University Rawalpindi, Rawalpindi 46000, Pakistan
3
Paul Engler College of Agriculture & Natural Sciences, West Texas A&M University, Canyon, TX 79016, USA
4
Department of Statistics and Mathematics, Arid Agriculture University Rawalpindi, Rawalpindi 46000, Pakistan
*
Authors to whom correspondence should be addressed.
Agriculture 2021, 11(12), 1206; https://doi.org/10.3390/agriculture11121206
Submission received: 27 September 2021 / Revised: 16 November 2021 / Accepted: 17 November 2021 / Published: 30 November 2021
(This article belongs to the Special Issue Agricultural Food Security and Economic Analysis)

Abstract

:
Innovation adoptions in agriculture sustain high total factor productivity (TFP) growth and overcome a potential production gap, which is beneficial for food security. Research and development (R&D) innovation adoption in agriculture sector is dependent on producers’ willingness to adopt, knowledge capital spillovers, and financial capacity. This research aims to investigate the impact of R&D innovation adoption and climate factors on agriculture TFP growth in Pakistan. The annual time series data were collected from different sources for the period of 1972–2020. For measuring the agriculture TFP, this study adopted the Cobb Douglas and Translog production functions. To analyze the impact of R&D innovation adoption and climate change on agricultural productivity, the dynamic autoregressive distributive lag (ARDL) and two-stage least square (TSLS) approaches were applied for regression analysis. The study outcomes highlight that the agricultural innovation adoption has a significantly positive impact on agriculture TFP growth in Pakistan with weak farmers’ absorptive ability. According to the results, agriculture tractors, innovative seed distribution, and fertilizer consumptions make a significantly positive contribution to agriculture TFP growth. Further, rainfall shows a positive and significant impact on agricultural productivity, where a moderate climate is beneficial for agricultural productivity. The estimation results contain policy suggestions for sustainable R&D adoption and agrarians’ absorptive ability. Based on the obtained results, it has been suggested that producers should focus on R&D innovation adoption to attain higher productivity. The government needs to emphasize innovative technology adoption, specifically to implement the extension services to increase farmers’ education, skills based training, and networking among the farmers to enhance their knowledge capital and absorptive ability. The farmers should also focus on the adoption of climate smart agriculture that can be achieved through the proper utilization of rainwater. For this purpose, the government needs to develop small community dams and large-scale dams for better use of rainwater harvesting.

1. Introduction

The Global Agriculture and Productivity (GAP) report of 2020 shows that an additional 71–100 million people fell into extreme poverty due to the COVID-19 shutdown, and 235 million people are at high risk of acute hunger [1]. Globally, 20–40% of crop yield is lost each year due to pest attacks, while around 81 million people’s food can be consumed by large swarms annually [1]. Agriculture innovation and its adoption is an essential element to overcome the global food security and poverty challenges. Climatic vulnerability harms the agriculture output and creates food insecurity problems, especially in developing countries [2]. These food insecurity challenges may even lead to a negative impact on a global scale. Sustainable agriculture research and development (R&D) and its spillover shocks, climatic protection strategies, water management policies, and technology adoption in the agriculture sector have significant potential to save the world from a harmful climate and upcoming food security challenges. The adoption of innovative technology in the agriculture sector provides mechanization and permits adaptation of the optimum level of input. Utilizing optimal agriculture inputs reduces the labor demand, monitoring cost, water scarcity challenges, and soil erosion. Further, an optimal input provides digital logistics and trustworthy consumers to the farmers, through which agrarians can maintain high yield and profitability. The combined collaborative efforts between government and private partners can enhance innovative adoptions in agriculture [3]. A country’s efforts toward agriculture innovation and adoption perform an essential role in farm and off-farm agriculture growth [4]. Agriculture innovations consolidate the higher level of output and sustain food availability in both developed and developing countries, which is also helpful in food accessibility through international agriculture trade.
The agriculture total factor productivity (TFP) increases when agrarians adopt and efficiently utilize innovative technology to attain a higher yield with fewer input resources. Additionally, technology adoption in the cropping sector (e.g., innovative and hybrid seeds, fertilizers, agri-technology, water efficiency, water management technology, hydroponic technology, etc.) intensifies the output, product efficiency, and profitability. Similarly, through the usage of animal genetics, breed varieties, and veterinary medicines, farmers can get more milk, meat, and eggs with fewer inputs [1,5,6,7]. Fundamental ways to mitigate agriculture risks are R&D, innovation adoption, and enhanced agronomic expertise. The farmer’s expertise improves through their access to education, extension services, training, skill-based workshops, community-led solutions, and networking with research institutions [8,9]. To enhance agricultural productivity and agrarians’ expertise, Steensland [1] highlights the policy priorities to adopt at a country level, which include R&D funding, extension services, agrarians’ training, accelerating innovation adoption, networking policies, and financial management.
To feed the 10 billion projected global population in 2050, global agriculture requires a higher annual TFP growth of 1.73% as compared to current TFP growth, which is at 1.63% [1]. In order to sustain the global food requirement, the government R&D spending in agricultural innovation is essential for agriculture productivity, food security, and environmental sustainability [9]. The government should provide agricultural innovation, such as rival goods, especially innovative seed varieties and other essential agriculture technologies. Further, the negligence of government towards agriculture R&D puts future agri-TFP growth at high risk [10]. Globally, the private sector accounted 52.5% of domestic agriculture R&D spending in 2011 as compared to 42% in 1980 [9]. A paradigm shift happened in agriculture R&D spending from public to private, and worldwide from higher income countries to middle-income countries (like USA to Brazil, China, and India) [9,11]. Further, Chai et al. [12] point out that the Chinese government spent the two-thirds of R&D funding on agriculture from their total, outspending the USA in both public and private sector agriculture expenditures.
Innovation adoption in agriculture is an essential element to increase farm level productivity, farmers’ welfare, and to ensure the food security. However, the technological spillovers and adoption are heterogeneous across the farms, regions, and countries, creating heterogeneity in farmers’ welfare, income equality, and poverty [13,14,15]. Similarly, Hurley et al. [16] argued that agriculture R&D spending has a high rate of return and provides 2.5–5% productivity over the non-R&D spenders. In addition, Diederen et al. [17] concluded that the farmers who adopt innovation early gain higher profitability over later adopters, so the period of innovation adoption is as important as innovation itself. Maffioli et al. [18] found the positive effect of technology adoption on agriculture output, whereas the improved adoption rate of seed varieties is certainly weak. Similarly, Gallardo & Sauer [19] investigated the role of laborsaving and technological adoption in agriculture output and found that innovative adoption performs a successive role in agriculture productivity. Bucci et al. [20] worked on precision agriculture technology (PAT) adoption in Italy and concluded that the chosen PAT farmers face various obstacles like geographical area, cultural barrier, limited information about benefits, small farm size, and low appreciating benefits over the cost of PAT chosen. Additionally, ref. [1,21] reviewed the existing literature and found that weak extension services are major hurdles in agriculture technology diffusion and earlier adoption.
Investment in agriculture innovation performs an influential role in agriculture growth. Evidence shows that agriculture investment has a high rate of return, and investment in tractor and tube-well has a significant impact on agriculture production in Punjab, Pakistan [22]. Chandio et al. [23] revealed that farmers’ financial access and technology adoption improved cereal production, whereas carbon emission and harmful climate negatively affected cereal output. Shabbir & Yaqoob [24] argued that agriculture innovative inputs have stabilized the cotton production in Pakistan and that hybrid seed, mechanization, and cotton yield area are important reasons for the growing output. Abdullah & Samah [25] studied the appreciative measures of agriculture technology and concluded that the farmers’ education, willingness to learn about advanced technology, updated knowledge of extension workers, extension services, and the area’s physical condition all perform an essential role in technology transformation and agriculture output. The updated knowledge of extension workers can help in training and educating the farmers to prepare for agriculture innovation adoption. In a similar line, Sjakir et al. [26] found that agri-field schools modernized agrarian expertise and extension services, playing an effective role in technology expansion. This revealed that the farmers who attended such field schools have improved knowledge and productivity.
In Pakistan, the agriculture share of GDP is decreased because of the lacking government consideration, poor agriculture development expenditure, weak extension services, technological availability, low agriculture financial services, farmers’ education, and lack of agrarians’ management skills [27,28]. The agriculture innovation adoption (like innovative seeds, new production process, water management strategies, fertilizers, pesticides, and advance agri-technology) and favored production environment have a noteworthy impact on the agriculture production of Pakistan. The innovation in agriculture technology has the potential to increase the output that leads to an increase in farmers’ income, consumption, standard of living, and poverty alleviation in rural areas [28,29,30,31]. Additionally, Ali & Behera [27] found that the wealthier and educated young farmers are more willing to adopt alternative irrigation technology, but the frequent energy shortage and lack of financial credit facilities creates hurdles in irrigation technology adoption. Similarly, Chandio et al. [32] argued that farm size, labor, credit, and fertilizer usage perform a positive role in rice productivity, while technical efficiency indicates that rice farmers are highly efficient with rice production technology.
This paper contributes to the literature in following ways. First, the perpetual inventory methodology (PIM) was used to calculate the net capital stock of the agriculture sector of Pakistan [33]. For necessary conditions, the initial capital stock is calculated from the gross capital investment of the agriculture in Pakistan. Second, the TFP is calculated by adopting the Cobb Douglas production function in the form of Translog function. Thirdly, the existing literature regarding agriculture innovation adoption focused on a cross-section framework, either cross-country or across the regions, while this study adopts the time series empirics for country-specific analysis. This research not only focused on technology adaption measures, but also incorporated the laborer absorptive capacity in the agriculture sector. Fourth, the literature review regarding technology adoption in the agriculture sector of Pakistan [13,34,35,36,37] is not readily available and most studies focused on analyzing the impact of agriculture inputs on value addition. This unique study comprehensively measured agriculture TFP growth and investigates the dependency relationship between technology adoption and absorptive ability factors on TFP growth. Fifth, simultaneously, numerous diagnostic tests were applied for quantitative results, and the problem of endogeneity was detected through a different sort of analysis. To avoid the endogeneity problem, two stage least square (TSLS) regression analysis was adopted to provide an instrumental specification to resolve the endogeneity problem.
At the beginning of the twenty-first century, R&D innovation revolutionized the agriculture production process and social well-being of the farmers. Globally, the increasing role of R&D in economic performance of agriculture sector is a key motivation of this research. R&D contributes through potential and intensive utilization of resources, increasing efficiency of human resources, availability, and adoptability of internal and external innovation. Agriculture R&D spending has a negligent share in the GDP of Pakistan, which is creating the potentiation gap in agriculture TFP growth. Existing literature suggested that agriculture performance of developed and developing countries is dependent on R&D spillovers, knowledge shocks, adoption of innovation, and farmers absorptive ability [38,39,40]. However, current studies have presented an overview of R&D spillover, but empirical studies are very rare in the case of Pakistan and not focused on farmers absorptive ability. The role of R&D spillovers in agriculture productivity growth in the context of empirical manners is missing in case of Pakistan. At present, the existing literature is unable to answer the following questions: Does technological adoption perform a productive role in agriculture output in Pakistan? Does the agriculture investment affect the volatility of TFP growth? Does the R&D innovation adoption perform a significant role in the agriculture TFP of Pakistan? Does the climatic variation affect the agriculture TFP growth in Pakistan? Does the human capital perform a fruitful role in agriculture TFP? Do the farmers have the absorptive capacity to adopt the agriculture innovation for higher output? By keeping in mind such research questions and gaps in the existing literature, this research comprises the following objectives: calculating the TFP growth in Pakistan and investigating the impact of technological adoption on agricultural TFP. The purpose of this research is to explore the role of R&D adoption in agriculture TFP growth in Pakistan. This study investigates the farmers absorptive ability regarding R&D innovation in the agriculture sector. This research also examines the effects of climatic variation in agriculture TFP.

2. Materials & Methods

The agriculture R&D investment and innovation adoption performs a productive role in agriculture output. Agriculture R&D brings innovative seed varieties, new sowing and harvesting techniques and technologies, and product efficiency to boost the agriculture output. R&D innovation enhances the irrigated water efficiency and protects the crops from climatic vulnerability. In this adoption, farmers’ education, management skill, ground-level experience, willingness, and financial capacity to opt for the innovation and absorptive ability are crucial factors. The evidence shows that agriculture R&D has a higher rate of return and the adoption of agriculture technology, fertilizers, and innovative seeds has a significant impact on agriculture TFP growth [22,23]. Numerous approaches are available to compute the TFP growth, such as index numbers, stochastic frontiers, ordinary least square, etc. However, such approaches are not applicable in this circumstance and considering the availability of data [41]. The agriculture TFP growth is measured through the Cobb Douglas and translog production functions due to data constraints. In addition, the Cobb Douglas-based production function measures the TFP at the aggregate level, which is appropriate to compute the TFP growth from conventional time-series data for labor and capital inputs, in addition to climatic variations. The essential assumption for the Cobb Douglas production function includes a constant return to scale in respective labor, capital, land, and climate factors in both production and input processes [42,43].
In this research, the Cobb Douglas production function is modified, according to the input requirements for agriculture cultivation and introduced climatic factors to quantify the agriculture TFP growth over the traditional production function. The purpose of incorporating climatic factors into agriculture production function is that the climatic factors are as important as capital, land, technology, and labor inputs, which are all essential for agriculture output. The adapted form of the agriculture production function is expressed as follows:
Y t = A t K t α L t β CF t δ
In Equation (1), Yt is agriculture output, Kt is net capital stock, Lt is agriculture labor, CFt representing the climatic factors, and At is representing agriculture TFP (ATFP). ‘t’ represents time period, and α, β, and δ represent the shares of net capital stock, agriculture labor, and climate factors, respectively. Through inverting Equation (1) into input-output ratio and applying the logarithm with its properties on both sides, ATFPt growth is calculated as follows.
A TFP t = Y t K t α L t β CF t δ
lnATFP t = ln   Y t ( α lnK t + β lnL t + δ lnCF t )
As a result, Equation (3) demonstrates ATFPt growth, which is measured through agriculture inputs, such as capital stock, labor, and climatic factors. To measure the net capital stock Kt in agriculture, the perpetual inventory method was adopted [44,45].

2.1. Net Capital Stock

The rise in stock of capital is called gross capital formation and can be measured from total investment expenditure on agriculture within the economy. The gross capital formation is known as net investment in the agriculture sector by excluding the depreciation and inflationary effects. The net capital stock is a difference between gross capital formation and fixed capital consumption (known as depreciation rate or capital consumption). Depreciation rate refers to wear and tear expenditures from fixed capital to maintain capital stock in its initial condition.
The perpetual inventory methodology (PIM) approach based on ref. [46] is applied for calculating the net capital stock in agriculture [33]. In this approach, the approximation of initial capital stock, growth rate, and depreciation was incorporated. The agriculture GDP growth rate of the previous period is assumed as the current capital growth rate [41]. The estimated formulations are as follows:
K t + 1 = I t + ( 1 d ) K t
In notation (4), ‘t’ represents the time, Kt+1 is net capital stock in the agriculture sector, It is gross capital stock, and ‘d’ is the depreciation rate. However, the study focused on measuring the net capital stock in the agriculture sector, which requires the initial capital stock. For this purpose, the gross fixed capital formation is used to calculate the initial capital formation. The method to measure the initial capital is as follows:
I o = I i g i + d
Here, in notation 5, Io is the initial level gross fixed capital stock in agriculture, where gi is average growth in capital formation. For this, the proxy of previous period agriculture GDP growth was taken [41]. The data of depreciation are collected from the Panne world table 10.1, to measure the net capital stock. Ref. [47] argued that depreciation is taken as the average life span of machinery (capital equipment) and growth of capital stock ‘g’ must be taken as the average growth of capital stock through the sample range.

2.2. Empirical Model

For agriculture productivity, the inclusive R&D factors, knowledge capital, labor force, and climatic condition have a fruitful effect on agriculture output and promote inclusive agriculture growth. Domestic R&D innovation is suitable because it matches with country land requirement, soil fertility, climatic conditions, water sources availability, rural culture, social and politic norms, and economic conditions. Internal R&D in agriculture has a compatibility with land requirement while internally developed innovative seeds, irrigation techniques, and agriculture equipment and machinery are more appropriate and productive than foreign imported technology [45,48]. The functional relationship of ATFP growth is determined to investigate the R&D adoption indicators, climate factors, and farmers absorptive ability as endogenous determinants of ATFP growth. Salim & Islam [43] worked on the significance of technology adoption and its role in agriculture ATFPt growth and concluded that the innovation adoption is an essential factor to increase the ATFPt, which is largely influenced through climate factors and knowledge capital. The technology adoption is highly productive and beneficial to improve internal innovation and knowledge absorptive ability [44]. R&D innovation in the agriculture sector is measured through the Cobb Douglas production function in the Translog form. The agriculture output performance is dependent on ATFP, labor efficiency, net capital stock and cultivation land. The domestic innovation spillover is essential to increase the agriculture ATFP growth, which is largely influenced through technological spillovers, farmers’ willingness to adopt, farmers’ absorptive capacity, and climate factors. The modified form of the Cobb Douglas production function as follows:
Y t = ATFP t K t α L t β M t γ CF t δ
In Equation (6), Yt is agriculture output performance while A T F P t , L t α , K t β ,   M t γ ,   and   C F t δ are ATFP, agricultural labor force, net agriculture capital stock, agriculture land, and climate factors, respectively. Here, ‘t’ represents time and α, β, γ, and δ represent the respective weights of capital, labor, land, and climatic factor in agriculture. The ATFPt growth is dependent on R&D adoption in agriculture, climate factors, and agrarians’ absorptive capacity.
ATFP t = A t RD t AC t CF t
In Equation (7), the dependent variable is ATFPt growth and independent variables are ( A t ,   R D t ,   A C t ,   C F t ) technical progress, R&D adoption, absorptive capacity (AC), and climate factors (CF). Taking the natural log and adding intercept and residual terms in Equation (7), the final estimated model is provided.
LnATFP t = α 0 +   α 1 A t + α 2 RD t + α 3 AC t +   α 4 CF t + ε t
Equation (8) is modified according to the availability of data and assumption concerns with the agriculture growth model [43]. In the estimated model, the interactive term of human capital is incorporated with R&D indicators to measure the farmers’ absorptive capacity [44]. The estimated model to examine the role of R&D innovation adoption, climatic factors, and absorptive capacity on ATFPt growth is as follows:
lnATFP t = a 0 + a 1 lnAI t + a 2 lnAEMP t + a 3 lnHC t + a 4 lnRF t + a 5 lnAT t +   a 6 lnFC t + α 7 lnSD t +   a 8 lnHCxRD t   + π t
Equation (9) is the final estimated model to investigate the behavior of technology adoption channels, absorptive ability, and climate factor in ATFPt growth in Pakistan. Here, ATFPt is agriculture total factor productivity, which is a dependent variable, while the independent variables AIt, AEMPt, HCt, RFt, ATt, FCt, SDt, and HCxRDt represent agriculture investment, employment, human capital, rainfall, agriculture tractor, fertilizer consumption, innovative seeds distribution, and interactive term to measure the absorptive ability. The description and measurement units of variables are given in Table 1. In Equation (9), alphas (α’s) represent the equation coefficients and π represents the residual term. Initially, the ATFPt is considered as an exogenous variable and measured through the growth accounting technique. At the second stage, ATFPt is regressed on the indicators of technological adoption and climatic indicators. This research concerns the multicollinearity and endogeneity problem caused by technological adoption (input) factors in the estimation of ATFP and interactive term to capture the absorptive ability. The endogeneity can be overcome by including the interactive term and lagged instrumental variables in the estimated model [41], which cause the multicollinearity problems. In the estimated model, the interactive term of human capital with R&D incorporated in the model is used to overcome the endogeneity problems and measures the absorptive ability. For final analysis, the dynamic autoregressive distributed lag (ARDL) model and TSLS instrumental variable approach is applied.

2.3. Data and Data Sources

The TFP growth of agriculture sector is accessed through time series data. The annual data were collected for the period of 1973–2020 from different sources. The secondary data were collected from domestic and foreign sources. The foreign sources consist of World Development Indicators (WDI) and Food and Agriculture Organization (FAO). Meanwhile, information from domestic sources is taken from Pakistan Meteorological Department (PMD) Pakistan Agricultural Research Council (PARC), Ministry of National Food Security & Research, Ministry of Finance Pakistan, Economic Survey of Pakistan, and Pakistan Bureau of Statistics. The secondary source data were collected from various issues of economic surveys of Pakistan, reports issued by PMD [49], and different statistical books issued by Pakistan Bureau of statistics.

3. Results and Discussion

The purpose of this research is to examine the impact of R&D adoption, climatic factors, and absorptive ability on ATFP growth. Time-series data are utilized for long-run analysis, namely, to assess trending behavior and causes of spurious analysis. To avoid spurious analysis through the data cleaning process, the problem of multicollinearity and endogeneity is detected. The problem of multicollinearity existed due to interactive terms, and for this purpose, different models are estimated through incorporating the proxies of R&D adoption. Three models are estimated to avoid the multicollinearity in the dynamic autoregressive distributive lag (ARDL) approach, while the TSLS instrumental variable technique is applied as a remedy to endogeneity problems and to investigate the behavior of coefficient cleaning in the endogeneity problem.

3.1. Agriculture TFP Growth

The calculated value of ATFPt is given in Figure 1. The estimates show positive and upward trending behavior of ATFPt in Pakistan. The average ATFPt growth remains at 1.29 percent in Pakistan from 1972 to 2020. The ATFPt growth was 1.45% in 2020, which is less than the world agriculture TFP growth that was 1.63% in 2020 [1]. To ensure food security in Pakistan, it is necessary to increase ATFP growth at least at the rate of population growth, which was 1.95% in 2020 [50]. The ATFPt growth was high from 2000 to 2008 because of government structural improvement, e.g., cemented canals, agriculture credit structure, etc. especially in Punjab, Pakistan. This was the season that the average agriculture growth remained at 4.83% from 2003 to 2008 in Pakistan [51]. During this period, the government focused on the agriculture sector in different ways, such as agriculture research, improving the water use efficiency, agriculture institutional development, and extension services to educate the farmers about agriculture innovation [23,51,52].

3.2. Stationarity Results

In nature, time series data are used to assess trending behavior, which may cause the problem of spurious analysis. To avoid the spurious analysis, it is important in investigate the stationarity behavior. For this instance, the augmented Dickey–Fuller (ADF) unit root test is applied, and the results are given in Table 2. The ADF test findings indicate that all variables are stationary at first, except rainfall (RFt) and agriculture land (ALt). This clearly indicates that ATFPt growth in Pakistan is highly dependent on previous period R&D adoption, skills, and experience based human capital. Additionally, for ATFPt growth, the climatic factors and under cultivation land constitute an important matter in the present period.

3.3. ARDL Estimates

The ATFPt growth of Pakistan is measured through Cobb Douglas and Translog production functions, and the estimated value of long-run analysis results are shown in Table 3. For empirical analysis, three different models are estimated to captures the role of R&D adoption and climatic sensitivity in ATFPt growth in Pakistan. The purpose of pertaining different models is to avoid multicollinearity and endogeneity challenges (results are given in Table 4. At the next stage, the endogeneity problem addressed through the TSLS method by selecting the appropriate instrumental variables. In estimated models, the dependent variable is ATFPt, while independent variables are AIt, ATt, SDt, FCt, and interactive term (INT)t. The INTt is buoyed to capture the laborer absorptive ability. The other control variables are ALt, AEMPt, HCt, and RFt. In all three models, the calculated value of ARDL bound test is greater than the upper bound, which specifies the significant rejection of the null hypothesis of no cointegration relationship among estimated variables. The empirical results of the ARDL bound test show that all three estimated models have long run cointegration association. The percentage share of control variables ALt, AEMPt, AIt, and RFt in agriculture TFP growth is 26%, 8%, 19%, and 17%, respectively. The contribution of R&D adoption variables ATt, SDt, and FCt to TFP growth is 28%, 24%, and 17%, respectively.
In model 1, the calculated value of agriculture land (ALt) has positive and significant impact on ATFPt growth with elasticity value of 0.2625. As cultivation land is increased with other efficient input factors, the agriculture productivity will increase in long-term. The computed coefficient of ALt shows a 1% increase in agriculture land increasing the ATFPt by 26%. The magnitude value ALt highlights that agriculture land holds a higher share in ATFPt growth and reduction in cultivation land reduces the agriculture productivity. Urbanization is the reason for a reduction in cultivation land and lowering of the agriculture share of GDP [53], creating food security challenges in Pakistan. The outcomes of the study are consistent with the findings of Villoria [48], who empirically found the increase in cropland area to be an agriculture productivity input to increase the ATFPt growth. Land ownership policies play a key role and land inputs made a remarkable contribution to agriculture output [54]. The coefficient value of AEMPt has a positive and significant impact on ATFPt growth in Pakistan. The AEMPt coefficient value is 0.085, which is significant at the 10% level. The calculated value of AEMPt indicates that agriculture labor has an 8% share in ATFPt growth. The AEMPt results illustrated that agriculture labor has a productive contribution to ATFPt growth, and therefore the efficient agriculture labor produces a fruitful effect on ATFPt growth. Further, the agriculture sector of Pakistan disguises unemployment, which is a reason for lower labor productivity share in agriculture output.
Agriculture investment is considered in the model to capture the as proxy of innovation adoption ratio, the result of agriculture investment (AIt) shows significantly positive effect on ATFPt growth. The coefficient value of AIt is 0.1914, which is significant at 5%. AIt result indicates that one percent increase in agriculture investment increases the ATFPt growth by 19%. The positive impact of AIt highlights that, in agriculture, investment in innovation adoption has a key contribution to agriculture output. The outcomes are consistent with the results of Ahmed & Javed [52], who argued that timely agriculture investment and innovation adoption as inputs perform a productive role in agriculture output. In addition, Aslam [55] highlights that lower agriculture investment in innovative technology represents a fundamental constraint, including advanced technology availability, institutional support, and socio-economic conditions for agriculture sector.
The variable rainfall is used as proxy of climatic sensitivity to check the effectiveness in ATFP growth. The moderate level of rainfall has a crucial role in agriculture production. The rainfall in detrimental conditions causes the climate vulnerability, which affects the agriculture production directly. The estimated value of rainfall (RFt) has a significantly positive impact on agriculture TFPt. This shows that effective RFt and its proper utilization have an advantage to accelerate the ATFPt by 0.7%. The RFt results indicate that sustainable rainfall with managerial efficiency has a positive role in agriculture productivity. Therefore, the adoption of climatic smart agriculture induces the agriculture output, which can be achieved through proper utilization and mechanism for rainwater. For better use of rainwater, the government water policies and farmers efficiency are very important. These findings are consistent with Olayide et al. [56], who argued that directly rain-fed irrigation has a limited impact on output, whereas the appropriate utilization of rainwater has a positive and long-term impact on agriculture production. The estimated result of human capital (HCt) shows an insignificant impact, while the interactive term (HC*ATt) has a negative and significant influence on ATFPt growth. The insignificant results of HCt and negative outcome of the interactive term highlight that the agriculture labor force has less absorptive ability about technological innovation. The estimated results are similar to the findings of dos Santos et al. [57] who found that the farmers who have high absorptive ability are those who are closer to research institutions and they have higher networking, while the lower networking farmers have less absorptive ability. Similarly, Onegina et al. [58] concluded that the labor productivity in the agriculture sector is dependent on capital spent to increase the labor efficiency. The agriculture human capital will be more efficient if government focused on the knowledge capital and skill development of the agriculture labor.
To capture the impact of R&D innovation adoption on ATFPt, different proxies are taken, such as agriculture tractors, innovative seed distribution, and agriculture fertilizer consumptions. The coefficient value of the agriculture tractor (ATt) has a positive and significant effect on ATFPt. The results of ATt highlight that a 1% increase in the number of tractors in the agriculture sector increases the ATFPt by 28%. This indicated that technological adoption in agriculture performs an essential role in agriculture productivity. The estimated value of LATt indicates that R&D expenditure on agriculture technology performs fruitful role in enhancement of ATFPt. Estimates directed that farmers should be a focus of LATt to increase the production efficiency and ATFPt growth. Based on outcomes, it is recommended that the farmers should focus on R&D innovation adoption to enhance their production and profitability. Additionally, the government should support the farmers to speed up the R&D innovation adoption and spillover process to enhance agriculture productivity. Similar results were found by Cavallo et al. [59] who argued that farmers with agriculture tractor adoption increase output, while the contract and large-scale farmers are more up to date in the adoption of innovations in agriculture.
In model 2, the policy variables ATt and interactive term are replaced with innovative seed distribution variable, while taking the control variables constant. The estimated behavior of all other variables are alike to the estimated model 1, except for the minor change in magnitude of the coefficients and the RFt variable behavior. The RFt variable has an insignificant impact on ATFPt growth. The innovative seed distribution (SDt) is utilized to capture the R&D innovative adoption in the agriculture sector. The coefficient value of SDt in agriculture has a positive and significant impact on ATFPt growth in Pakistan. This illustrated that innovative seed distribution has a dynamic role in agriculture output. The estimated value of SDt shows that a 1% increase in innovative seed adoption in agriculture increases ATFPt growth by 24%. The innovative seeds perform a crucial role in agriculture output and hybrid seed technologies are more disease secure, climatic resilient, drought resistant, etc. through which farmers can attain higher output. Similarly, Adolwa et al. [60] concluded that the agriculture productivity is low where the access to innovative seeds is low, and heterogeneity exists in seed distribution among farmers. The coefficient value of the interactive term of seeds distribution and human capital has a negative and significant impact. The negative coefficient of INTt infers that the knowledge capital about innovative seed adoption has less absorptive ability, farmers have weak knowledge about innovative seed distribution, and its early adoption is neglected.
In model 3, the R&D innovation adoption is captured through fertilizer consumption by taking the control variables as in the first model. The coefficient value of fertilizer consumption has a positive and significant impact on ATFPt growth in Pakistan. This indicates that farmers FCt for high yield has a key contribution to increasing FCt, optimally increasing the ATFPt by 17% annually. The findings are inconsistent with the outcomes of Raza et al. [36], who found an insignificant impact of fertilizers on agriculture output with a lower adaptive rate. The coefficient value of the interactive term of FCt and HCt has negative and significant impact. The negative coefficient of INTt supports that the knowledge capital about optimal fertilizers utilization has less absorptive ability and farmers have weak knowledge to gain the actual outcome of fertilizer consumption.
The ARDL short-run estimates are given in Table 5 where the short-run coefficients have an insignificant impact on ATFPt growth in Pakistan. In all three models, the coefficient value of short-run control variables ALt, AEMPt, AIt, HCt, and RFt and their lag coefficients demonstrate similar behavior as in the long run and mostly have an insignificant impact on ATFPt growth. The lag coefficient value of HCt has a significantly positive impact on ATFPt. The lag value of HCt indicated that education with field experience positively contributes towards ATFPt growth. Additionally, the R&D innovation indicators show a positive and significant impact on ATFPt growth in the short run. In the multivariate time series model, the convergence behavior, stochastic trend, and speed of adjustment of the dependent variable are attained through the error correction model (ECM). The ECMt(−1) coefficients values in all three models are negative and statistically significant, showing that the estimated models have convergent behavior towards equilibrium. The higher value of ECM (−1) revealed that if there is a disequilibrium in R&D input, adoption and input shocks can be adjusted with higher speed during the given period. The agriculture R&D innovation inputs perform an essential role in bringing ATFPt into its steady-state position [53]. The ECMt (−1) coefficient value is higher due to two cropping seasons (Rabi and Kharif) in Pakistan, with heterogeneously located areas, so if the production of one crop is lower, its production deficiency will recover either during the same crop or during the next crop [53,61].

3.4. Endogeneity and Diagnostic Test Results

The variables highly correlated with error create the endogeneity problem and the chosen lag period value is used as the instrumental variable. After identifying the instrumental variables, the endogeneity test for the validation of unbiased selection instruments was applied. As the estimated variables are first order stationary, to avoid the spurious results, the endogeneity problem was initially detected through Hansen’s (1982) J-test [62]. The null hypothesis of the endogeneity test is that the variables behave endogenously. The results of endogeneity test and residual diagnostic estimates are given in Table 5. The estimated result of J-stat is insignificant, which is evidence of the endogeneity problem in the estimates. Further, the selected instruments are uncorrelated with the residual term and independent variable. This selection resolves the problem of non-stationarity, heteroscedasticity, and autocorrelation in the estimated model [62,63]. To resolve the endogeneity problem, a one-year lag of highly correlated variables was used as an instrumental variable in the two stage least square (TSLS) approach [64]. Further, the interactive term is helpful in resolving the endogeneity problem. The TSLS is an instrumental variables approach, which efficiently handles the endogeneity and spurious regression problems [63]. The residual diagnostic estimates highlight that there is no evidence of heteroscedasticity and autocorrelation in the estimated model. Further, the data are normally distributed.

3.5. Two Stage Least Square Results

The results of TSLS are given in Table 6, where the dependent variable is ATFPt, while independent variables are AIt, AEMPt, ATt, SDt, FCt, HCt, and interactive term (HC *ATt). The estimated value of the F-statistic is high and significant, which shows that model is overall significant and likely useful for policy implication. The value of R-square is 0.93, which indicates that the estimated model is well-fitted, and 93% variation in ATFPt growth is explained by the selected independent variables. In general, the TSLS analysis shows that agriculture investment, agriculture land, seed distributions, tractors, fertilizer consumption, and rainfall perform a positive and significant role in ATFPt. The estimated value of agriculture employment and human capital has an insignificant impact on ATFPt, whereas the calculated value of the interactive term have a negative and significant impact on ATFPt. The agriculture labor is less efficient in R&D innovation absorption due to over-employment and poor knowledge of R&D innovation.
Agriculture investment is considered as proxy of R&D capital. The result of agriculture investment (AIt) shows significantly positive effect on agriculture productivity growth. The coefficient value of AIt is 0.016, which is significant at 10%. This indicates that a 1% increase in agriculture investment increases the ATFPt growth by 1.6%. The positive impact of agriculture investment highlights that an increase in agriculture investment has key contribution to productivity. The outcomes are consistent with the results of Ahmed & Javed [52], who argued that timely agriculture investment and innovation adoption as inputs performs productive role in agriculture output. The estimated value of AEMPt has a negative and insignificant impact on agriculture ATFPt. The insignificant impact on AEMPt on agriculture productivity may be due to overemployment in the agriculture sector. Similarly, Onegina et al. [58] concluded that the labor productivity in the agriculture sector is dependent on capital spent for increasing the labor efficiency.
The variable rainfall is used as proxy of climatic sensitivity, as a moderate level of rainfall has a crucial role in agriculture production. The rainfall in detrimental condition caused climate vulnerability, which directly affects the agriculture production. The estimated value of rainfall (RFt) has a significantly positive impact on ATFPt. This shows that effective RFt and proper utilization of rainwater led to more accelerated agriculture productivity by 9.7%. The RFt results indicate that sustainable rainfall with managerial efforts has a positive role in agriculture productivity. Therefore, the adoption of climatic smart agriculture induces the agriculture output, which can be achieved through proper utilization of rainwater. For the better use of rainwater, the government water policies and farmer efficiency are very important. The findings are consistent with Olayide et al. [56], who argued that directly rain-fed irrigation has a limited impact on output, whereas the appropriate utilization of rainwater has a positive and long-term impact on agriculture production.
The coefficient value of the agriculture tractor (ATt) is positive and significant. The ATt results highlight that a one percent increase in the number of tractors in agriculture sector increase the ATFPt by 7.3% in Pakistan. This indicated that technological adoption in the agriculture sector of Pakistan performs an essential role in agriculture productivity. As a result, the farmers should focus on innovation adoption to enhance their production and profitability. The farmers’ technological adoption behavior speeds up the spillover process and agriculture output. Similar results were found by Cavallo et al. [59] who argued that farmers’ agriculture tractors adoption behavior increases the spillover impact, while the contract and large-scale farmers are more up to date in their adoption of innovation in agriculture. The coefficient value of SDt in agriculture has a positive and significant impact on ATFPt growth in Pakistan. The estimated value of SDt shows that one percent increase in innovative seed in agriculture leads to increase the ATFPt growth by 12 percent. The innovative seeds perform a crucial role in agriculture output and crops are more disease protected, climatic resilient, drought resistant, etc. through which farmers can attain higher output. Similarly, Adolwa et al. [60] concluded that the agriculture productivity in those areas where there is access to innovative seeds is low, and heterogeneity exists in seed distribution among farmers. The coefficient value of fertilizer consumption has a significant impact on ATFPt growth in Pakistan. This indicates that farmers are willing to adopt the fertilizers for higher agriculture output. The findings are consistent with the outcomes of Raza et al. [36], who found significant fertilizers impact on agriculture output.
The estimated result of human capital (HCt) presents a insignificant and positive effect, while the interactive term (HC*ATt) shows significant and negative effect on ATFPt growth. The insignificant results of HCt and negative coefficient of the interactive term highlights that agriculture labor has weak knowledge capital with lower R&D absorptive ability. The results indicate that agriculture labor has less absorptive efficiency of innovative technology at an early stage. The estimated results are similar to the findings of dos Santos et al. [57] who found that farmers have high absorptive ability those who are closer to research institutions have higher networking, while the lower networking farmers have less absorptive ability. Onegina et al. [58] concluded that the government needs to focus on agriculture knowledge capital, extension services to enhance the knowledge spillover in agriculture, educate the farmers about earlier innovation adoption, and increase the absorptive ability to gain high productivity. Farmer expertise, knowledge capital, and best practices for technology absorption have an effective role to play in agricultural development in Pakistan.
The authors in this study have used the time series data from 1973–2020 to assess the impact of R&D on agricultural productivity growth in Pakistan. Authors had data limitations as is true with many studies based on time series data. Meanwhile, these data can also be split into two different periods (1973–2000 and 2001–2020) to investigate the structural difference. The use of a structural break in data analysis can be more beneficial to determine structural differences and investigate the before and after shocks of R&D. However, the limited number of observations, 28 and 20, respectively, in the time series data available for two structural breaks (1973–2000 and 2001–2020) may create the spurious analysis problem as a normal distribution assumption may not hold and it may mislead the analysis of structural difference and make the test invalid.

4. Conclusions and Recommendations

This study investigates the role of R&D adoption in agriculture TFP growth in Pakistan. For the empirical analysis, the annual time series data were collected spanning from 1973 to 2020. The problem of multicollinearity existed due to interactive terms, and for this purpose, different models were estimated through incorporating the proxies of the R&D innovation adoption. Three models are estimated to avoid the multicollinearity in the dynamic ARDL approach, while the TSLS instrumental variables technique is applied as a remedy to the endogeneity problems and to investigate the behavior of the estimated coefficient cleaned from the endogeneity problem. It can be concluded from the analysis results that R&D adoption presents a fruitful impact on agriculture TFP growth in Pakistan. Agriculture innovation adoption spending, seed distribution, fertilizer consumptions, and agriculture tractors each perform a positive role in agriculture TFP growth. This indicates that the adoption of agricultural R&D innovation is a fundamental source of higher agriculture productivity. However, the human capital and agriculture employment show insignificant results, while interactive terms indicate a negative and significant impact on agriculture TFP in Pakistan. This highlights that the agriculture labor force has less innovation adoption knowledge and absorptive ability for R&D innovative in agriculture sector of Pakistan. The results indicate that sustained rainfall with managerial efforts play a productive role in agriculture productivity. Consequently, the adoption of climate smart agriculture induces higher agriculture output, which can be achieved through the proper utilization of rainwater. To improve agricultural output and farmers’ absorptive ability, the following policy measures are required:
  • The government and research institutions should increase the agriculture R&D innovation expenditures to increase agriculture productivity.
  • The research institutions and government should focus on innovative seed development (like hybrid seeds) and its early spillovers for higher agriculture productivity.
  • Urbanization has caused a reduction in cultivation land and lowered the agriculture output. So, the government should focus to enhance the under-cultivation of land to avoid food security challenges in Pakistan.
  • The government must develop and implement the extension services to educate the farmers about technological innovation and efficient resource utilization.
  • The government and technology developing agencies should focus on farmers’ expertise, knowledge-based training, skills-based workshops, capacity building, and community-led experiences to improve the absorptive ability of new technology.
  • Farmers should also focus on the adoption of climate smart agriculture, which can be achieved through a proper utilization of rainwater. For this purpose, the government needs to develop small community dams and large-scale dams for the timely use of rainwater.

Author Contributions

This research manuscript is part of the Ph.D. dissertation of M.U., and he generates research idea and scientific approaches for model development, collected data, data analysis and discussion, and bibliography. G.H. was supervised in data collection, model development, data estimation, review of the analysis results, and other technical aspects. A.S. was responsible for literature review, results translation, and review of the manuscript. L.K.A. was responsible for introduction write-up, general guidance, helped in model development, organized the whole manuscript and review of manuscript, conclusion & policy recommendations, and incorporated the reviewers’ suggestions. Being statistician, M.H. in model construction, data issues, and data analysis. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding. The APC was funded by Paul Engler College of Agriculture & Natural Sciences, West Texas A & M University, Canyon, TX 79016 USA.

Institutional Review Board Statement

Not Applicable.

Informed Consent Statement

Not Applicable.

Acknowledgments

The research manuscript has been supported though International Research Supports Fellowship Program (IRSFP) by Punjab Higher Education Commission (PHEC), Pakistan, through Usman’s institution, PMAS- Arid Agriculture University, Rawalpindi, Pakistan. Authors also acknowledge the acceptance and support of West Texas A&M University to host the Visiting Research Scholar from Pakistan.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Steensland, A. 2020 Global Agricultural Productivity Report: Productivity in a Time of Pandemics; Thompson, T., Ed.; Virginia Tech College of Agriculture and Life Sciences: Blacksburg, VA, USA, 2020. [Google Scholar]
  2. Eriksen, S.; Schipper, E.L.F.; Scoville-Simonds, M.; Vincent, K.; Adam, H.N.; Brooks, N.; West, J.J. Adaptation interventions and their effect on vulnerability in developing countries: Help, hindrance or irrelevance? World Dev. 2021, 141, 105383. [Google Scholar] [CrossRef]
  3. OECD. New Technologies and Digitalization Are Transforming Agriculture and Offering New Opportunities to Improve Policy. Available online: https://www.oecd.org/agriculture/events/oecd-global-forum-on-agriculture (accessed on 15 February 2021).
  4. Dechezleprêtre, A.; Martin, R.; Mohnen, M. Knowledge Spillovers from Clean and Dirty Technologies; Centre for Economic Performance, London School of Economics and Political Science: London, UK, 2014. [Google Scholar]
  5. Alston, J.M. Reflections on agricultural R&D, productivity, and the data constraint: Unfinished business, unsettled issues. Am. J. Agric. Econ. 2018, 100, 392–413. [Google Scholar]
  6. Thompson, T.; Gyatso, T. Technology Adoption for Improving Agricultural Productivity in Sub-Saharan Africa. Available online: https://globalagriculturalproductivity.org/ (accessed on 15 May 2021).
  7. USDA ERS. International Agricultural Productivity. Available online: https://www.ers.usda.gov/data-products/international-agricultural-productivity (accessed on 4 April 2021).
  8. Liu, J.; Wang, M.; Yang, L.; Rahman, S.; Sriboonchitta, S. Agricultural productivity growth and its determinants in south and southeast asian countries. Sustainability 2020, 12, 4981. [Google Scholar] [CrossRef]
  9. Rawat, S. Global volatility of public agricultural R&D expenditure. Adv. Food Secur. Sustain. 2020, 5, 119. [Google Scholar]
  10. Fuglie, K.O.; Nin-Pratt, A. Agricultural productivity: A changing global harvest. In 2012 Global Food Policy Report; International Food Policy Research Institute: Washington, DC, USA; pp. 14–25.
  11. Alston, J.M.; Pardey, P.G. Innovation, Growth, and Structural Change in American Agriculture. In The Role of Innovation and Entrepreneurship in Economic Growth; University of Chicago Press: Chicago, IL, USA, 2020. [Google Scholar]
  12. Chai, Y.; Pardey, P.G.; Chan-Kang, C.; Huang, J.; Lee, K.; Dong, W. Passing the food and agricultural R&D buck? The United States and China. Food Policy 2019, 86, 101729. [Google Scholar]
  13. Chavas, J.P.; Nauges, C. Uncertainty, learning, and technology adoption in agriculture. Appl. Econ. Perspect. Policy 2020, 42, 42–53. [Google Scholar] [CrossRef] [Green Version]
  14. Llewellyn, R.S.; Brown, B. Predicting Adoption of Innovations by Farmers: What is Different in Smallholder Agriculture? Appl. Econ. Perspect. Policy 2020, 42, 100–112. [Google Scholar] [CrossRef]
  15. Rehman, A.; Jingdong, L.; Khatoon, R.; Hussain, I.; Iqbal, M.S. Modern agricultural technology adoption its importance, role and usage for the improvement of agriculture. Life Sci. J. 2016, 14, 70–74. [Google Scholar]
  16. Hurley, T.M.; Rao, X.; Pardey, P.G. Re-examining the reported rates of return to food and agricultural research and development. Am. J. Agric. Econ. 2014, 96, 1492–1504. [Google Scholar] [CrossRef]
  17. Diederen, P.; Van Meijl, H.; Wolters, A.; Bijak, K. Innovation adoption in agriculture: Innovators, early adopters and laggards. Cah. D’economie Sociol. Rural. 2003, 67, 29–50. [Google Scholar]
  18. Maffioli, A.; Ubfal, D.; Vazquez-Bare, G.; Cerdan-Infantes, P. Improving technology adoption in agriculture through extension services: Evidence from Uruguay. J. Dev. Eff. 2013, 5, 64–81. [Google Scholar] [CrossRef]
  19. Gallardo, R.K.; Sauer, J. Adoption of labor-saving technologies in agriculture. Annu. Rev. Resour. Econ. 2018, 10, 185–206. [Google Scholar] [CrossRef] [Green Version]
  20. Bucci, G.; Bentivoglio, D.; Finco, A.; Belletti, M. Exploring the impact of innovation adoption in agriculture: How and where Precision Agriculture Technologies can be suitable for the Italian farm system? In Proceedings of the Paper Presented at the IOP Conference Series: Earth and Environmental Science, Ancona, Italy, 1–2 October 2018. [Google Scholar]
  21. Takahashi, K.; Muraoka, R.; Otsuka, K. Technology adoption, impact, and extension in developing countries’ agriculture: A review of the recent literature. Agric. Econ. 2020, 51, 31–45. [Google Scholar] [CrossRef]
  22. Kiani, A.K.; Iqbal, M.; Javed, T. Total factor productivity and agricultural research relationship: Evidence from crops sub-sector of Pakistan’s Punjab. Eur. J. Sci. Res. 2008, 23, 87–97. [Google Scholar]
  23. Chandio, A.A.; Jiang, Y.; Akram, W.; Adeel, S.; Irfan, M.; Jan, I. Addressing the effect of climate change in the framework of financial and technological development on cereal production in Pakistan. J. Clean. Prod. 2021, 288, 125637. [Google Scholar] [CrossRef]
  24. Shabbir, M.S.; Yaqoob, N. The impact of technological advancement on total factor productivity of cotton: A comparative analysis between Pakistan and India. J. Econ. Struct. 2019, 8, 1–16. [Google Scholar] [CrossRef] [Green Version]
  25. Abdullah, F.A.; Samah, B.A. Factors impinging farmers’ use of agriculture technology. Asian Soc. Sci. 2013, 9, 120. [Google Scholar] [CrossRef]
  26. Sjakir, M.; Azima, A.M.; Hussain, M.Y.; Zaimah, R. Learning and technology adoption impacts on farmer’s productivity. Mediterr. J. Soc. Sci. 2015, 6, 126. [Google Scholar] [CrossRef] [Green Version]
  27. Ali, A.; Behera, B. Factors influencing farmers’ adoption of energy-based water pumps and impacts on crop productivity and household income in Pakistan. Renew. Sustain. Energy Rev. 2016, 54, 48–57. [Google Scholar] [CrossRef]
  28. Chandio, A.A.; Yuansheng, J. Determinants of adoption of improved rice varieties in northern Sindh, Pakistan. Rice Sci. 2018, 25, 103–110. [Google Scholar] [CrossRef]
  29. Ali, A.; Abdulai, A. The adoption of genetically modified cotton and poverty reduction in Pakistan. J. Agric. Econ. 2010, 61, 175–192. [Google Scholar] [CrossRef]
  30. Hashmi, M.S. Land distribution, technological changes and productivity in pakistan’s agriculture: Some explanations and policy options. Management 2011, 1, 51–74. [Google Scholar]
  31. Renkow, M. Differential technology adoption and income distribution in Pakistan: Implications for research resource allocation. Am. J. Agric. Econ. 1993, 75, 33–43. [Google Scholar] [CrossRef]
  32. Chandio, A.A.; Jiang, Y.; Gessesse, A.T.; Dunya, R. The nexus of agricultural credit, farm size and technical efficiency in Sindh, Pakistan: A stochastic production frontier approach. J. Saudi Soc. Agric. Sci. 2019, 18, 348–354. [Google Scholar] [CrossRef]
  33. Barro, R.J.; Sala-i Martin, X. Economic Growth; The MIT Press: Cambridge, MA, USA, 2004. [Google Scholar]
  34. Akhtar, K.; Pirzada, S.S. SWOT analysis of agriculture sector of Pakistan. J. Econ. Sustain. Dev. 2014, 5, 127–134. [Google Scholar]
  35. Ali, A. Impact of agricultural extension services on technology adoption and crops yield: Empirical evidence from Pakistan. Asian J. Agric. Rural Dev. 2013, 3, 801. [Google Scholar]
  36. Raza, M.H.; Shahbaz, B.; Bell, M.A. Review based analysis of adoption gap and training needs of farmers in Pakistan. Int. J. Agric. Ext. 2017, 4, 185–193. [Google Scholar]
  37. Wang, Z.; Ali, S.; Akbar, A.; Rasool, F. Determining the influencing factors of biogas technology adoption intention in Pakistan: The moderating role of social media. Int. J. Environ. Res. Public Health 2020, 17, 2311. [Google Scholar] [CrossRef] [Green Version]
  38. Alston, J. The Benefits from Agricultural Research and Development, Innovation, and Productivity Growth. OECD Food Agric. Fish. Pap. 2010, 31, 1–27. [Google Scholar] [CrossRef]
  39. Ugochukwu, A.I.; Phillips, P.W. Technology Adoption by Agricultural Producers: A Review of the Literature. In From Agriscience to Agribusiness; Kalaitzandonakes, N., Carayannis, E.G., Grigoroudis, E., Rozakis, S., Eds.; Springer: Cham, Switzerland, 2018; pp. 361–377. [Google Scholar]
  40. Adetutu, M.O.; Ajayi, V. The impact of domestic and foreign R&D on agricultural productivity in sub-Saharan Africa. World Dev. 2020, 125, 104690. [Google Scholar]
  41. Sharif, N.; Chandra, K.; Mansoor, A.; Sinha, K.B. A comparative analysis of research and development spending and total factor productivity growth in Hong Kong, Shenzhen, Singapore. Struct. Chang. Econ. Dyn. 2021, 57, 108–120. [Google Scholar] [CrossRef]
  42. Coe, D.T.; Helpman, E. International R&D Spillovers. Eur. Econ. Rev. 1995, 39, 859–887. [Google Scholar]
  43. Salim, R.A.; Islam, N. Exploring the impact of R&D and climate change on agricultural productivity growth: The case of Western Australia. Aust. J. Agric. Resour. Econ. 2010, 54, 561–582. [Google Scholar]
  44. Coe, D.T.; Helpman, E.; Hoffmaister, A.W. International R&D spillovers and institutions. Eur. Econ. Rev. 2009, 53, 723–741. [Google Scholar]
  45. Läpple, D.; Renwick, A.; Cullinan, J.; Thorne, F. What drives innovation in the agricultural sector? A spatial analysis of knowledge spillovers. Land Use Policy 2016, 56, 238–250. [Google Scholar] [CrossRef]
  46. Griliches, Z. Issues in assessing the contribution of research and development to productivity growth. Bell J. Econ. 1979, 10, 92–116. [Google Scholar] [CrossRef]
  47. Kuo, F.Y.; Young, M.L. A study of the intention–action gap in knowledge sharing practices. J. Am. Soc. Inf. Sci. Technol. 2008, 59, 1224–1237. [Google Scholar] [CrossRef]
  48. Villoria, N.B. Technology spillovers and land use change: Empirical evidence from global agriculture. Am. J. Agric. Econ. 2019, 101, 870–893. [Google Scholar] [CrossRef] [Green Version]
  49. Government of Pakistan, Pakistan Meteorological Department. Climate Data Processing Centre. Available online: http://www.pmd.gov.pk/cdpc/home.htm (accessed on 1 March 2021).
  50. Government of Pakistan, Ministry of Finance. Budget Statement 2009. Available online: https://www.finance.gov.pk/survey_2021.html (accessed on 20 March 2021).
  51. Government of Pakistan, Ministry of Finance. Agriculture. Available online: https://www.finance.gov.pk/survey/chapter_10/02_agriculture.pdf (accessed on 20 March 2021).
  52. Ahmed, V.; Javed, A. National study on agriculture investment in Pakistan. Available online: https://www.think-asia.org/handle/11540/6822 (accessed on 12 April 2021).
  53. Usman, M.; Hameed, G.; Saboor, A.; Almas, L.K. Research and Development Spillover, Irrigation Water Use and Agricultural Production in Pakistan. WSEAS Trans. Environ. Dev. 2021, 17, 840–858. [Google Scholar] [CrossRef]
  54. Abdullahi, H.S.; Mahieddine, F.; Sheriff, R.E. Technology impact on agricultural productivity: A review of precision agriculture using unmanned aerial vehicles. In International Conference on Wireless and Satellite Systems; Springer: Berlin/Heidelberg, Germany, 2015; pp. 388–400. [Google Scholar]
  55. Aslam, M. Agricultural productivity current scenario, constraints and future prospects in Pakistan. Sarhad J. Agric. 2016, 32, 289–303. [Google Scholar] [CrossRef]
  56. Olayide, O.E.; Tetteh, I.K.; Popoola, L. Differential impacts of rainfall and irrigation on agricultural production in Nigeria: Any lessons for climate-smart agriculture? Agric. Water Manag. 2016, 178, 30–36. [Google Scholar] [CrossRef] [Green Version]
  57. dos Santos, J.A.; Roldan, L.B.; Loo, M.K.L. Clarifying relationships between networking, absorptive capacity and financial performance among South Brazilian farmers. J. Rural Stud. 2021, 84, 90–99. [Google Scholar] [CrossRef]
  58. Onegina, V.; Megits, N.; Antoshchenkova, V.; Boblovskyi, O. Outcome of capital investment on labor productivity in agriculture sector of Ukraine. J. East. Eur. Cent. Asian Res. 2020, 7, 12–25. [Google Scholar] [CrossRef]
  59. Cavallo, E.; Ferrari, E.; Bollani, L.; Coccia, M. Attitudes and behaviour of adopters of technological innovations in agricultural tractors: A case study in Italian agricultural system. Agric. Syst. 2014, 130, 44–54. [Google Scholar] [CrossRef]
  60. Adolwa, I.S.; Schwarze, S.; Buerkert, A. Impacts of integrated soil fertility management on yield and household income: The case of Tamale (Ghana) and Kakamega (Kenya). Ecol. Econ. 2019, 161, 186–192. [Google Scholar] [CrossRef]
  61. Ullah, A.; Arshad, M.; Kächele, H.; Khan, A.; Mahmood, N.; Müller, K. Information asymmetry, input markets, adoption of innovations and agricultural land use in Khyber Pakhtunkhwa, Pakistan. Land Use Policy 2020, 90, 104261. [Google Scholar] [CrossRef]
  62. Semadeni, M.; Withers, M.C.; Trevis Certo, S. The perils of endogeneity and instrumental variables in strategy research: Understanding through simulations. Strateg. Manag. J. 2014, 35, 1070–1079. [Google Scholar] [CrossRef]
  63. Orji, A.; Mba, P.N.; Peter, N. Foreign Private Investment, Capital Formation and Economic Growth in Nigeria: A two stage least square approach. J. Econ. Sustain. Dev. 2010, 6, 57–63. [Google Scholar]
  64. Siller, M.; Schatzer, T.; Walde, J.; Tappeiner, G. What drives total factor productivity growth? An examination of spillover effects. Reg. Stud. 2021, 55, 1129–1139. [Google Scholar] [CrossRef]
Figure 1. Agriculture Total Factor Productivity in Pakistan. Source: Author’s Calculation.
Figure 1. Agriculture Total Factor Productivity in Pakistan. Source: Author’s Calculation.
Agriculture 11 01206 g001
Table 1. Variables and Description.
Table 1. Variables and Description.
Abbreviation VariablesMeasurement Units
TFPtTotal Factor Productivity Measured through Cobb Douglas and Translog Production Function
AItAgriculture InvestmentMillion Rupees
ALAgriculture Land Percentage share in total Land
ATtAgriculture TractorTotal in numbers
AEMPtAgriculture EmploymentPercent share in total Employment
HCtHuman Capital indexIndex developed by panne world table 10.1
RFtAverage Annual RainfallMillimeter
FCtFertilizer Consumption in (000) tons000 Tonnes
SDInnovative Seeds Distribution Tonnes
INTtInteractive term of Human capital with Agriculture Tractors, Seed Distributions, and Fertilizers ConsumptionsTo Capture the Absorptive Ability
Table 2. The ADF Test Results.
Table 2. The ADF Test Results.
VariablesLevelFirst Difference
T-Statp-ValueT-Statp-Value
TFPt0.0060(0.6796)−6.1420(0.0000) *
ALt2.8426(0.0236) *5.1769(0.0000)
AIt−1.3215(0.1693)−3.2986(0.0016) *
LAEMPt−1.1515(0.6874)−8.2817(0.0000) *
LHCt−1.2886(0.6270)−6.6419(0.0000) *
LRFt−5.4196(0.0003) *−5.9350(0.0001)
LATt0.0791(0.9607)−8.4761(0.0000) *
SDt(1.3291)(0.4932)6.1674(0.0000) *
LFCTt−0.1679(0.9345)−5.0333(0.0002) *
* Indicates the integration level of all variables.
Table 3. ARDL Long Run Coefficient and Bound Test Results.
Table 3. ARDL Long Run Coefficient and Bound Test Results.
Model 1Model 2Model 3
VariablesCoefficientProbCoefficientProbCoefficientProb
C−5.65590.0000 ***3.94840.0276 **−2.31730.0000 ***
LALt0.26250.0188 **0.12860.0292 **0.07270.2535
LAEMPt0.085470.0921 *0.02060.84210.19930.8016
LAIt0.19140.0102 **0.02980.0001 ***0.51340.0001 ***
LHCt0.28760.42200.25600.19340.05910.3562
LRFt0.00730.0146 **0.01020.73510.17260.0008 ***
INTt−0.064310.0515 *−0.06890.0066 **−0.58870.0002 ***
LATt0.28650.0086 ***
LSDt 0.24750.0282 **
LFCt 0.17690.0001 ***
ARDL Bounds TestF-Statistic5.1832F-statistic5.0530F-statistic16.7837
***, **, * represents the level of significance at 1, 5 and 10 percent respectively.
Table 4. Diagnostic tests results.
Table 4. Diagnostic tests results.
Endogeneity TestProb-Value
J-statistic25.436(0.1466)
Instrument Rank19
Difference in J-stats3.4283(0.9047)
Restricted J-statistic31.460
Unrestricted J-statistic28.032
Autocorrelation
Prob. Chi-Square0.1830
Heteroscedasticity Test
F-Statistic0.0524(0.4225)
Normality Test
Jarque–Bera Test2.004(0.367)
Table 5. Short Run and cointegration coefficient.
Table 5. Short Run and cointegration coefficient.
Model 1Model 2Model 3
VariablesCoefficientProbCoefficientProbCoefficientProb
D(LTFP(−1))0.09530.54360.32320.62310.52190.2710
D(LAL)0.25420.0494 **0.25240.0917 *−0.07020.2808
D(LAL(−1))0.04370.72400.21150.7429−0.02260.6749
D(LAEMP)0.04090.1120−0.01140.83990.02710.5102
D(LAEMP(−1))0.38260.3411−0.04350.4813−0.10380.0067 ***
D(LAI)0.01820.8102−0.00020.94750.12970.2510
D(LAI(−1))0.07670.42680.01470.0255 **−0.08610.4476
D(LHC)0.74680.13560.01420.0717 *0.01060.3214
DLHC(−1))0.72190.13170.01850.0743 *0.04910.0018 ***
D(LRF)0.00450.0251 **0.00560.74570.47050.1002
D(LRF(−1))−0.00150.22160.00540.7366−0.68950.0131 **
D(INT)−1.48140.1305−0.03100.0684 *−2.46240.0077 ***
D(INT(−1))−1.41690.1264−0.04500.0552 *−0.08150.9539
D(LAT)0.97250.3580
D(LAT(−1))0.49620.0842 ***
D(LSD) 0.15920.0083 ***
D(LSD(−1)) 0.17000.0136 ***
D(LFCT) 0.69580.3146
D(LFCT(−1)) −0.88480.2822
ECM (−1)−1.21280.0001 ***−0.55360.0019 ***−1.15770.0000 ***
Autocorrelation (Breusch–Godfrey)0.0862
(0.9176)
0.5222
(0.5983)
2.2928
(0.1376)
Heteroskedasticity (Breusch–Pagan–Godfrey)1.2323
(0.3035)
1.3594
(0.2341)
0.3984
(0.9840)
Normality (Jarque–Bera)2.6202
(0.2697)
4.3351
(0.1035)
1.8697
(0.3925)
***, **, * represents the level of significance at 1, 5 and 10 percent respectively.
Table 6. Results of Two Stage Least Square.
Table 6. Results of Two Stage Least Square.
VariableCoefficientp-Values
C1.84280.0138 **
LAIt0.01610.0005 ***
LAEMPt−0.08830.4002
LALt0.21510.0173 ***
LHCt0.69500.3063
LRFt0.09760.0000 ***
INTt−0.04530.0038 ***
LATt0.07310.0637 *
LSDt0.12090.0392 **
LFCTt0.07180.0189 **
F-statistic133.126(0.000) ***
R-squared0.93
***, **, * represents the level of significance at 1, 5 and 10 percent respectively.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Usman, M.; Hameed, G.; Saboor, A.; Almas, L.K.; Hanif, M. R&D Innovation Adoption, Climatic Sensitivity, and Absorptive Ability Contribution for Agriculture TFP Growth in Pakistan. Agriculture 2021, 11, 1206. https://doi.org/10.3390/agriculture11121206

AMA Style

Usman M, Hameed G, Saboor A, Almas LK, Hanif M. R&D Innovation Adoption, Climatic Sensitivity, and Absorptive Ability Contribution for Agriculture TFP Growth in Pakistan. Agriculture. 2021; 11(12):1206. https://doi.org/10.3390/agriculture11121206

Chicago/Turabian Style

Usman, Muhammad, Gulnaz Hameed, Abdul Saboor, Lal K. Almas, and Muhammad Hanif. 2021. "R&D Innovation Adoption, Climatic Sensitivity, and Absorptive Ability Contribution for Agriculture TFP Growth in Pakistan" Agriculture 11, no. 12: 1206. https://doi.org/10.3390/agriculture11121206

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop