State Regulation of the Digital Transformation of Agribusiness in the Context of the Climate Crisis Intensification

Abstract

The research states that the exacerbation of the climate crisis observed in recent years is accompanied by an increase in ground-level temperatures, natural disasters, loss of water resources, and other extreme weather events, which significantly impact the economy, water, and food security of water-dependent countries and the expected consequences shortly. For this purpose, during this research, data from the Bureau of National Statistics of the Agency for Strategic Planning and Reforms of the Republic of Kazakhstan were studied, and a sample of private indicators of the country’s agribusiness digitalization potential was made, which were further normalized to construct a mathematical model of the correlation between the level of digitalization of the agricultural sector and the volume of water consumed by agribusiness. The feasibility of using agricultural notes (electronic agricultural receipts) in Kazakhstan’s agribusiness as an innovative tool for attracting funds to develop agricultural production is justified. It is highlighted that the agricultural note has the potential to become a successful tool for attracting funds for the digitalization of the agricultural sector, provided it acquires the status of a full-fledged market product, in which state regulation of Kazakhstan’s agribusiness digital transformation plays a significant role.

1. Introduction

The aggravation of the climate crisis observed in recent years is accompanied by an increase in surface temperatures, natural disasters, loss of water resources, and other extreme weather events which have a rather painful impact on the economy, water, and food security of the water-dependent countries of Central Asia, which includes the Republic of Kazakhstan (RK).

Kazakhstan is the ninth largest country in terms of area, but for many years, it has remained water-dependent with a gradually increasing problem of water deficit, because firstly, seven out of eight river basins in Kazakhstan are transboundary, i.e., most of the water comes from neighboring countries; secondly, Kazakhstan is located in an arid, dry climate zone, with the southern and eastern regions of the country in mountainous and forested zones, the western regions in the zones of the Caspian Sea and river basins, and the central and northern regions in steppe zones with minimal vegetation and water deficit, which is becoming more pronounced with climate change; thirdly, over the past 50–60 years, with the increase in temperature, the glaciers of the Central Asian mountains have retreated more than one kilometer and lost almost a third of their volume, which has had negative consequences for the water potential of the country (Suleimenova 2023) (in particular, the shallowing of rivers such as the Tobol, Irtysh, Yesil, etc.).

Understanding the importance of solving this problem, the Government of Kazakhstan has been actively working for several years on coordinated and integrated basin-based water management and transboundary cooperation within the framework of the Convention on the Protection and Use of Transboundary Watercourses and International Lakes (1992), which Kazakhstan joined in 2001, the Convention on the Law of the Non-navigational Uses of International Watercourses (1997), ratified by Kazakhstan in 2024 (Information Portal of the Regional Environmental Center for Central Asia 2024), the Joint Statement of the Governments of the C5+1 Countries as of 25 September 2023, and other interstate agreements (US Embassy and Consulate in Kazakhstan 2023).

However, the reduction in the volume of water reaching Kazakhstan from neighboring countries because of climate change has reached a critical level. This necessitates the implementation of more decisive measures to avert a water and food catastrophe (Eldesov 2021). According to World Bank forecasts, by 2030, the volume of fresh water in Kazakhstan will decrease by five times, reaching 23 km³, which is comparable to the annual water consumption rate (Chynybaeva 2023). This implies that in the forthcoming years, the water deficit in Kazakhstan will amount to approximately 15% of the total water intake (Nurbay 2022). This should be regarded as a genuine threat to water and food security, uncontrolled migration, the economic development of the country, and the potential for conflict in the region (UNDP 2021).

Of particular concern is the anticipated launch of the Kosh-Tepa channel in Afghanistan in 2028, which experts predict will precipitate a water crisis in the region. However, as T. Oroskulov correctly asserts, “The fundamental issue of water scarcity in Central Asian countries is not the quantity of water available, but rather the volume of consumption and losses” (Matyashova 2023).

Statistical data indicate that the primary consumer of water in the RK is irrigated agriculture, which annually accounts for approximately 70% of the total water intake (Bureau of National Statistics 2023). Concurrently, water loss in primary, inter-farm, and intra-farm canal systems due to the high degree of wear of the irrigation infrastructure is estimated to be approximately 50% (Skakova 2023). As outlined in the report “Efficient Irrigation and Water Conservation in Central Asia,” (Vinokurov et al. 2023) which was authored by experts from the Eurasian Development Bank, the rational use of water reserves and the elimination of water losses would enable Kazakhstan to not only reduce the volume of its shortage but also to form a strategic reserve for the future. Consequently, in 2024, the Government of Kazakhstan approved the Concept for the Development of the Water Resources Management System of the RK for the period 2024–2030. This document outlines a number of urgent measures for the construction of reservoirs and the reconstruction of hydraulic structures, irrigation systems, and group water pipelines. These measures are designed to reduce unproductive water losses during transportation from 50% to 25%. This reduction will increase the water resources available to the RK by 2.4 cubic kilometers annually (Skakova 2023).

Nevertheless, the implementation of this concept will only mitigate the existing problem of water loss but will not resolve it entirely. The achievement of near-zero losses is possible only through the implementation of innovative digital technologies designed to conserve water. Consequently, the modification of regulatory frameworks for the digital transformation of agribusiness within the context of the state’s strategic policy should be accorded particular priority.

Therefore, this study is focused on finding possible ways to modify the mechanism of the digital transformation of agribusiness in the Republic of Kazakhstan through the introduction of innovative tools of state regulation.

The goal of this scientific research is to modify the mechanism of state regulation of the digital transformation of agribusiness of the RK by expanding the list of its toolkit with “agrarian notes”—a tool for attracting funds to the agrarian sector of the economy for the implementation of the Water Resources Management Program of the RK for 2020–2030 and the creation of “smart irrigation”.

To achieve this goal, the following three tasks were formulated. The first is to establish the factors that constrain the development of agribusiness in Central Asian countries in the context of the worsening climate crisis. The second is to analyze the policy of state regulation of water intake by agribusiness in Kazakhstan and the possibility of creating “smart irrigation”. The third is to search for ways to modify the mechanism of state regulation of the digital transformation of agribusiness.

The structure of the article includes an introduction, analysis of literary sources, goals and objectives of the study, materials and methods, results of the development of a mechanism of state regulation of the digitalization of the transformation of agribusiness in the context of the aggravation of the climate crisis, a discussion of the research results, and conclusions.

2. Literature Review

An analysis of publications gives grounds to assert that the topic under consideration has a fairly solid research base, as evidenced by a number of annual reports of the World Meteorological Organization, The World Bank, the Eurasian Development Bank, reports of rating and analytical centers, scientific publications, and speeches. In particular, Tri Waluyo substantiates that the use of digital technologies in the agricultural sector has a significant positive effect since the development of precision farming based on remote sensing, the introduction of integrated databases and cloud services, mobile solutions, control and accounting sensors, and other digital elements contribute to the operational efficiency of agribusiness and its sustainable development (Waluyo 2024). The author emphasizes that increasing the interest and motivation of farmers in the implementation of digital technologies should be carried out through co-financing of expenses in the form of subsidies from the state, the introduction of “digital” benefits, and tax reduction (Waluyo 2024).

Zhou et al. (2023), using 30 Chinese provinces as an example, highlight the enormous practical value of agribusiness digitalization in improving the total factor productivity of green agriculture (AGTFP). Analyzing a number of strategic documents of the Central Government of China (2022), namely the “Digital Countryside Development Strategy”, the “Outline of the Digital Countryside Development Strategy”, the “Digital Rural Development Action Plan (2022–2025)”, and the results of their implementation, the authors emphasize the role of government incentives for the digitalization of the agricultural sector as a driving force for the development of green agriculture and the creation of digital villages (Zhou et al. 2023).

Wang et al. (2023, 2024), having studied the provinces of the PRC, prove that the creation of digital villages and the digitalization of agribusiness significantly accelerate the economic development of not only the agricultural sector but also the territories. Among the mechanisms for stimulating the digital transformation of agribusiness, the authors recognize state assistance to entrepreneurship (Wang et al. 2023; Wang et al. 2024).

Pashkov and Mazhitova (2021), emphasizing the distinctive features of the digital transformation of agriculture in Kazakhstan and the North Kazakhstan region (the oldest region of dryland farming), argue that the main measures for further digitalization of agribusiness in the RK should be innovative and paternalistic (state stimulation of agro-formations, etc.) measures.

Research by Dankova et al. (2022), having studied the level of digitalization of irrigated agriculture in Central Asia, substantiates the need for preferential lending and state support for the digital modernization of irrigation in the region in order to reliably deliver (without losses) irrigation water and organize a high level of control over its consumption.

Vinokurov et al. (2021, 2022, 2023) focus on the mechanism of subsidizing the introduction of digital technologies in irrigation, emphasizing the effectiveness of its application using the example of water-dependent European countries.

Berbel et al. (2019) substantiate the feasibility of digitalizing the process of water consumption by agribusiness as a fundamental basis for implementing the country’s environmental fiscal policy using the example of EU countries with both abundant water potential (in particular, Germany, Denmark, the Netherlands, etc.) and its deficit (in particular, Portugal, Spain, Italy, etc.). The authors emphasize that the accumulated environmental tax should be redistributed between budget levels and partially directed to the digital transformation of irrigation systems of agribusiness.

Thus, we come to the conclusion that in most cases, indirect regulation instruments are used to regulate the process of the digitalization of agribusiness, such as state support, subsidies, preferential lending, fiscal regulation, etc. However, in the context of the polycrisis that has been observed in recent years, their application is complicated by the state budget deficit. Therefore, the modification of the mechanism of state regulation of the digital transformation of agribusiness by expanding the list of its instruments in the context of the aggravation of the climate crisis is of particular importance.

3. Methodology

The main hypothesis is that the modification of the state mechanism of the digitalization of water use in agribusiness will create “smart irrigation” and the topic will thereby ensure a significant reduction in water losses through timely measures for resource-saving use of the country’s water potential. The materials for this study were data from the United Nations Economic Commission for Europe, the Information Portal of the Regional Environmental Center for Central Asia (2024), the Eurasian Development Bank, the Information Portal of the Regional Environmental Center for Central Asia, the Agency for Strategic Planning and Reforms of the RK, the Bureau of National Statistics, as well as the Official Information Resource of the Prime Minister-Ministry of the Republic of Kazakhstan (2024).

In light of the findings of the aforementioned studies and the availability of data on the digitalization of agribusiness in Kazakhstan, the parsing method was selected to ascertain the comprehensive index of digitalization in the RK, which involves the collection of data from web resource pages.

Evaluating the impact of digitalization on the resource efficiency of agribusiness, with a particular focus on the rationality of fresh water utilization for irrigation, we apply the entropy method to assess data variability and uncertainty because of the varying scales and comparability issues of the statistical data presence (Han et al. 2012; Zong et al. 2021). This approach standardizes the data into a normalized form, determines the significance of each indicator, and constructs an integral index reflecting the overall situation based on multiple factors.

As highlighted by recent studies (Dou et al. 2023; Zhou 2022; Rather et al. 2023), working with small datasets in machine learning poses significant challenges, including limited data diversity, noise, imbalanced distributions, and high dimensionality, which can reduce the generalizability and effectiveness of models, as they struggle to learn and generalize from limited data. Unlike humans, who can excel at learning from a few observations, artificial intelligence models struggle to learn and generalize effectively from scarce data (Jin et al. 2022). The lack of sufficient data often hinders prediction accuracy, especially with imbalanced samples (Safonova et al. 2023). Sample size is a crucial factor in time series analysis and regression modeling because it impacts the reliability and validity of the results (Hecht and Zitzmann 2020). However, we are constrained by the availability of open-source data and unfortunately do not have the means to increase the sample size. Advanced techniques like transfer learning and data augmentation are typically required to improve performance with small datasets (Jin et al. 2022; Zhou 2022).

In time series regression, limited data can lead to inaccurate model fitting, poor generalization (Hyndman and Athanasopoulos 2018), and increased risk of overfitting. Models like ARIMA or SARIMA may struggle with unstable or biased parameter estimates, and traditional validation techniques may be less effective (Box et al. 2015; Woodward et al. 2017; Montgomery et al. 2015; Shmueli and Polak 2024).

By quantifying the uncertainty inherent in the data, the entropy method allows decision-makers to assign weights that reflect the reliability or importance of each factor (Zhu et al. 2020). In environmental and socio-economic assessments, where uncertainty is prevalent, the entropy method is often used to assign weights to different indicators or factors based on their contributions to the total uncertainty of the system (Zhang et al. 2023).

The choice of the entropy method over other methods can be justified based on several factors. The entropy method provides a measure of uncertainty or randomness in a dataset. In digital agribusiness, decisions often need to be made despite uncertain data. The entropy method provides a way to quantify and incorporate this uncertainty into decision-making processes. Agribusiness data come from diverse sources, such as satellite imagery, IoT sensors, and market reports. These sources often produce data of varying quality and reliability. Entropy helps in managing this heterogeneity by identifying which data sources are more consistent and which are more variable, aiding in the selection of reliable data for analysis (Wihartiko et al. 2023).

Agribusiness operates in dynamic environments where conditions change rapidly (e.g., weather, pest outbreaks). The entropy method can be used to continuously monitor the uncertainty in the data, allowing for adaptive management strategies that respond to changing conditions in real time. Unlike some methods that may only focus on mean values or variance, entropy captures the full distribution of the data, making it a more comprehensive measure of uncertainty. This is particularly important in agribusiness, where understanding the full range of potential outcomes is crucial. In addition, the entropy method does not assume any specific distribution for the data, making it a flexible tool for analyzing agribusiness data, which may not follow standard distributions. Agribusiness digitalization data can be both discrete (e.g., number of farms adopting a technology) and continuous (e.g., rainfall measurements). The entropy method can be applied to both types of data, providing a unified approach to handling different data forms.

Normalization of the indicators, i.e., bringing them to a single scale, was carried out using the following formula:

$${\tilde{x}}_{ti}={\displaystyle \frac{x_{ti}}{{\displaystyle {\sum }_{t=1}^{10}{x}_{ti}}}}$$

(1)

where $t$ is a level of the series of the indicator under consideration.

Entropy $E_i$ for each indicator $i$ is calculated as

$$E_i=-{\displaystyle \frac{1}{\ln 10}}{\displaystyle {\sum }_{t=1}^{10}{\tilde{x}}_{ti}\ln {\tilde{x}}_{ti}}$$

(2)

The information coefficient $d_i$ for each indicator $i$ is calculated as

$$d_i=1-E_i$$

(3)

The weight of each indicator $w_i$ is determined as

$$w_i={\displaystyle \frac{d_i}{{\displaystyle {\sum }_{i=1}^m{d}_i}}}$$

(4)

where $m$ is a number of indicators.

The integral indicator $I_{tj}$ for each level $t$ is calculated as the weighted sum of normalized values:

$$I_{tj}={\displaystyle {\sum }_{t=1}^{10}{w}_i}{\tilde{x}}_{ti}$$

(5)

where $j=1,2$ are indices of potentials taken into account for the integral indicator of digitalization.

$I_1$ is an integral indicator of technical potential of the digitalization of the agricultural sector, and $I_2$ is an integral indicator of human resources potential for digitalization of the agricultural sector.

Nevertheless, in order to ascertain the extent to which the agricultural sector is prepared for digitalization, it is more appropriate to utilize a comprehensive indicator of the digitalization of agribusiness.

The comprehensive integral indicator of the digitalization $I_{cifrat}$ for each level $t$ will be calculated as the arithmetic mean of the values of the integral indicators:

$$I_{cifrat}=(I_1+I_2)/2$$

(6)

Missing data in statistical compilations are obtained via linear interpolation using adjacent values. Linear interpolation is a commonly used method to handle missing data, especially when the amount of missing data is small. In our case, when the missing data points are few and scattered and the gaps between the known data points are small, linear interpolation can provide a good approximation. This is because the assumption of a linear trend is more likely to hold over short intervals. Linear interpolation maintains the overall trend of the data without introducing significant biases or distortions. This is important when the goal is to preserve the underlying patterns in the dataset, such as gradual increases or decreases over time. It was chosen because of its minimal complexity; it is computationally simple and easy to implement, making it a practical choice for handling missing data. Unlike other imputation methods, such as polynomial interpolation or spline methods, linear interpolation does not require making strong assumptions about the nature of the data beyond the assumption of linearity between points. This makes it a safer and more conservative choice when only a few data points are missing.

Regression analysis is used to infer causal relationships between the independent and dependent variables: the consumption of fresh water by agribusiness is dependent upon a number of factors, from which we take into account the level of digitalization, stress load on water resources, climate changes, and gross output of agricultural products. Therefore, the development of the mathematical model here is based on regressive analy-sis, as exemplified by the approach taken by Ohanisian et al. (2022).

The generalized least squares (GLS) method was used to find the coefficients of the nonlinear multivariate model. This method allows for efficient estimation of model parameters by accounting for potential heteroscedasticity or correlations in the residuals, making it particularly useful in complex regression models (Gujarati 2021). GLS is employed to improve statistical efficiency and reduce the risk of drawing erroneous inferences, as compared to ordinary least squares and weighted least squares methods.

To determine the most suitable form of the nonlinear models, we employed the method of directed enumeration, which systematically tests a variety of functional forms, including linear, multiplicative, polynomial, exponential, and logarithmic dependencies. By calculating the deviation potentials (Ohanisian et al. 2022; Shyshkanova 2018), we were able to identify which model best represented the underlying relationship between the variables. The selection of the optimal model was further refined by calculating the criterion for the best fit to real-world data, following the guidelines of Greene (2018). This criterion relies on evaluating the goodness-of-fit through a weighting function applied to each predicted value in the partial likelihood estimation equation (Wahid et al. 2021). The aim is to minimize the sum of squared deviations to assess how well the model’s predictions align with observed values. After applying the inverse nonlinear transformation, the model that exhibited the smallest sum of squared deviations and a determination coefficient closest to one was selected, ensuring both the robustness and accuracy of the predictions. The calculation algorithm was implemented by the authors in Python 3.0.

The following tests were carried out: the Kwiatkowski–Phillips–Schmidt–Shin (KPSS) test to check the stationarity of the series; the Farrar–Glauber Chi-square test for detecting the strength of the multicollinearity over the whole set of explanatory variables; the Durbin–Watson test and the Q-Ljung-Box to detect the presence of autocorrelation; and the Spearman’s test to detect the presence of heteroscedasticity. The model validation was also conducted using an F-test.

4. Results of the Search for Ways to Modify the Mechanism of State Regulation of the Digital Transformation of Agribusiness in the Context of the Climate Crisis

In the context of the deepening climate crisis, the issue of water resource deficit is becoming increasingly acute, especially in the countries of Central Asia, among which Kazakhstan is recognized as the most vulnerable. In this region, the temperature is rising faster than the global average, leading to a reduction in the area of glaciers, which entails a decrease in river runoff and creates risks for agriculture and food security in the region. Therefore, the water management policy of the Central Asian countries is subordinated to the interests of agriculture and the provision of irrigated lands with irrigation water. In the country structure of water consumption for irrigation needs in Central Asia, Kyrgyzstan (92.7%) and Uzbekistan (92.3%), historically specializing in cotton cultivation, dominate. They are followed by Tajikistan (74.5%), Kazakhstan (62.7%), and Turkmenistan (61.4%); see

Table 1. Indicators of irrigated land and water intake for the needs of agribusiness at the end of 2022.

	Unit of Measurement	World	Central Asia	Kazakhstan	Kyrgyzstan	Tajikistan	Turkmenistan	Uzbekistan
Agricultural lands	million hectares	4744.46	283.81	214.00	10.37	4.92	33.84	25.69
Cultivated land (arable and pasture)	million hectares	1561.45	38.53	29.68	1.36	1.04	2.00	4.44
Irrigated lands	million hectares	331.71	10.11	2.243	1.03	0.76	1.80	4.28
% of cultivated land	%	21.24	26.25	7.56	75.14	73.29	90.00	96.45
Realized potential of irrigated lands	%	63.92	68.03	59.54	45.62	48.28	76.50	87.09
Precipitation	mm/year	1170.66	368.2	250	533	691	161	206
Internal renewable water resources	km3/year	42,808.60	194.49	64.35	48.93	63.46	1.41	16.34
Total water intake	km3/year	4031.86	127.27	24.56	7.66	9.90	26.24	58.90
Water withdrawal for agricultural needs relative to total water withdrawal	%	-	78.9	62.7	92.7	74.5	61.4	92.3

Source: Vinokurov et al. (2023).

.

However, among the countries of Central Asia, the most critical situation is in Kazakhstan since its water resources are quite limited compared to other states. Due to geographical peculiarities, seven out of eight water basins of Kazakhstan are transboundary, as a result of which Kazakhstan is largely dependent on the water policy of neighboring countries (People’s Republic of China, Russian Federation, and Central Asian countries).

In this respect, the Aral-Syrdarya (91%), Zhaiyk-Caspian (82%), Shu-Talas (74%), and Balkhash-Alakol (48%) water basins are the most vulnerable, and the Tobyl-Torgai (12%) and Ertis (20%) water basins are the least vulnerable (Vinokurov et al. 2023), the degree of water resources use of which is characterized by the data of

Table 2. Water resources utilization level within the framework of water management river basins of the RK in 2022.

Water Management Basins	Water Resources (km3) *		Water Intake, (km3)	Pressure on Water Resources Kuse ***
	Local	Total (Local and Cross-Border)		Local, (%/category)		Total (Local and Cross-Border), (%/category)
Aral-Syrdarya	2.16	18.7	10.7	495.4	V	57.2	IV
Balkhash-Alakol	16.5	29	4.1	24.8	III	14.1	II
Ertis	26.5	33.5	3.8	14.3	II	11.3	II
Esil	2.7	2.7	0.4	14.8	II	14.8	II
Zhaiyk-Caspian	3.1	11	2.4	77.4	V	21.8	III
Nura-Sarysu **	0.9	1.6	1.4	155.6	V	87.5	V
Tobyl-Torgai	1.7	2.1	0.1	5.9	I	4.8	I
Shu-Talas	1.0	3.7	2.1	210.0	V	56.8	IV

Notes: * River water resource indicator values are taken for average water years; ** without taking into account water flowing through the Kanysh Satpayev Canal; *** I—low load, II—moderate; III—high, IV—very high, V—critical. Source: Vinokurov et al. (2023).

.

In 2023, the Ministry of Agriculture of the RK drew attention to the threat of water shortages for irrigation of water-intensive crops in Kyzylorda and Turkestan provinces, which depend on the water resources of the Aral Sea basin. In Zhambyl province, farmers began experiencing water shortages in mid-April 2023 as a result of the Ismail canal closure. In Atyrau province, the main source of water is the Zhaiyk (Ural) River, along which 80% of farms are located, but it is becoming more and more difficult to bring water to consumers every year because the infrastructure built in the 1960s is deteriorating.

Due to the lack of irrigation water in Kazakhstan, the area planted with rice has been reduced for the second consecutive year. It is also planned to reduce the area allocated for cotton cultivation in favor of crops that require less water (Vinokurov et al. 2023). In light of the anticipated rise in surface air temperature across Kazakhstan, with an expected increase of 0.8–1.2 °C per annum, it is likely that there will be an increase in evaporation of moisture in river catchments. A reduction in the quantity of water flowing into rivers, a decline in the level of the Caspian Sea and Lake Balkhash, an increase in the demand for water resources for economic purposes, and an expansion of irrigation on irrigated lands will result in a rise in the real threat to both the food security of the country and its sustainable development (

Table 3. Forecast values of river flow resources of the Republic of Kazakhstan taking into account climate change and anthropogenic pressures by 2030, km³.

Water Management Basins	Local Resources		Inflow		Summarized
	Total	Including Outflow Outside the RK (Return Outflow)	Total	Including Those Formed in the Territory of Neighboring Countries	Total
Aral-Syrdarya	3.17	0.48	14.4	13.9	17.1
Balkhash-Alakol	16.6	0.99	12.5	11.5	28.1
Ertis	26.5	1.31	7.13	5.82	32.3
Esil	2.47	-	-	-	2.47
Zhaiyk-Caspian	3.08	0.97	8.63	7.66	10.7
Nura-Sarysu	1.96	-	-	-	1.96
Tobyl-Torgai	1.88	-	0.59	0.59	2.47
Shu-Talas	1.01	-	3.21	3.21	4.22
Total:	56.7	3.75	46.5	42.7	99.4

Sources: On Approval of the State Program for Water Resources Management of the Republic of Kazakhstan for 2020–2030 (2024), On Approval of the Concept for the Development of the Water Resources Management System of the Republic of Kazakhstan for 2024–2030 (2024).

).

The majority of water withdrawal in Kazakhstan is attributed to the agricultural sector, with 67% of the total volume being utilized by agribusiness. The total area of irrigated lands in 2022 was 2243.4 thousand hectares, of which only 1144.3 thousand hectares were in use. This implies that approximately 50% of the infrastructure is in a state of disrepair and requires maintenance and repair. The surface (furrow) method of crop irrigation is employed on approximately 80% of the utilized lands. An area of 185.8 thousand hectares is irrigated by sprinkling, while an area of 73.0 thousand hectares is irrigated by drip irrigation. The irrigation systems are operated with equipment that is no longer state-of-the-art. The following irrigation systems are also in use: DM “Fregat”, “Dnepr”, “Volzhanka”, and others (Vinokurov et al. 2023). Consequently, between 2020 and 2022, the average volume of water withdrawn for agricultural purposes was 11.9 cubic kilometers. Of this total, 77% was utilized for regular irrigation on an area of 1.9 million hectares, while the remaining 2.7 km³ was distributed between limanic irrigation, hayfield flooding, agricultural water supply, and pasture watering (

Table 4. Dynamics of water withdrawal for regular irrigation by water management basins of the RK during 2010–2022.

No	Water Management Basins	Years
		2010–2019			2020			2021			2022
		Irrigated Lands, thousand ha	Water Withdrawal, million. m3	Specific Consumption, thousand m3/ha	Irrigated Lands, thousand ha	Water Withdrawal, million. m3	Specific Consumption, thousand m3/ha	Irrigated Lands, thousand ha	Water Withdrawal, million. m3	Specific Consumption, thousand m3/ha	Irrigated Lands, thousand ha	Water Withdrawal, million. m3	Specific Consumption, thousand m3/ha
1	Aral-Syrdarya	606	7154	11.8	642	7456	11.6	536	6920	12.9	590	6781	11.5
2	Balkhash-Alakol	396	3111	7.8	456	3401	7.5	453	3310	7.3	312	3347	10.7
3	Ertis	66	227	3.4	49	165	3.4	48	158	3.3	52	174	3.3
4	Esil	4	8	1.9	6	10	1.8	9	14	1.5	8	5	0.6
5	Zhaiyk-Caspian	10	46	4.4	12	46	3.9	12	48	4.1	15	47	3.2
6	Nura-Sarysu	19	75	3.9	24	74	3.1	9	74	7.9	20	74	3.7
7	Tobyl-Torgai	6	13	2.3	6	13	2.2	7	13	2.0	8	20	2.3
8	Shu-Talas	145	1075	7.4	84	936	11.2	104	936	9.0	139	1040	7.5
	Total	1254	11,710	9.3	1277	12,101	9.5	1177	12,101	10.3	1144	11,489	10.0

Note: According to reports of basin inspections on regulation of water resources use and protection of the Water Resources Committee of the Ministry of Water Resources and Irrigation of the RK. Sources: Kazakhstan in figures (2023) and Bureau of National Statistics (2023).

).

Studies (Vinokurov et al. 2023) have demonstrated that approximately one-third of the volume of water consumed by agribusiness is lost in the main and inter-farm canal systems due to the high percentage of wear and tear of water management fixed assets. The average age of irrigation inter-farm and on-farm infrastructure is more than 50 years, with large main canals being even older. The unsatisfactory technical condition of the irrigation infrastructure has a detrimental impact on the quality of services provided by state water management organizations (main and inter-farm irrigation structures and canals) and irrigated land owners (on-farm structures and irrigation networks). This ultimately results in significant economic losses.

According to estimates, the annual economic damage from irrational water use for irrigation is close to 3% of GDP. It is estimated that each dollar invested in water-saving technologies gives about USD 5 of return (Vinokurov et al. 2023). However, small agro-companies as well as farmers are not able and not motivated enough to implement digital water-saving technologies.

Today, out of 1.9 million hectares of irrigated land, water-saving technologies are applied only on 16% (Figure 1). At the same time, in the southern regions of the country, which account for the major share of irrigated agriculture, the level of implementation of water-saving technologies is only 3% of the total area of irrigated land. That is extremely unacceptable in the current realities since with the growth of water withdrawal, water losses are also increasing (in particular, according to the data of 2022, the share of water losses in irrigated agriculture amounted to about 50%) (Kydyrbaeva 2024).

Moreover, in regular irrigation, there has been a tendency to increase water consumption rates from 8.5 m³/ha in 2010 to 10 m³/ha in 2022. According to the Kazakh Research Institute of Water Management, this is due to climate change. Thus, on average, the number of irrigations during the growing season has increased from five to six.

It is imperative that immediate and decisive action be taken to rectify the current situation (Dvigun et al. 2022). Kassym-Jomart Tokayev, Head of the State, emphasized the need to prioritize the acceleration of the implementation of various economic initiatives (Economic Course of a Fair Kazakhstan 2023). The implementation of advanced digital water-saving technologies will result in an increase in the area of irrigated land using water-saving technologies to 2.5 million hectares by 2030. This will entail the irrigation of up to 150,000 hectares per year (Official Information Resource of the Prime Minister-Ministry of the Republic of Kazakhstan 2024). This will result in the annual saving of approximately 2.1 km³ of water, while also increasing the yield of agricultural products by a factor of 1.5 to 2 (Kydyrbaeva 2024). However, the path to the digitalization of Kazakhstan’s agribusiness is fraught with difficulty, despite the fact that according to the report published by the Global CIO portal in 2023, Kazakhstan entered the top 30 most digitally developed countries in the world in 2022, took 51st place in the world ranking according to the ICT Development Index, and 58th place according to the Network Readiness Index. In order to encourage farmers to utilize water-saving technologies, the Ministry of Agriculture of the Republic of Kazakhstan has initiated a program of subsidies for the creation of irrigation systems and the purchase of drip irrigation equipment. This program is part of the Water Resources Management Program for the period 2020–2030. Furthermore, the proportion of subsidies for expenses related to irrigation equipment, including sprinkler irrigation equipment, has been increased from 50% to 80% (Kydyrbaeva 2024). Additionally, the construction of new reservoirs with a capacity of 5–7 km³ has been initiated, and the digitalization process of irrigation has been activated. This process is expected to reduce water consumption per 1000 GDP from 91.2 to 73.0 m³.

Nevertheless, when making decisions regarding the implementation of certain measures, it is essential to have a clear understanding of the readiness of agribusiness for digital transformation. In order to ascertain this, we conducted a study of the data provided by the Bureau of National Statistics of the Agency for Strategic Planning and Reforms of the RK. We then selected a number of indicators of the digitalization potential of the country’s agribusiness (

Table 5. Key indicators of the digitalization potential of Kazakhstan’s agribusiness during 2013–2022.

Years	Number of Reporting Agricultural Companies	Resource Potential of Digitalization of Agribusiness in Kazakhstan
		Technical							Human Resources, Persons
		Use of ICT	Digitalization of Communication with Government Agencies	Digitalization of Business Processes				Digitalization of Product Sales	Number of ICT Specialists	Number of Employees Using the Internet
				Using Cloud Computing	Business Process Automation	Use of RFID Technology	Electronic Documentation
2013	5308	1731	1349	4	3	-	871	89	188	5174
2014	5414	1956	1623	7	4	3	1073	146	192	6212
2015	5449	2177	1856	12	8	7	1377	187	120	7862
2016	6049	2389	2119	19	13	12	1572	231	403	9764
2017	6467	2254	2541	89	24	14	1844	248	312	10,468
2018	7112	3573	2739	162	43	18	2148	364	156	11,680
2019	6505	3893	3094	193	86	22	2540	360	238	13,377
2020	7136	4375	3481	288	57	31	3032	268	195	14,897
2021	7589	4459	3792	354	73	27	3461	235	331	16,133
2022	8521	5158	4377	245	428	52	4260	284	262	16,839

Sources: Kazakhstan in figures (2023) and Digital Kazakhstan (2024).

). To conduct a comprehensive analysis of digitalization in the agrarian sector of Kazakhstan, we utilized publicly available statistical data. These data points are divided into two clusters representing different types of potential: technical (x₁—use of ICT; x₂—digitalization of communication digitalization with government agencies; x₃—using cloud computing; x₄—business process automation; x₅—use of RFID technology; x₆—electronic documentation; x₇—digitalization of product sales) and human resources (x₈—number of ICT specialists; x₉—number of employees using the Internet), each of which, to one degree or another, characterizes the ability and readiness of agribusiness for digital transformation.

The dynamics of the digitalization of agribusiness in Kazakhstan are characterized by the data presented in

Table 6. Dynamics of indicators of the digitalization of agribusiness in Kazakhstan during 2013–2022.

Years	Integral Indicator of Technical Potential	Integral Indicator of Human Resources Potential	Comprehensive Integral Indicator
2013	0.009	0.062	0.035
2014	0.015	0.067	0.041
2015	0.024	0.060	0.042
2016	0.034	0.126	0.080
2017	0.056	0.111	0.083
2018	0.087	0.085	0.086
2019	0.124	0.109	0.116
2020	0.132	0.108	0.120
2021	0.150	0.141	0.146
2022	0.370	0.130	0.250

.

Since the objective of this study is to assess the impact of the level of digitalization of agribusiness on the volume of fresh water withdrawal for agricultural needs in Kazakhstan, based on statistical data available in the public domain (

Table 7. Dynamics of indicators of agribusiness load on water resources during 2013–2022.

Years	Fresh Water Withdrawal by Agribusiness, Million m3	Water Resources Use Indicators
		Stress			Climate
		Irrigated Land Area, Thousand Hectares	Fresh Water Intake for Irrigation (Regular and Estuary), Million m3	Level of Stress on Water Resources, %	Average Annual Temperature, °C	Average Annual Precipitation, mm Hg
2013	16,056.5	1142.38	9486	31.25	7.20	320.7
2014	21,257.5	1128.03	9704	32.01	5.74	317.3
2015	19,691	1091.09	10165	30.04	7.16	315.7
2016	19,818.5	1309.44	9350	30.01	7.22	441.3
2017	20,952.3	1330.81	9833	31.14	7.10	313.7
2018	16,805.6	1365.49	9782	32.65	5.85	323.2
2019	19,782.7	1290.52	10300	32.62	7.29	294.2
2020	22,280.3	1277.02	9684	34.10	7.66	270.7
2021	21,298.1	1177.11	9119	34.01	7.32	271.5
2022	20,315.9	1144.21	8520	33.92	7.52	311.0

Sources: Bureau of National Statistics (2023), Annual Bulletin of Monitoring the State and Change of Climate in Kazakhstan: 2015–2022 (2023), and WMO (2024a, 2024b).

), an analysis was conducted of other indicators that have a significant impact on the use of the country’s water potential. In instances where data were absent from the statistical collections of the RK, they were obtained through linear interpolation using neighboring values.

In order to perform further calculations, two clusters were selected for analysis: the stress indicators of the load on water resources and the indicators of the impact of climate change. The indicators in the first cluster are unidirectional in nature. In contrast, the indicators within the second one are bidirectional. This is because the average annual temperature exerts a negative influence on the intake of fresh water. That is to say, as temperature increases, the demand for fresh water rises, while the renewability of natural water resources declines.

Conversely, precipitation exerts a positive influence, whereby an increase in precipitation reduces the demand for fresh water and increases renewable water resources. Consequently, the values of the average annual precipitation indicator are replaced with inversely proportional ones, thereby aligning the two indicators in terms of their influence on the water intake factor.

Next we determined the integral indicators of stress load on water resources I_stress and the climate change I_climat in the same way as we did earlier using Formulae (1)–(5). Gross agricultural output is also linked to water withdrawal. Statistical data for gross agricultural output during 2013–2022 are presented in the

Table 8. Gross agricultural output during 2013–2022, million KZT.

	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022
Gross output of agricultural products (services)	2386.1	3143.7	3307.0	3684.4	4092.3	4497.6	5177.9	6363.9	7549.8	9521.0
Gross crop production	1313.0	1739.4	1825.2	2047.6	2249.2	2411.5	2817.7	3687.3	4387.2	5808.3
Gross livestock production	1064.3	1393.8	1469.9	1621.5	1810.9	2050.5	2319.5	2637.5	3116.9	3658.8
Agricultural services	8.8	10.5	11.8	15.3	10.8	12.1	14.0	9.9	11.2	14.2

Sources: Bureau of National Statistics (2023) and Kazakhstan in figures (2023).

.

We also normalized the gross output of agricultural products ${\tilde{G}}_{agri}$ and fresh water withdrawal by agribusiness ${\tilde{W}}_{agri}$ using Formula (1). In order to evaluate the influence of digitalization and other pertinent variables on the consumption of pressed water, we proposed the following multivariate regression equation, which employs the methods described above in paragraph 3:

$${\tilde{W}}_{agri}=0.134\cdot I_{cifra}^{-0.197}\cdot {\tilde{G}}_{agri}^{0.514}\cdot {15.394}^{I_{stress}}\cdot I_{climat}^{0.022}$$

(7)

To use the GLS method, the regression model was preliminary linearized to the following view $Z=b_0+a_1{Y}_1+a_2{Y}_2+b_3{Y}_3+a_4{Y}_4$, where $a_0$, $a_1$, $a_2$, $a_3$, and $a_4$ are unknown regression coefficients defined by GSL; $Z=\ln {\tilde{W}}_{agri}$; $b_0=\ln a_0$; $Y_1=\ln I_{cifra}$; $Y_2=\ln {\tilde{G}}_{agri}$; $Y_3=I_{stress}$; $b_3=\ln a_3$; and $Y_4=\ln I_{climat}$.

Here, the KPSS test was used for checking a null hypothesis that an observable time series is stationary around a deterministic trend. We computed the KPSS test statistic relative to all transformed variables; see

Table 9. Test statistics.

	Z	${\mathit{Y}}_{\mathbf{1}}$	${\mathit{Y}}_{\mathbf{2}}$	${\mathit{Y}}_{\mathbf{3}}$	${\mathit{Y}}_{\mathbf{4}}$
KPSS test statistic	0.0576	0.0592	0.1777	0.2147	0.08450
Spearman’s test statistic	-	−0.3818	−0.3818	0.0303	−0.1151

Source: Authors’ computation.

. The critical level for it is 0.216 at a significance level $\alpha =0.01$. All the test statistics were less than the critical value, so the test failed to reject the null hypothesis. We can make a conclusion that the series considered here are stationary.

According to the Farrar–Glauber test, the chi-square statistic was calculated as ${\chi }^2$ = 16.36, and it was found to be less than the critical value of ${{\chi }_{cr}}^2$(0.99;6) = 16.81 at the desired level of significance, $\alpha =0.01$. Therefore, we cannot reject the assumption of the linear independence of the variables, so we assume that there is no problem of multicollinearity in the developed model.

The Durbin–Watson test statistic was computed as $d=0.29$. To test for autocorrelation at a significance of $\alpha =0.01$, the test statistic was compared to lower $d_l=$ 0.23 and upper $d_u=$ 2.19 critical values, so the test was inconclusive, as $d_l<d<d_u$ at $\alpha =0.01$. The Q-Ljung-Box test is an alternative method to detect autocorrelation at multiple lags (not just lag-1), making it a broader test for detecting higher-order autocorrelation. We computed the Q-statistic = 0.6663 for lag-3, and the critical value was ${{\chi }_{cr}}^2$(0.99;3) = 11.345. Therefore, the Q-Ljung-Box test failed to reject the null hypothesis that the data are independently distributed. The conclusion can be that there is no significant autocorrelation present in the data at the significance level $\alpha =0.01$.

An important assumption assumed by the classical linear regression model is that the error term should be homogeneous in nature. To detect the presence of heteroscedasticity, we computed the Spearman’s rank correlation coefficients relative to all independent variables; see Table 9. The critical value of Spearman’s rank correlation coefficient was 0.782 when $\alpha =0.01$, n = 10. In our case, heteroscedasticity was considered insignificant as all test statistics were less than the critical value, so we cannot reject the hypothesis about homoscedastisity and we can consider that heteroscedasticity is not present in the model.

Figure 2 depicts a multivariate regression (7) with fixed indicators of the missing axes at the level of 2022.

As illustrated in Figure 2 and in accordance with Formula (7), the selected indicators exhibit varying effects on the studied factor of water intake.

The calculated coefficient of determination for this model was $R^2$ = 0.782, which is close to one, indicating sufficient density between the variables. The critical value of the Fisher coefficient was $F_{kr}=$ 5.192 for a degree of freedom of $k_1=$ 4 and $k_2=$ 5, and the significance level was 0.05. The calculated value of the Fisher coefficient is $F$ = 6.321. A Fisher criterion-based assessment $F$ > $F_{kr}$ revealed that the proposed mathematical model is 95% reliable in representing the statistical data and that an economic analysis can be conducted based on the adopted model.

Notably, digitalization exerts a positive influence on this factor, indicating that the growth of digitalization contributes to reducing the water intensity of agricultural production. The gross output of agricultural products, stress load on water resources, and climate change have a negative effect on fresh water intake: That is, as these variables increase, there is an increase in fresh water consumption.

The mathematical model (7) suggests that an increase in the agribusiness digitalization indicator by 75% compared to the 2022 indicator, that is, to the level of $I_{cifra}$ = 0.438, will result in a reduction in water intake by Kazakhstan’s agribusiness by approximately 18%. This is based on the assumption that other values remain fixed at the 2022 level. Consequently, the projected fresh water intake in this case will be $W_{agri}$ = 16,452.786 million cubic meters.

In order to ascertain the necessity for financial resources, it is essential to evaluate the financial expenses associated with digitalization in agricultural enterprises $F_{cost}$ over the period between 2013 and 2022 (

Table 10. The financial expenses associated with digitalization in agricultural enterprises during 2013–2022, million KZT.

Years	ICT Costs	Including
		Purchase of Software	Software Development	Employee Training		Costs of Paying for Services of Third-Party ICT Bodies
				Total	Training in Digital Skills
2013	539.0	36.2	3.5	1.9	0.7	87.1
2014	600.5	44.3	2.2	2.4	1.3	151.5
2015	480.0	54.3	0.4	1.0	-	197.6
2016	742.8	102.0	2.7	1.6	0.9	129.0
2017	947.2	94.9	2.4	6.7	4.8	172.9
2018	948.7	81.4	2.8	3.9	1.6	179.7
2019	1027.4	144.9	2.6	3.3	1.9	219.1
2020	1421.1	158.0	0.6	6.7	5.4	299.2
2021	1413.9	133.0	15.5	4.9	3.7	453.9
2022	2428.2	194.7	21.8	8.7	3.6	972.6

Source: Digital Kazakhstan (2024).

).

The KPSS test statistic was equal to 0.1341 relative to $I_{cifra}$ and 0.1121 relative to $F_{cost}^{1/3}$, which are less than the critical value, so the test failed to reject the null hypothesis. We can make a conclusion that the series considered here are stationary.

Using the OLS method, we compiled a nonlinear dependence of the digitalization indicator on the level of financial costs:

$$I_{cifra}=-0.260+0.036F_{cost}^{1/3}$$

(8)

The Durbin–Watson test statistic was computed as $d=2.044$. As its value was more than 2, we considered $d_{*}=4-2.044=1.956$. At a significance of $\alpha =0.01$ the critical values were $d_l=$ 0.88 and upper $d_u=$ 1.32, so, as $d_u<d_{*}$, we can conclude that there is no significant autocorrelation present in the data at the significance level $\alpha =0.01$.

Spearman’s rank correlation coefficient was 0.261, which is less the critical value of 0.782 at $\alpha =0.01$, n = 10. Therefore, we cannot reject the hypothesis about homoscedasticity and heteroscedasticity is considered insignificant.

The correlation coefficient of the linearized model was $r$ = 0.971, and the determination coefficient for this model was $R^2$ = 0.942. Both values are close to one, indicating sufficient density between the variables. The critical value of the Fisher coefficient was $F_{kr}=$ 5.318 for a degrees of freedom of $k_1=$ 1, $k_2=$ 8, and a significance level $\alpha =0.05$. The calculated value of the Fisher coefficient was $F$ = 32.549, indicating a reliability of 95%. Therefore, it can be concluded that an economic analysis can be conducted on the basis of the adopted model.

Figure 3 shows the mathematical model (8), represented by a solid line. The dotted lines illustrate the 95% confidence interval of the regression, while the dots represent the statistical data.

Equation (8) demonstrates that in order to achieve an overall integrated digitalization indicator that is 75% greater than the 2022 value, it is necessary to increase digitalization costs by approximately threefold compared to 2022. This equates to KZT 7.1 billion. Otherwise, the consequences of the implementation of climate threats may be the aggravation of interstate contradictions, the development of new sources of environmental instability, the disruption of socio-economic development programs, etc. (Medeu et al. 2018).

In order to achieve this objective, the Government of Kazakhstan enacted the Law of the RK “On Amendments and Additions to Certain Legislative Acts of the RK on Communications, Digitalization, Improving the Investment Climate and Eliminating Excessive Legislative Regulation” on 21 May 2024, as Legislative Act No. 86-VIII (2024). This legislative measure provides for amendments and additions to a number of legislative acts of the RK on digitalization and improving the investment climate (On Amendments and Supplements to Certain Legislative Acts of the Republic of Kazakhstan on Communications, Digitalization, Improving the Investment Climate and Eliminating Excessive Legislative Regulation 2024).

According to the Law of the RK No. 86-VIII, as well as the Concept of Development of the Water Resources Management System of the RK for 2024–2030, the invested funds should be directed to the digitalization of at least 3500 km of irrigation canals; the digitalization of water consumption and loss accounting during transportation along main and inter-farm canals; coverage of the water management infrastructure with digital technologies; the development of the digital geoservice (flood.gharysh.kz) for flood modeling and forecasting; the development of the interactive geoinformation platform for water resources of the RK (https://test-gidro.gharysh.kz/about), etc. That is, both legislative and strategic documents of the government of Kazakhstan clearly define investment vectors. At the same time, there are no innovations regarding their attraction.

Currently, the project of the RSE “Kazvodkhoz” on the digitalization of the water channel “K-19” (2021–2025) is being implemented in Kazakhstan, which includes 119 main and economically significant main and inter-farm irrigation canals located in the Almaty, Zhambyl, Turkestan, and Kyzylorda regions, with a total water intake of 6 cubic km. The project is financed in the amount of KZT 192 billion from the republican budget and international financial organizations, in particular, the IDB (Bondal 2021).

Nevertheless, the potential for public investment in the digitalization of the agricultural sector is constrained by budgetary limitations. In light of these considerations, it becomes pertinent to explore potential avenues for financing the digital transformation of agribusiness from the private sector. However, attracting such investors is challenging, particularly given the significant obstacles, including insufficient guarantees of payback and return of funds, the lack of understanding of investment prospects, transparency of reporting, and others.

In world practice, this issue is resolved by attracting ESG investments (Tkachenko et al. 2023), which have been actively scaled up in recent years. However, unfortunately, Kazakhstan’s agribusiness cannot take advantage of this opportunity, since Kazakhstan has not ratified the standards for the formation of ESG reporting and has not developed a mechanism for verifying the indicators of the environmental and social impact of agricultural companies and their good corporate governance.

Moreover, no single instrument is a panacea (including ESG investments) and therefore is not capable of solving all existing problems. The desired result can be achieved by using various instruments, each of which has certain properties and its own advantages. Thus, along with ESG investments, it is advisable to use agricultural notes—non-emission securities certifying the unconditional obligation of the debtor, secured by collateral, to deliver agricultural products or pay money to the creditor in accordance with the terms determined by such a security.

An agrarian note may be represented by an account within the securities depository accounting system, which displays information regarding the agrarian note and other pertinent details within the Register of Agrarian Notes. Consequently, the Central Depository and depository institutions within the RK will be responsible for the accounting and circulation of these agrarian notes. In other words, farmers will be able to independently issue an electronic agrarian note from a personal e-cabinet in a specially created register of agrarian notes, without incurring the cost of notary services (Figure 4).

The Register of Agrarian Notes (hereinafter referred to as the Register) is an information and communication system that must display information on the issuance, content and change of details, termination and encumbrance of agrarian notes, as well as the beginning of the compulsory execution of obligations under agrarian notes based on a special extract from the Register. In this case, the holder and administrator of the Register must be the Central Securities Depository of the RK.

The agrarian notes may be classified as either commodity or financial. A commodity note is repaid with a fixed volume of a specific product, whereas a financial note is repaid with cash. The decision regarding the form of agrarian note to be utilized is at the discretion of the creditor. Upon fulfillment of the agrarian note, the agreement is deemed to be terminated. In the event of a violation of the terms or conditions of repayment of the agrarian note, the creditor is entitled to initiate enforcement proceedings through the courts. In the event that a farmer issues a commodity agrarian note, the repayment is accomplished through the transfer of the collateral to the creditor. In the event that the farmer has issued a financial agrarian note, the collateral is transferred for safekeeping and subsequently sold in order to repay the debt. Upon the complete repayment of the debt, the agrarian note is deemed to be closed.

The mechanism of agrarian notes is quite easy to use, which is especially important for small and medium-sized farmers. In addition, it allows for faster receipt of financial and material resources from creditors without transaction costs. Consequently, the demand for this instrument will grow from farmers. Moreover, the higher it is, the more offers will come from creditors since this is how the market works.

However, for the agrarian note to become a viable instrument for securing funding for the digitalization of the agricultural sector, it must attain the status of a full-fledged market product. This is one of the key objectives of state-led regulation of the digital transformation of agribusiness in Kazakhstan.

5. Discussion of the Study Results on the Search for Ways to Modify the State Regulation Mechanism of the Agribusiness Digital Transformation in Comparison with Others

The present study places significant emphasis on the critical role of state support and capital investment in the digitalization of agribusiness. This makes it different from previous studies of other researches. In particular, the studies by Kuandykova et al. (2023) provide an in-depth analysis of the evolving legal framework for digitalization in the industry. Their work considers challenges at every level of the transformation process—from ensuring basic Internet access for farmers to enhancing the legal regulation of public administration in digitalization management. However, the authors give secondary importance to issues such as state support and capital investment in the digitalization of agribusiness, which, in our view, is a critical oversight. The creation of “smart irrigation” systems in Kazakhstan’s agribusiness sector requires substantial investments and a clear understanding of potential funding sources.

Ehlers et al. (2022) have developed four scenarios for the digitalization of the agri-food sector. These scenarios include the baseline digitalization of the sector in line with current trends, strong digitalization of regulatory governance, the adoption of autonomous farming technologies, and the digitalization of the food business (Ehlers et al. 2022). These scenarios reveal various gaps in achieving the objectives of European agricultural policy. However, no ways to eliminate these problems were shown. Our study, on the other hand, clearly defines the necessary investments and methods for securing them, ensuring the attainment of goals, like the creation of “smart irrigation” in Kazakhstan.

Many researches emphasize another problem—despite its fragmented nature, the current legal framework surrounding the digitalization of agriculture is progressing more slowly than technological advancements. Precedents show that laws are often reactive, placing public policy in a position where it must not only anticipate the digital transformation of agribusiness but also actively guide it towards sustainability (MacPherson et al. 2022). One of the conclusions of the present work strongly asserts that the most effective way to achieve this is by empowering farms to utilize agricultural notes.

Research conducted by Zhou et al. (2022) expands the existing research by integrating environmental regulation, digital transformation, and agricultural productivity into a single framework through a deep theoretical extension of the “Porter hypothesis” (Zhou et al. 2022). However, the authors neglected to incorporate a scenario approach in their analysis. In contradiction, the present study employs a scenario approach effectively to predict and discuss results. Therefore, unlike other studies, our research clearly outlines strategies to accelerate the digitalization of agribusiness and drive its economic development. We demonstrated the feasibility of the predicted outcomes and highlighted the current limitations in this field as well as the vectors of further research on this topic.

6. Conclusions

This study’s findings indicate that the intensification of the climate crisis observed in recent years, accompanied by rising ground-level temperatures and the loss of water resources in the RK, significantly affects the country’s economy as well as its water and food security. The complexity of water resource management in Kazakhstan is emphasized by the fact that seven out of eight river basins are transboundary, meaning that the majority of water resources originate from neighboring countries. However, the main issue is not the amount of water available but rather its consumption and losses.

According to statistical data, irrigated agriculture is the primary water consumer in Kazakhstan, using approximately 70% of the total water intake annually. Losses of irrigation water in main, inter-farm, and on-farm canal systems, due to the high degree of wear and tear on the irrigation infrastructure, are about 50%. Thus, with the rational use of water resources and a reduction in water losses, Kazakhstan can not only mitigate water shortages but also create a strategic reserve for the future. Therefore, in 2024, the government of Kazakhstan approved the Concept for the Development of the Water Resources Management System of the RK for 2024–2030, which includes several urgent measures to mitigate water scarcity, particularly through the digitalization of agribusiness.

However, to implement state regulatory instruments for the digitalization of agribusiness, it is essential to clearly understand the sector’s readiness for digital transformation. To this end, data from the Bureau of National Statistics of the Agency for Strategic Planning and Reforms of the RK were studied, and specific indicators of the country’s agribusiness digitalization potential were selected, normalized, and used to build a mathematical model of the relationship between the level of digitalization in the agricultural sector and its water consumption.

The application of the developed mathematical model during the study proved the necessity of significantly increasing funding for the digital transformation of the agricultural sector. However, the opportunities for state support in the digitalization of agricultural production are limited by budget constraints. Consequently, there is a need to seek financing sources for the digital transformation of agribusiness from the private sector. Attracting private sector investment is challenging due to significant obstacles, such as insufficient guarantees of return on investment, lack of understanding of investment prospects, transparency of reporting, and more.

In recent years, Kazakhstan has created a legal framework for the state regulation of agribusiness digitalization, which has only outlined investment directions. However, there are no innovations regarding the attraction of funds for implementing digital transformation projects in the agricultural sector. Therefore, we propose using “agrarian notes” (electronic agricultural receipts) in Kazakh agribusiness as an innovative state regulatory tool for the development of agricultural production. For agrarian notes to become a successful tool for attracting funds for the digitalization of the agricultural sector, they must gain the status of a full-fledged market product, a process in which state regulation should play a leading role. Its main vectors should be the following:

-

State support, as well as the formation of a favorable environment for attracting investment in the creation of “smart irrigation”;

-

Launch of a single platform as a source of comprehensive information on the possibility of using “agricultural notes” in agribusiness, experience of their application, legal online consultations, etc.;

-

Ratification of CSDR and ESRS on the formation of corporate reporting on the sustainable development of agribusiness;

-

Ensuring transparency of the use of “agricultural notes” in agribusiness from the initial point to its repayment;

-

Organizing the transfer of knowledge on transactions with “agricultural notes”.

Further prospects of the present study will be aimed at studying the stakeholder approach to the digitalization of irrigation in the RK on the basin principle. Also, the application of machine learning for comparing and improving the proposed mathematical model can be used.