INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1640
www.rsisinternational.org
Paddy Production Analysis in Non-Granary Areas
Nur Aliah Najwa Ahmad Mahyudin, Nora Mohd Basir
*
Faculty of Computer and Mathematical Sciences, University Technology MARA (UiTM) Cawangan
Negeri Sembilan, Kampus Seremban, Negeri Sembilan
*Corresponding Author
DOI:
https://dx.doi.org/10.47772/IJRISS.2025.910000139
Received: 27 September 2025; Accepted: 04 October 2025; Published: 06 November 2025
ABSTRACT
Currently, it has been shown that Malaysia has a low production of paddy. That is one of the reasons why
Malaysia has been importing rice from other countries. To help Malaysia overcome the rice output shortage,
this study has decided to determine whether there is an effect of planted area and paddy yield towards paddy
production in a long-run scenario. This study focuses on analysing the paddy production of non-granary areas
in Johor, Melaka, Negeri Sembilan, Pahang, Pulau Pinang, and Terengganu from 2011 until 2020. The paddy
production that is being analysed is during the main season in Malaysia. In order to determine the presence of
the long-run relationship, this study employed the Pedroni cointegration test. It was found that the majority of
the statistical tests from the Pedroni cointegration test are significant at 5% level of significance. It can be
concluded that there is a long-run relationship between planted area, paddy yield, and paddy production. Thus,
it is recommended for future researchers to investigate in detail these effects to help maximise the paddy
production, especially in the non-granary areas.
Keywords— cointegration test, paddy production, panel data, stationary tests
INTRODUCTION
The three most important food crops in the world are rice, wheat, and maize. When taken as a whole, rice makes
up almost half of the calories consumed by people worldwide [9]. Besides, rice has also been one of the staple
foods in many countries in the world. Countries located in the Association of Southeast Asian Nations (ASEAN),
also known as the Southeast Asian region, have been categorised as the countries that consume the most rice. In
2023, a study found that Malaysia only consumes 106 kilograms of rice per person every year, which is lower
than in some of the other Southeast Asian countries [14]. Table 1 shows the total consumption of rice in Malaysia
from 2019 to 2024 [10].
Table I Total Consumption of Rice in Malaysia
Year
Rice Consumption(million metric tons)
2019
2.75
2020
2.85
2021
2.90
2022
2.90
2023**
2.91
2024**
2.91
** indicates the estimation of the rice consumption in the future
In 2021, Malaysia produced roughly 1.68 million metric tons of rice, which was 180 thousand metric tons more
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1641
www.rsisinternational.org
than the production in 2020 [11]. It can be seen that the production of rice does not align with the rice
consumption, as the current production is still unable to meet the demands of the Malaysian market. Naturally,
as the demand for rice increases, production should also rise to meet the demand. While Malaysia produced 67%
of its own rice, the remaining amount is imported from Pakistan, Vietnam, and Thailand [12]. Thus, the
Malaysian government is preparing to increase the local rice production by 2025 [11].
Essentially, Malaysia needs to produce more rice. To maximise production, Malaysia should focus on land and
climate change. Malaysia has two seasons, which are the main season and the off-season, where the activities of
paddy cultivation, from the preparation of land to harvesting, start [6]. Based on Table 2, it can be seen that the
non-granary areas in Johor, Melaka, Perak, and Terengganu have decreasing productivity status. Meanwhile,
Kedah, Negeri Sembilan, and Perlis have optimum productivity, while productivity in Kelantan and Pahang has
an increasing pattern.
Table 2 Summary of Productivity Status
Non-Granary States
Productivity Status
Johor
Decreasing
Kedah
Optimum
Kelantan
Increasing
Melaka
Decreasing
Negeri Sembilan
Optimum
Pahang
Increasing
Perak
Decreasing
Perlis
Optimum
Terengganu
Decreasing
Researchers believe that many aspects should be considered when discussing the causes of paddy production.
A study found that rice production can be affected by human capital, labour, wages, wetland, urban population,
and rice prices [3]. However, technology has barely any effect on production. Next, it is also important to
explore both short- and long-run factors that affect the paddy production. A study on rice production in Nepal
revealed that in both short- and long-term, rice production is positively influenced by cultivated area, fertiliser
consumption, and agricultural credit [5].
The average yield of cleaned paddy is the mean crop of paddy after the process of drying, cleaning, and when
the moisture content is at exactly 14% [6]. Win et al. (2020) found that during the dry season, there is no
significant difference in the yield regardless of the types of water management. On the contrary, the yield differs
with the alternate wetting and drying system having a higher paddy yield than in continuous flooding [13]. Next,
it was stated that the paddy plantation is divided into two parts [6]. The granary areas refer to the dominant
irrigation schemes, which are about 4,000 hectares minimum, while the outside granary areas are referred to as
the areas with minor or outside the irrigation system.
A study by Win et al. (2020) stated that he used both water management system, continuous flooding (CF) and
alternate wetting and drying (AWD), to manage water in maintaining the rice production. He also found that
during the wet season, CF needed more irrigation than AWD as CF irrigates 1,835 mm water while AWD
irrigates 1,522 mm of water [13].
Nevertheless, the current paddy production is not enough to cover the population in Malaysia. Since the granary
areas are likely to have optimum and increasing productivity of rice, this study attempts to focus on the non-
granary areas during the main season to help increase the productivity. Hence, the study decides to determine
whether there is a long-term relationship between planted area, paddy yield and paddy production.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1642
www.rsisinternational.org
METHODOLOGY
Source of Data
This study used data from the Paddy Production Survey Report 2020/2021 by the Department of Agriculture,
Peninsular Malaysia. The data provided is categorised as panel data, which is a combination of cross-sectional
data and time series data [15]. The annual data utilises the information from six states in Malaysia specifically
to non-granary states from 2011 until 2020, with the total observation of 60, with 2 independent variables. The
six non-granaries states are Johor, Melaka, Negeri Sembilan, Pahang, Pulau Pinang, and Terengganu, while the
independent variables that are being investigated are the planted area and paddy yield. Table 3 shows the variable
of interest in this study.
Table 3 List of Variables
Variable
Description
Measurement
Area
Planted area in hectare
Ratio
Yield
Average yield of cleaned paddy in kilogram/hectare
Ratio
Production
Wetland paddy production in metric ton
Ratio
Stationary Test
Since panel data is a combination of time-series and cross-sectional data, the study needs to test for model
stationary using unit root test. The study decides to use the Hadri Lagrange Multiplier test proposed by Hadri in
2000. Hadri has a null hypothesis of stationary, while in contrast, the alternative hypothesis of at least one series
has a unit root [7]. The components of the Hadri test are represented by Equation 1 [1].
(1)
Where
is the deterministic component while
is a random walk,
(2)
And
is known as a stationary process while
. The partial sum of the residuals,
is written
as in Equation 3
(3)
While the LM statistic is represented in Equation 4
(4)
Hadri (2000) believes that the standardized statistics are as in Equation 5 and 6 [7]:
(5)
and
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1643
www.rsisinternational.org
(6)
As T→∞ followed by N→∞.
Cointegration Test
To verify the stationary variables, unit root test is run [2]. It is a crucial step in determining whether the stationary
first-difference variables, I(1), are cointegrated or not. The cointegration test also able to check whether a long-
run relationship among the variables exist or not [4]. The study chooses to test the cointegration using Pedroni
(1999) residual-based cointegration test. It is constructed based on Engle-Granger (1987) as the test statistic is
from the residuals of the panel static regression [2].
The test statistics and p-values that will be generated by the Pedroni (1999) are panel and weighted panel of v-
statistic, ρ-statistics, PP-statistic, and ADF-statistic. Besides, there are also group mean approach of ρ-statistics,
PP-statistic, and ADF-statistic [8]. The panel and weighted panel statistics are from pooled data along the within-
dimension while the group statistics are from pooled data along the between-dimension [2]. According to
Barbieri, L., Pedroni had designed the following test statistics [2]:
1.
Panel -statistic:
(7)
2.
Panel -statistic:
(8)
3.
Panel -statistic (non-parametric):
(9)
4.
Panel -statistic (parametric):
(10)
5.
Group -statistic:
(11)
6.
Group -statistic (non-parametric):
(12)
7.
Group -statistic (parametric):
(13)
As mentioned before, the null hypothesis of the Pedroni residual cointegration test is that there is no
cointegration relationship between the non-stationary variables. So, in order to estimate a model in a long-run
relationship, the study must reject the null hypothesis. To reject the null hypothesis, the study must ensure that
the majority of the test statistics and their p-values are significant. Only then can it be proven that there is a long-
run relationship between planted area, paddy yield and paddy production.
RESULTS AND DISCUSSION
Stationary Test
Since panel data is a combination of cross-sectional data and time series data, the stationary assumption must
be checked. This study uses the Hadri Lagrange Multiplier (LM) Stationary test as stated in Table 4.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1644
www.rsisinternational.org
Table 4 Hadri Lm Test
Variables
Level
First Difference
AREA
1.443e-03 *
0.5037
YIELD
1.203e-06 *
0.7009
PRODUCTION
1.035e-04 *
0.6687
* significant at 0.05.
Based on Table 4, at the level, all p-values are significant at 5%. So, the study can conclude that area, yield, and
production are not stationary at a level. Hence, the variables need to be tested again at the first difference level.
Next, it is proven that area, yield, and production are stationary at first difference as the p-values are
insignificant. It can be concluded that all variables are stationary at first difference and integrated at first order,
I(1). Thus, the study can proceed to check the cointegration between the variables to determine the existence of
a long-run relationship between them.
Cointegration Test
Since the study aims to determine the presence of a long-run relationship between planted areas, paddy yield,
and paddy production, a cointegration test is performed. If cointegration exists, it can be proven that a long-run
relationship exists.
Table 5 Pedroni Residual Cointegration Test
Test
Statistic
Probability.
Alternative hypothesis: common AR coefficients.
(within-dimension)
Panel v-statistic
-1.2141
0.8877
Panel rho-statistic
1.1603
0.8770
Panel PP-statistic
-1.7310
0.0417 *
Panel ADF-statistic
-1.7291
0.0419 *
Weighted Panel v-statistic
-1.5085
0.9343
Weighted Panel rho-statistic
0.7018
0.7586
Weighted Panel PP-statistic
-6.9495
0.0000 *
Weighted Panel ADF-statistic
-5.4273
0.0000 *
Alternative hypothesis: common AR coefficients.
(between-dimension)
Group rho-statistic
1.6467
0.9502
Group PP-statistic
-7.6406
0.0000 *
Group ADF-statistic
-4.8891
0.0000 *
* significant at 0.05.
Six out of eleven of the p-values from Table 5 are less than 0.05. Hence, it can be concluded that the variables
are cointegrated with each other. Thus, the study can conclude that there is a long-run relationship between
planted areas, paddy yield, and paddy production.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1645
www.rsisinternational.org
CONCLUSION AND RECOMMENDATION
Conclusion
Rice is the staple food in Malaysia. So, it is normal when the demand for rice is higher compared to other foods.
So, when the demand for rice increases, the paddy production should increase by the same amount of rice
produced. However, since the market demands in Malaysia are not met, then, the study decides to explore the
possible factors that could improve the local paddy production. Hence, the study investigates the dataset of paddy
production, planted area, and paddy yield during the main season from 2011 until 2020 for a few non-granary
states in Malaysia. Next, the Hadri LM test is used to test the stationary, and it is found that planted area, paddy
yield and paddy production are not stationary at the level but stationary at first difference. Hence, the Pedroni
residual cointegration test is run. From the cointegration test, the study found that all the variables are
cointegrated to each other. The study can conclude that there is a long-run relationship between planted area,
paddy yield, and paddy production.
Recommendation
The findings confirm the existence of a long-run relationship between planted area, paddy yield, and paddy
production. Future research should further explore the extent and mechanisms through which these factors
influence production outcomes over time. It is also recommended that subsequent studies incorporate data
covering both the main and off-seasons to provide a more comprehensive understanding of paddy production
dynamics. Furthermore, translating these empirical insights into concrete policy interventions would enhance
the study’s applied significance and ensure practical benefits for Malaysian agriculture. Policymakers and
agricultural agencies should consider developing optimised planting strategies tailored to regional soil and
climatic conditions, implementing yield-enhancement initiatives such as the adoption of high-yield paddy
varieties, improved irrigation management, and precision farming technologies. These evidence-based measures,
supported by continuous farmer training and performance monitoring, could strengthen productivity in non-
granary areas and contribute meaningfully toward Malaysia’s long-term objective of achieving rice self-
sufficiency and sustainable agricultural development.
ACKNOWLEDGMENT
I would like to sincerely thank everyone for their unwavering support and commitment, which ensured the
successful publication of this article.
REFERENCES
1. Baltagi, B. H., & Kao, C. (2000). Nonstationary panels, cointegration in panels and dynamic panels: A
survey. Advances in Econometrics
2. Barbieri, L. (2008). Panel Cointegration Tests: A Survey. Rivista Internazionale Di Scienze Sociali,
116(1), 3–36. https://www.jstor.org/stable/41625199
3. Bashir, A., & Yuliana, S. (2019). Identifying factors influencing rice production and consumption in
Indonesia. Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi Dan Pembangunan, 19(2)
https://doi.org/10.23917/jep.v19i2.5939
4. Bouchoucha, Yahyaoui, I., & Bouchoucha, N. (2019). The long-run relationship between ODA, growth
and governance: An application of FMOLS and DOLS Approaches
5. Chandio, A. A., Jiang, Y., Ahmad, F., Adhikari, S., & Ain, Q. U. (2021). Assessing the impacts of climatic
and technological factors on rice production: Empirical evidence from Nepal. Technology in Society, 66
101607. https://doi.org/10.1016/j.techsoc.2021.101607
6. Department of Agriculture Peninsular Malaysia (2021). Paddy Production Survey Report Malaysia, Main
Season 2020/2021.
7. Giulietti, M., Otero, J., & Smith, J. (2009). Testing for stationarity in heterogeneous panel data in the
presence of cross-section dependence. Journal of Statistical Computation and Simulation, 79(2), 195–203.
https://doi.org/10.1080/00949650701719136
8. Örsal, D. D. K. (2007). Comparison of panel cointegration tests. Www.econstor.eu.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1646
www.rsisinternational.org
http://hdl.handle.net/10419/25201
9. Pariona, A. (2019). What are the World’s Most Important Staple Foods? WorldAtlas.
https://www.worldatlas.com/articles/most-important-staple-foods-in-the-world.html
10. Statista. (2023a). Annual Rice Consumption in Malaysia 2019-2024.
https://worldpopulationreview.com/country-rankings/rice-consumption-by-country
11. Statista. (2023b). Production of Rice in Malaysia 2013-2021.
https://www.statista.com/statistics/794700/rice-production-volume-malaysia/
12. Makhtar, S., Abidin, I. S. Z., & Islam, R. (2022). Reviewing food security on paddy production: A
conceptual paper. International Journal of Industrial Management, 16(1), 51-58.
13. Win, E. P., Win, K. K., Bellingrath‐Kimura, S. D., & Oo, A. Z. (2020). Greenhouse gas emissions, grain
yield and water productivity: a paddy rice field case study based in Myanmar. Greenhouse Gases: Science
and Technology, 10(5), 884–897. https://doi.org/10.1002/ghg.2011
14. World Population Review. (2024). Rice Consumption by Country 2023. Worldpopulationreview.com.
https://worldpopulationreview.com/country-rankings/rice-consumption-by-country
15. Yohandoko, S., & Supriyanto, S. (2023). Panel Data Analysis on Rice (Paddy) Production in Indonesia
2018-2021. International Journal of Mathematics, Statistics, and Computing, 1(3), 44–53.
https://doi.org/10.46336/ijmsc.v1i3.27
