The Philippines used to be tagged as an agriculture-based economy. However, in the recent statistics, agriculture, fishery and forestry directly account for only 10% of the economy’s Gross Domestic Product while industry and services sectors account for 31% and 59%, respectively. These figures convey that the Philippines is now more aptly considered as a service-based economy rather than what it has been categorized before. 

The importance of agriculture looms larger when it comes to employment. In the 2014 statistics, around 31% of the Philippine labor force are from the agriculture sector. While a huge portion of the population relies on agriculture, studies suggest that Philippine agriculture has continually faltered in the past decades. Sector growth decelerated and public investments on agriculture declined albeit it remains an identified key player in accelerating inclusive growth. 

This lacklustre performance in productivity growth is considered a major constraint in the sector’s development. To address this, the government has considered augmenting its budget on agricultural development for the purpose of instituting various policies and programmes aimed at strengthening the sector for it to continue performing its lead role in combating poverty and food sufficiency especially in the rural communities.

While there have been a number of strategies and programs that aim to further improve the agriculture sector, it is also imperative to investigate what other factors could be affecting the sector outside the basic production function theory or the input-output relationship. The present study sought to find out more of the macroeconomic data that affects the declining performance of the Philippine agriculture sector. Furthermore, it also explored the link between agriculture spending and governance towards agricultural output.

The main objective of the study is to determine the factors affecting the declining performance of the Philippine agricultural sector. It explores the effect of demographic, climatic and institutional factors to the agricultural output in the Philippines. In addition, the study also investigated the mediating effect of governance between spending and agricultural output.

Result of the study would provide additional input for formulating policies and strategies in the improvement of the agricultural sector. More specifically, the result of the study would be a valuable input in academic instruction that discusses on agriculture sector performance.

The study utilized secondary data which are sourced from online statistical site such as the Country Stat of Philippine Statistics Authority (PSA), World Bank and Transparency International. Period under study runs from 1995 to 2016. The percent contribution of the agriculture sector to GDP was the dependent variable. Factors hypothesized to affect the sectors performance include: population growth (annual growth increase), CO₂ damage (Adjusted savings: carbon dioxide damage (% of GNI)) and expenditure in the agriculture sector (% to total GDP). Governance expressed in the CPI or the corruption perceptions index (released by Transparency International) is hypothesized to mediate the effect of spending to agricultural output. Figure 1 shows the working model anchored from literatures outlined previously.

The study used path analysis to explore the links among variables. Developed by Sewall Wright, path analysis is a method employed to determine whether or not a multivariate set of non-experimental data fits well with a particular (a priori) causal model. It provides a graphical way to represent the assumed theory and this was found appropriate for the study. Bootstrap method was employed to confirm the mediation between spending and agricultural output with governance. The model was estimated using Amos v.21.

The study employed secondary data sourced from reputable online statistical platforms, including the Philippine Statistics Authority’s CountrySTAT, the World Bank, and Transparency International. Covering the period from 1995 to 2016, the research focused on the agriculture sector’s percent contribution to GDP as the dependent variable. Independent variables hypothesized to influence this contribution included annual population growth, CO₂ damage (measured as adjusted savings from carbon dioxide damage as a percentage of GNI), and agricultural sector expenditure (as a percentage of total GDP). Governance, represented by the Corruption Perceptions Index (CPI) from Transparency International, was analyzed as a mediating factor in the relationship between spending and agricultural output. A path analysis, developed by Sewall Wright, was used to test the theoretical model, supported by a bootstrap method to validate the mediation effect. The model estimation was conducted using AMOS version 21, with the framework anchored in existing literature and illustrated in Figure 1.

Figure 1: Working Model

Variable’s Trend 

Shown in Figure 2 are the variable’s trend used in the study. Agricultural output shows a declining trend in the period under study. Percent increase in population shows a relatively steady state with an average of 1.97% for the year’s analysis. Meanwhile, agriculture spending expressed as percentage to total GDP, displays an erratic movement with the highest budget posted in 2008 at around 7% and the lowest is in 1998 tagged at 3.23%. Moreover, the Adjusted savings: carbon dioxide damage outlined as percent of Gross National Income displays a relatively unwavering trend. Lastly, the corruption perceptions index measured from 1 to 10 where 10 is the cleanest shows an improved trend in 2011 where it increased from 2.6 to 3.4 score and that follows a relatively stable trend until 2015.

Output Model

Figure 3 shows the output diagram with the standardized coefficients. From the working model, it can be glimpsed that the exogenous variables (population, spending and carbon emission) were correlated. These correlations are modification indices generated in the estimate calculation that improved the model fit indices. Backed-up literature to explain the correlations among these variables were no longer covered by the study.

On the other hand, Table 1 shows the model fit indices of the output model. The model do not seemed to satisfy all the suggested model fit values. By agreement, a non-significant chi-square (i.e., with p-value≥0.05) suggests good model fit which means that the data and the model is not significantly different or the data fits with the model. The comparative fit index (CFI) = 0.967 satisfies the 0.95 or higher value for good fit. However, Tucker-Lewis Index (TLI) is equal to 0.833 which is lower than the recommended good model fit of 0.95. The Root Mean Square Error of Approximation (RMSEA) = 0.096 did not comply with the suggested good model fit which is less than or equal to 0.05 [1].

Table 2 shows the regression weights of the output model. Population and carbon damage are found significant variables to affect agricultural output. Population and agricultural output are positively related. The coefficient implies that for every 1 unit increase in population, agricultural output will increase by 2.46-unit, ceteris paribus. Meanwhile, carbon damage and agricultural output are inversely related interpreted as for every 1 unit increase in carbon damage, agricultural output will decrease by 1.03 unit, holding other things constant.

Table 1: Summary of the Model Fit Indices for the Output Model.

Table 2: Regression Weights of the Output Model

Figure 2: Philippine Agriculture Output, Population, Agriculture Spending and Carbon Damage, Cpi Trend, 1995-2016

Figure 3: Output Path Diagram with the Standardized Coefficients

Moreover, the mediating effect of governance (expressed in the corruption perceptions index) is also exemplified in Table 2. The paths of spending to agricultural output and spending to governance are not significantly different from 0. But the path from governance to agricultural output is found significant having negative coefficient. This suggests full mediation of governance in the spending to output nexus. The mediation path was confirmed by employing bootstrap. The direct effect of governance on agricultural output is significantly different from zero at 0.001 level (p = .002 two-tailed) while the direct path from spending to agricultural output was not significant. 

However, governance and agricultural output has inverse relationship. Specifically, for every 1 unit increase in the corruption perceptions index, agricultural output will decrease by .581 units. This relationship deviates in the common knowledge that better governance translate to positive output such as Baldacci et al. [2], Kaufmann et al. [3] and Kaufman et al. [4]. In so far from this reading there is no literature that would support the findings. 

Population and Agricultural Output

The positive relationship between agricultural output and population growth is supported by Oduwole [5] however, deviates with Oyewole who states that higher population will have a negative impact on the level of output. 

The positive relationship of population to agricultural output explains that higher or increase in population will mean more demand for food and this will signal producers to produce more thus, increasing productive activity in the economy.

On the contrary, following Malthus proposition, faster growth in population than food may lead to famine, disease, and even war. This is evident in the Philippine setting where population increase is higher than the level of agricultural output. Annual increase in agricultural output in the Philippines is negative while population increase on the average is posted at 1.75% each year for the period 2000 to 2014 [6]. 

Hence, although increase in population will increase agricultural output, there is a need to check and ascertain the sufficiency of food production against the level of demand by the population. 

CO₂ and Agricultural Output

The negative relationship between CO₂ damage expressed as percentage to GNI and agricultural output follows the findings of Yohannes [7], Edoja et al. [8] and Mulatu et al. [9]. 

With agriculture’s dependence on optimal temperature and water availability, climate change has been and will continue to be a critical factor affecting the productivity of different activities within the sector. On the study of Charlotte Benson, he cited some disaster losses in Philippine agriculture due to different types of hazards. Department of Agriculture research on rice yield also indicated reduction in coefficients due to drought. In every stages of rice development, there’s an occurrence of stresses and the volume of yield is affected by the stress.

The foregoing discussion calls for the need to strengthen adaptive capacity and mitigation strategies related to climate change in order to improve agricultural output.

The estimates on the determinants of declining Philippine agricultural output are based on available data from the World Bank database. Among the three variables hypothesized to determine the variation in the Philippine agricultural output, population and CO₂ are found significant. 

Population and agricultural output are positively related. This explains that an increase in population increases demand for food and consequently becomes a motivation for the producers to produce more.  Thus, increasing the level of output.

CO₂ damage and agricultural output are inversely related. This implies that damage caused by CO₂ negatively affects the Philippine agricultural output.

Governance has fully mediated the effect of agricultural spending to agricultural output.

1. Albright, J. "Confirmatory factor analysis using AMOS, LISREL and Mplus." Retrieved from /~statmath/stat /all/cfa/cfa15.html.

2. Baldacci, E. et al. "International monetary fund working paper." WP/04/217, 2004.

3. Kaufmann, D. et al. "Governance matters." World Bank, Development Economics Research Group, 1999.

4. Kaufman, D. et al. "Governance matters III: governance indicators for 1996, 1998, 2000, and 2002." World Bank Economic Review, vol. 18, no. –, 2004, pp. 253–287.

5. Oduwole, O.O. "Effects of population growth on the agricultural output in Nigeria." Retrieved from https://www.academia.edu/9148475/Effects_of_Population_Growth_On_Agricultural_output_in_Nigeria?auto=download.

6. Philippine Statistics Authority. "Untitled." Retrieved from www.psa.gov.ph, 2015.

7. Yohannes, H. "A review on relationship between climate change and agriculture." Journal on Earth Sci Clim Change, vol. 7, no. 2, 2016. Retrieved from http://dx.doi.org/ 10.4172/2157-7617.1000335.

8. Edoja, E. et al. "Dynamic relationship among CO2 emission, agricultural productivity and food security in Nigeria." Cogent Economics & Finance, vol. 4, no. –, 2016. http://dx.doi.org/10.1080/23322039.2016.1204809.

9. Mulatu, D.W. et al. "The impact of CO2 emissions on agricultural productivity and household welfare in Ethiopia: a computable general equilibrium analysis." Environment for Development. Discussion Paper Series, March 2016, EfD DP 16-08.