The World’s Largest Open Access Agricultural & Applied Economics Digital Library

This document is discoverable and free to researchers across the

globe due to the work of AgEcon Search.

Help ensure our sustainability.

Give to AgEcon Search

AgEcon Search
http://ageconsearch.umn.edu
[email protected]

Papers downloaded from AgEcon Search may be used for non-commercial purposes and personal study only.
No other use, including posting to another Internet site, is permitted without permission from the copyright
owner (not AgEcon Search), or as allowed under the provisions of Fair Use, U.S. Copyright Act, Title 17 U.S.C.
 Spatial Price Transmission in the Retail Rice Markets of the Philippines
1,2*
Bathan, B.M.
1
Department of Agricultural and Resource Economics,
Faculty of Economics, Kasetsart University
2
Department of Agricultural and Applied Economics, College of Economics
and Management, University of the Philippines Los Baños

Selected paper prepared for presentation at the 1st ASEAN University Symposium for
Sustainable Food System, Faculty of Economics, Kasetsart University, Bangkok, Thailand
18-19 April 2024.

Copyright 2024 by Bathan, B.M. All rights reserved. Readers may make verbatim copies of
this document for non-commercial purposes by any means, provided that this copyright
notice appears on all such copies.

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
 Spatial Price Transmission in the
Retail Rice Markets of the Philippines
Bates M. Bathan*
Department of Agricultural and Resource Economics,
Faculty of Economics, Kasetsart University
Department of Agricultural and Applied Economics, College of Economics
and Management, University of the Philippines Los Baños
*Corresponding author Email: [email protected], [email protected]

Abstract
Rice markets in the Philippines recently experienced various policy shocks, including the implementation of the
Rice Tariffication Law (RTL) in March 2019, and the imposition of price freeze during the COVID-19 pandemic
in 2020, and mandated ceiling prices on milled rice in September 2023. This study aims to examine the spatial
price transmission of regular milled rice across 17 regional retail markets. By analyzing monthly price data from
January 2012 to December 2023, the study seeks to describe structural breaks in price series, assess the speed and
magnitude of price adjustments, determine the nature and direction of price transmissions, and identify factors
influencing spatial price transmissions. Various analytical techniques, such as descriptive analysis, unit root and
structural break tests, Toda-Yamamoto causality tests, and the autoregressive distributed lag technique, are
employed to analyze the price series. The findings reveal significant positive and negative structural breaks. Both
Central Luzon, the top producer of paddy rice, and Eastern Visayas, the main consumer of regular milled rice,
serve as reference markets. The National Capital Region also functions as a central market. The speed of price
adjustments ranges from 1.18 to 3.60 months, while asymmetric price transmission is observed in two out of four
spatially integrated regional markets. Cagayan Valley exhibits almost one-to-one price transmission, whereas the
other reference markets have an elasticity of price transmission of less than 0.80. The study recommends
monitoring prices, disseminating market information, investigating unfair trade practices, improving NFA
inventory management, and developing targeted price and market policy measures.

Keywords: Autoregressive distributed, asymmetric cointegration, spatial market

integration, structural break
JEL code: C22, D40, and Q11
SDG goals: End hunger, achieve food security and improved nutrition and promote
sustainable agriculture

Introduction
Food price inflation exacerbates food insecurity by limiting economic access to food.
According to the Philippine Statistics Authority (PSA), the year-on-year changes in the
Consumer Price Index (CPI) for food in the Philippines were 3.57 and 4.59 percent between
2012-2021 and 2018-2023 (PSA 2024). In the National Capital Region (NCR), the primary
consumer hub in the country, food price inflation rates were 4.16 and 4.63 percent, while areas
outside the NCR recorded rates of 3.46 and 4.58 percent during the same periods (PSA 2024).
Additionally, the CPI for rice in the Philippines (1.72% and 0.32%), the NCR (2.06% and
0.52%), and areas outside the NCR (1.69% and 0.30%) has generally increased despite the
decline in 2019 and 2020 (PSA 2024). Rice, as a staple food for 109.04 million Filipinos,
accounts for over a third of the average calorie intake, 13.1 percent of total household
expenditure, and a third of total food consumption (Valera et al. 2022). This makes it a priority
crop for policy interventions aimed at achieving food security goals.

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
1
As an archipelago composed of three main island groups (Luzon, Visayas, and Mindanao),
food security in the Philippines is best assessed by examining smaller geographic levels such
as regions, provinces, and municipalities/cities, which have varying levels of local production,
imports (as a net rice importer), and local consumption. Table 1 shows that while all regions,
except for the NCR, produce paddy rice, the majority is concentrated in Central Luzon,
Cagayan Valley, and Western Visayas. However, the per capita consumption of regular milled
rice is higher in Eastern Visayas, Western Visayas, and the Bicol Region.

Table 1. Paddy rice production and per capita consumption of regular milled rice, by region,
Philippines
Regions Production Share Per Capita
(mt) (%) Consumption
(2010=2023) (kg)
(FIES, 2018)
REGION III (Central Luzon) 3,444,894 18.56 70.76
REGION II (Cagayan Valley) 2,519,135 13.57 76.25
REGION VI (Western Visayas) 2,157,250 11.62 102.03
REGION I (Ilocos Region) 1,802,246 9.71 75.20
REGION XII (SOCCSKSARGEN -South Cotabato, 1,271,280 6.85 81.66
Cotabato, Sultan Kudarat, Sarangani, and General
Santos)
REGION V (Bicol Region) 1,251,332 6.74 98.13
MIMAROPA - Mindoro Occidental, Mindoro 1,113,179 6.00 97.00
Oriental, Marinduqe, Romblon, and Palawan
REGION VIII (Eastern Visayas) 935,354 5.04 104.03
REGION X (Northern Mindanao) 724,469 3.90 64.47
BARMM - Bangsamoro Autonomous Region in 651,967 3.51 77.82
Muslim Mindanao
REGION IX (Zamboanga Peninsula) 650,534 3.50 73.08
REGION XIII (Caraga) 490,765 2.64 81.30
REGION XI (Davao Region) 451,318 2.43 58.57
CAR - Cordillera Administrative Region 404,067 2.18 47.12
CALABARZON - Cavite, Laguna, Batangas, Rizal, 402,431 2.17 42.03
and Quezon
REGION VII (Central Visayas) 294,170 1.58 34.39
National Capital Region (NCR) 0.00 0.00 46.72
Philippines 18,564,392 100.00 66.52
Source of basic data: PSA (2024) and Family Income and Expenditure Survey (FIES) (2018)

Additionally, Table 2 reveals that some regions, such as Eastern Visayas and BARMM, face
deficits in their own rice production and are located far from ports where imported rice is
unloaded. These regional differences require specific interventions and strategies to ensure the
achievement of food security goals.

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
2
Table 2. Classification of provinces by rice self-sufficiency status, and proximity to port where
imported rice is disembarked
Rice self- Proximity to port where imported rice is disembarked
sufficiency Near Region Far Region
Surplus Zamboanga del Zamboanga Nueva Ecija Central Luzon
Sur Peninsula
Occidental MIMAROPA Cagayan Cagayan Valley
Mindoro
Oriental Mindoro MIMAROPA Isabela Cagayan Valley
Sultant Kudarat SOCCKSARGEN Apayao CAR
Tarlac Central Luzon Kalinga CAR
Bukidnon Northern Nueva Viscaya Cagayan Valley
Mindanao
Iloilo Western Visayas Ilocos Norte Ilocos Region
Capiz Western Visayas Quirino Cagayan Valley
Antique Western Visayas Aurora Central Luzon
Zamboanga Zamboanga North Cotabato SOCCKSARGE
Sibugay Peninsula N
South Cotabato SOCCKSARGEN Agusan del Sur Caraga
Pangasinan Ilocos Region Biliran Eastern Visayas
Zamboanga del Zamboanga Palawan MIMAROPA
Norte Peninsula
Camarines Sur Bicol Region Abra CAR
Ilocos Sur Ilocos Region Leyte Eastern Visayas
Guimaras Western Visayas Maguindanao BARMM
Lanao del Norte Northern Ifugao CAR
Mindanao
Davao del Sur Davao Region
Compostella Davao Region
Valley
Marginally Agusan del Norte Caraga Camarines Norte Bicol Region
Sufficient
La Union Ilocos Region
Pampanga Central Luzon
Bataan Central Luzon
Aklan Western Visayas
Zambales Central Luzon
Negros Western Visayas
Occidental
Sorsogon Bicol Region
Bohol Central Visayas
Deficit Davao Oriental Davao Region Western Samar Eastern Visayas
Masbate Bicol Region Northern Samar Eastern Visayas
Catanduanes Bicol Region Southern Leyte Eastern Visayas
Albay Bicol Region Eastern Samar Eastern Visayas
Zamboanga City Zamboanga Misamis Northern
Peninsula Occidental Mindanao

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
3
 Rice self- Proximity to port where imported rice is disembarked
sufficiency Near Region Far Region
Surigao del Sur Caraga Lanao del Sur BARMM
Bulacan Central Luzon Romblon MIMAROPA
Davao del Norte Davao Region Marinduque MIMAROPA
Surigao del Norte Caraga Mountain CAR
Province
Quezon CALABARZON Benguet CAR
Sarangani SOCCKSARGEN Batanes Cagayan Valley
Negros Oriental Central Visayas Sulu BARMM
Dinagat Islands Caraga Tawi-Tawi BARMM
Laguna CALABARZON Basilan BARMM
Siquijor Central Visayas
Misamis Oriental Northern
Mindanao
Camiguin Northern
Mindanao
Batangas CALABARZON
Rizal CALABARZON
Cavite CALABARZON
Davao City Davao Region
Cebu Central Visayas
Source: Department of Agriculture (2018)

Rice, a highly sensitive and political crop, has been a topic of concern in Philippine politics
since the 1950s (Intal & Garcia 2005) and has been a focal point for policy interventions under
every administration. During the 2022 presidential elections in the Philippines, President
Ferdinand R. Marcos, Jr. campaigned on the promise of reducing the retail prices of milled rice
to PhP 20 per kilogram (Presidential Communications Office 2023). Since assuming office on
June 30, 2022, the lowest prices were recorded in July 2022 (PhP 39.32/kg) and January 2023
(PhP 39.50/kg), while a record high of PhP 48.48 per kilogram was reached in December 2023
(PSA 2024). Various administrations have implemented production policies such as providing
subsidies for seeds and fertilizers, farm and post-harvest machinery and equipment,
constructing irrigation facilities and farm-to-market roads, and offering support services and
training programs. They have also implemented price and market policies, such as support
prices for paddy rice, ceiling prices for milled rice, and subsidized rice from the National Food
Authority (NFA), a government-owned and controlled corporation. In terms of trade policies,
the country has transitioned from long-standing protectionism through import quotas and the
NFA’s control of imports to the liberalization of its rice market through the Rice Tariffication
Law (RTL), which came into effect on March 5, 2019.

The RTL eliminated restrictions on rice imports and exports and replaced them with tariff
systems (RA 11203, 2019). In-quota tariffs based on the 2012-level minimum access volume
of 350,000 metric tons were set at 35% and 40% for ASEAN and non-ASEAN countries, and
out-quota tariffs of 50% for most favored nations and 180% for others (Estadilla 2022 & RA
11203 2019). Another provision of the RTL involves the reduction of regulatory functions and
downsizing of the trading operations of the NFA (RA 11203 2019). Other provisions include

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
4
the imposition of a special duty on rice in cases of sudden and extreme price fluctuations and
the establishment of the Rice Competitiveness Enhancement Fund (RA 11203, 2019).

On March 8, 2020, President Rodrigo R. Duterte declared a state of public health emergency
due to the COVID-19 pandemic (Official Gazette 2020). Proclamation 922 initiates the
implementation of a price freeze on basic necessities for a maximum of 60 days in accordance
with the Price Act.

On August 31, 2023, the national government issued Executive Order (EO) 39, imposing
mandated ceiling prices on regular and well milled rice at PhP 41 and 45 per kilogram (Official
Gazette 2023). The justifications for this policy are the alleged illegal price manipulation by
traders, the Russia-Ukraine conflict, India’s ban on rice exports, and the unpredictability of oil
prices in the world market. Through EO 42, this policy was lifted on October 4, 2023 (Official
Gazette 2023).

These policy shocks may have affected the price behavior of regular milled rice differently in
regional markets of the Philippines. The study uses the Global Bai-Perron L breaks method to
identify multiple break dates corresponding to structural changes in price series, and dummy
variables in the cointegration models to test the significance of these structural breaks.

If the Law of One Price holds in spatially related markets, a long-run price relationship can be
established between reference and peripheral markets. This means that a shock occurring in
one market is transmitted to another. Figure 1 illustrates that retail price movements in selected
regional markets mirror those in the NCR. However, their price differentials with the NCR
vary the most in the case of Caraga, followed by Western Visayas and the Bicol Region. In an
efficient market, the price differential should equal the transaction cost between two spatially
integrated markets. Consequently, any shock in a reference market (e.g. NCR) should be fully
(i.e., magnitude of elasticity of price transmission) and rapidly (i.e., speed of price adjustments)
transmitted to a peripheral market (e.g. Bicol Region, Western Visayas, or Caraga) (Meyer &
von Cramon-Taubadel 2002, Briones 2019, & Rose and Paparas 2023). If the peripheral market
responds similarly to a positive or a negative shock in the reference market, the nature of price
transmission is said to be symmetric. However, if the peripheral market reacts differently to a
positive or a negative shock, there is asymmetric price transmission (APT). According to
Meyer and von Cramon-Taubadel (2002), spatial APT may occur due to asymmetric
adjustment costs, asymmetric information, market power, and asymmetric price reporting.
These factors, along with the impacts of policy and external shocks, provide crucial
information for determining policy measures aimed at improving market efficiency.

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
5
 60.00

50.00

40.00

30.00

20.00

10.00

-
November

December

October

November

December

October

November
February
July

February
July
March

March
September

September

September
January

April

May

January

April

May

January

April
August

August
June

June

June
2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023

Phil ippines NCR Bicol Region Western Visayas Caraga

Figure 1. Nominal retail price of regular milled rice in selected regions of the Philippines
Source of data: Philippine Statistics Authority (2024)

To date, only the study of Briones (2019) which used regional price data on well milled rice
from 1990 to 2016 addressed spatial price transmission for rice at the retail level. His study
found long-run integration in the regional markets. This study contributes to the limited
literature on rice price transmission in the Philippines by using retail price series of regular
milled rice after the Rice Tariffication Law, conducting unit root tests that account for
seasonality and one unknown structural break, including structural breaks in the cointegration
models, and testing for asymmetry in price transmission. This study aims to determine
structural breaks in price series, assess the speed and magnitude of price adjustments,
determine the nature and direction of price transmissions, and identify factors influencing
spatial price transmissions.

This paper is structured as follows. The second section describes the secondary data used, their
sources, and methods of data analysis. Section 3 presents the main results and discusses these
results. The final section concludes and proposes policy recommendations.

Methodology
Secondary data on retail prices of regular milled rice and Consumer Price Index (CPI) from 17
regions between January 2012 and December 2023 were obtained from the OpenSTAT website
of the Philippine Statistics Authority (PSA). Descriptive analysis, unit root and structural break
tests, Toda-Yamamoto causality tests, and the autoregressive distributed lag (ARDL)
cointegration technique performed through Eviews 12 were used for the analysis.

The PSA reported two sets of retail price series with base years 2012 and 2018. To merge these
two sets and create a longer series, the prices were adjusted to a common base year. Monthly
CPIs for the periods of 2012-2021 and 2018-2023 were rebased by setting 100 as the CPI for
January 2018. The ratio between the current and previous month’s CPI (index relative) was
computed, and then multiplied by the previous month’s rebased CPI to obtain the current
month’s new CPI (BLS 2023). To determine deflated prices, the nominal prices were divided
by the new CPI multiplied by 100.

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
6
Descriptive statistics, including mean, maximum, minimum, standard deviation, and
coefficient of variation (CV) were used to describe the price series. The normality of data was
assessed using the Jarque-Bera (JB) test.

Various unit root tests were conducted in the study. The Augmented Dickey-Fuller (ADF) test
based on the Schwarz Information Criterion (SIC) and the Phillips-Perron (PP) test based on
the Newey-West using the Bartlett kernel (NW-Bk) were used to evaluate the null hypothesis
of series containing a unit root (i.e., non-stationary). The Kwiatkowski-Phillips-Schmidt-Shin
(KPSS) LM test based on the NW-Bk was used to assess the null hypothesis of series being
stationary.

To account for seasonality or structural breaks, the study also utilized the Hylleberg-Engle-
Granger-Yoo (HEGY) test for seasonal unit roots and a breakpoint unit root test that allows for
one unknown structural break. The HEGY test based on the Akaike Information Criterion
(AIC) considers a deterministic (seasonal dummies) and a stochastic seasonal process. If at
least one unit root is present in its autoregressive representation, then the stochastic seasonal
process is non-stationary (Darne & Diebolt 2002). The breakpoint unit root test determined a
break date based on the lag selection criterion of SIC with a maximum lag of 13. A break type
of innovation outlier and breakpoint selection based on the Dickey-Fuller minimum t-statistic
were used in the analysis. The endogenously determined date corresponds to a structural
change occurring in a specific regional rice market. The HEGY and breakpoint unit root tests
were performed using both constant and constant and trend models. The results of all the unit
root tests show that the log-transformed price series are either stationary at level (I(0)) or
integrated of order one (I(1)). This is a necessary condition for estimating an ARDL model, in
addition to the dependent variable being I(1).

To test for multiple structural breaks in the individual price series and for each ARDL model,
the study employed the Global Bai-Perron L Breaks method with a trimming percentage of 15
percent and a maximum break of five. This test compares information criteria for 0 to M
globally determined breaks. The test statistic employed heteroscedasticity and autocorrelation
consistent (HAC) covariances, pre-whitening with lags equal to one, a Quadratic-Spectral
kernel, and an Andrews bandwidth, assuming common data distribution. Structural breaks,
selected based on a modified Schwarz criterion (LWC), may correspond to shocks in regional
retail markets for regular milled rice.

The Toda-Yamamoto (TY) causality test was conducted before and after the ARDL estimation
(Toda & Yamamoto 1995). Its main advantage over the Granger causality test is that it does
not require all variables in the vector autoregressive (VAR) model to have the same order of
integration. The initial optimal lag length is determined using the LR – sequential modified LR
test statistic, FPE – Final prediction error, AIC – Akaike information criterion, SIC – Schwarz
information criterion, and HQ – Hannan-Quinn information criterion. Diagnostic tests such as
Inverse Roots of the Characteristic AR Polynomial (all roots have modulus less than one and
lie inside the unit circle), Autocorrelation LM test (no serial correlation), Normality test
(residuals are multivariate normal), and White Heteroskedasticity test (homoscedastic) are also
conducted to determine the final optimal lag length (p). The maximum order of integration
(dmax) was set at 1.0 and VAR models were estimated at VAR(p+dmax).

The TY test is conducted before the ARDL estimation to identify the dependent and
independent variables that should be included in each ARDL model. The dependent variable
represents the price of a peripheral market, while the independent variable(s) represent the

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
7
price(s) of one or more reference market(s). It is then conducted after the ARDL estimation to
determine the direction of causality. For Central Luzon, Davao Region, Northern Mindanao,
and Zamboanga Peninsula with seasonal unit root at 𝜋 frequency (Appendix Table 1), seasonal
dummy variables were included in the VAR models. To avoid ARDL models that may be
affected by simultaneous causality bias, only models showing a unidirectional relationship
were considered. This means that the regional rice markets in a specific ARDL model are
assigned distinct roles as either peripheral or reference markets.

The study utilized the ARDL technique due to its several advantages (Nkoro and Uko 2016).
It is particularly effective in distinguishing between dependent and independent variables when
there is only one long-run relationship. It can also detect cointegrating vectors even in cases
with multiple cointegrating vectors. Additionally, the ARDL model can be transformed into an
Error Correction Model (ECM) through a simple linear transformation that combines short-run
and long-run adjustments. The ECM model includes optimal lags to adequately capture the
data generating process within the general-to-specific modeling frameworks. From the ECM,
the speed of price adjustments, in months, is computed as -1/ECT where ECT refers to the error
correction term, indicating the rate at which the system adjusts towards long-run equilibrium
following short-run shocks.

The presence of cointegration between dependent and independent variables in the ARDL
models was determined using the F-bounds test to establish a long-run relationship and the t-
bounds test to determine whether the equilibrating relationship is non-generate, generate, or
nonsensical, as proposed by Pesaran and Shin (1999), Pesaran, Shin, and Smith (2001), and
Shin, Yu, and Greenwood-Ninmo (2014). The linear ARDL (p,q) cointegration model, with
the peripheral market’s price (Y) and the n-reference market’s price (Xn, n=1,2,3) time series,
is presented as follows:
p q
∆Yt = β0 + ρYt−1 + σZt + φX1,t−1 + φX2,t−1 + φX 3,t−1 + ∑i=1 λi ∆Yt−i + ∑i=0 δi ∆X1,t−i +
∑qi=0 δi ∆X2,t−i + ∑qi=0 δi ∆X 3,t−i + νt

where p and q refer to the lengths of lags for Y and X; Z t is a vector of regressors containing
trend and/or endogenously determined structural breaks; and vt is a stochastic error term, that
is independently and identically distributed.

To test the non-linear relationship between the dependent and independent variables, the
asymmetric cointegrating equation is expressed as follows:
+ − + − +
∆Yt = β0 + ρYt−1 + σZt + φ+ X1,t−i + φ− X1,t−i + φ+ X2,t−i + φ− X2,t−i + φ+ X3,t−i +
− − p−1 q + + q − − q + +
φ X3,t−i + ∑i=1 λi ∆Yt−i + ∑i=0 δi ∆X1,t−i + ∑i=0 δi ∆X1,t−i + + ∑i=0 δi ∆X2,t−i +
∑qi=0 δ− − q + + q − −
i ∆X 2,t−i + ∑i=0 δi ∆X 3,t−i + ∑i=0 δi ∆X 3,t−i + νt

where + and – represent positive and negative shocks in the n-reference market’s price (Xn) as
they affect the peripheral market’s price (Y). Wald tests were also performed to assess the
presence of long-run and short-run asymmetry.

To establish model stability, the following diagnostics were conducted: Jarque-Bera normality
test, Breusch-Godfrey serial correlation LM test, Breusch-Pagan-Godfrey heteroscedasticity
test, Autoregressive Conditional Heteroskedasticity (ARCH) test, Ramsey Regression

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
8
Equation Specification Error Test (RESET), Cumulative Sum of Recursive Residuals test, and
Cumulative Square Sum of Recursive Residuals test.

Results and Discussion

Description of the Retail Price Series
After adjusting for regional inflation, Table 3 shows that 82 percent of regional markets have
retail prices for regular milled rice that are below the average. Most provinces in Caraga and
Central Visayas do not produce enough paddy rice to meet their own needs. On the other hand,
although the majority of provinces in Zamboanga Peninsula have a surplus, the region has the
second highest average price after Central Visayas. This indicates that having a surplus at the
farm level does not always result in lower retail prices.

When it comes to price volatility, the CV shows that the highest and lowest dispersions are
observed in BARMM (0.11) and CAR (0.06) (Table 3). Only five regions exhibit a normal
distribution of deflated retail price series, as determined by the Jarque-Bera test (Table 3). In
the subsequent analysis, log-transformed deflated retail price series are utilized.

Table 3. Descriptive statistics of deflated retail price (January 2018=100) of regular milled rice,
by region, January 2012-December 2023 (n=144)
Regions Production Share Per Capita
(mt) (%) Consumption (kg)
(2010=2023) (FIES, 2018)
REGION III (Central Luzon) 3,444,894 18.56 70.76
REGION II (Cagayan Valley) 2,519,135 13.57 76.25
REGION VI (Western Visayas) 2,157,250 11.62 102.03
REGION I (Ilocos Region) 1,802,246 9.71 75.20
REGION XII (SOCCSKSARGEN -South 1,271,280 6.85 81.66
Cotabato, Cotabato, Sultan Kudarat, Sarangani, and
General Santos)
REGION V (Bicol Region) 1,251,332 6.74 98.13
MIMAROPA - Mindoro Occidental, Mindoro 1,113,179 6.00 97.00
Oriental, Marinduqe, Romblon, and Palawan
REGION VIII (Eastern Visayas) 935,354 5.04 104.03
REGION X (Northern Mindanao) 724,469 3.90 64.47
BARMM - Bangsamoro Autonomous Region in 651,967 3.51 77.82
Muslim Mindanao
REGION IX (Zamboanga Peninsula) 650,534 3.50 73.08
REGION XIII (Caraga) 490,765 2.64 81.30
REGION XI (Davao Region) 451,318 2.43 58.57
CAR - Cordillera Administrative Region 404,067 2.18 47.12
CALABARZON - Cavite, Laguna, Batangas, 402,431 2.17 42.03
Rizal, and Quezon
REGION VII (Central Visayas) 294,170 1.58 34.39
National Capital Region (NCR) 0.00 0.00 46.72
Philippines 18,564,392 100.00 66.52
Source of basic data: PSA (2024)

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
9
Results of the Unit Root and Structural Break Tests
Appendix Table 1 provides the conclusions for various unit root tests, including ADF, PP,
KPSS, HEGY, and breakpoint tests. Four regions, using the constant model, and nine regions,
using the constant and trend model, were found to be integrated of order one. Regions with
inconclusive order of integration are classified as either I(0) or I(1) but not I(2). This allows
for the estimation of ARDL models.

The results of the breakpoint unit root test, considering one endogenously determined structural
break in both constant and constant and trend models, identified January 2019, August 2023,
and September 2023 as the break dates with positive coefficients. Normally, paddy and rice
prices in January are lower due to the harvest months running from October to December in
the Ilocos Region and Central Luzon, and from January to March in Cagayan Valley, Central
Luzon, and Western Visayas. However, a positive structural break occurred most frequently in
January 2019, which can be attributed to the deficit in domestic rice stocks. The NFA appeared
to have not purchased enough from local paddy rice farmers and delayed rice imports in 2018
(Tolentino & dela Pena 2020). Another possible explanation is the suspicion of hoarding and
price manipulation by rice traders. In anticipation of the passage of the Rice Tariffication Law,
wholesalers and/or retailers may have postponed the sales of their rice stocks in January 2019,
leading to a decrease in supply and an increase in price in the retail market.

The positive structural breaks in August and September 2023 confirm the “current surge in
retail prices of rice (EO 39),” which led to the imposition of mandated ceiling prices on regular
and well-milled rice. Based on month-on-month changes in real retail prices of regular milled
rice using PSA (2024) data, the growth rates between August-September and September-
October substantially declined. For regions whose structural break is either August or
September 2023, the growth rates changed from 1.52 to -4.87 percent in the Ilocos Region,
11.75 to -8.62 percent in Central Luzon, and 10.39 to -3.59 percent in the Davao Region. As
the policy ended on October 4, 2023, prices between October and November for the three
regions rise again by 2.07, 5.08, and 4.17 percent.

Based on the multiple breakpoints determined using the Global Bai-Perron L Breaks method,
positive structural breaks are observed in October and December 2013, January-March and
September 2014, July 2015, March and July-September 2018, and January 2019, and negative
structural breaks in October-December 2015, March and April 2016, October-December 2020,
and January 2021 (Table 4). The break dates November and December 2013 and the first
quarter of 2014 correspond to the negative effects of Typhoon Haiyan on paddy rice
production, which occurred in November 2013 (GMA News 2013). As discussed earlier, the
structural breaks ion 2018 can be attributed to the inability of NFA inability to stabilize the
domestic rice supply, while the structural breal in January 2019 can be explained by suspected
uncompetitive practices among rice traders.

Table 4. Results of the Global Bai-Perron L breaks method in determining multiple structural breaks
Regions Break Dates
Model: Intercept Model: Intercept and Trend
NCR 4: 2013M12 [+], 2017M02 , 2019M01 2: 2014M03ns, 2018M09ns
ns

[+], 2022M04ns
CAR 3: 2014M09 [+], 2019M01ns, 3: 2013M10ns, 2019M01ns,
ns
2020M11 2022M04ns
Ilocos Region 4: 2014M03 [+], 2015M12 [-], 3: 2014M04ns, 2018M09ns,
ns
2018M09 [+], 2020M11 2022M04ns

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
10
 Regions Break Dates
Model: Intercept Model: Intercept and Trend
Cagayan 4: 2013M10 [+], 2015M07 [+], 3: 2015M04ns, 2018M10ns,
Valley 2018M10ns, 2020M11ns 2022M04ns
Central Luzon None with seasonal dummies None with seasonal dummies
CALABARZ 3: 2013M10 [+], 2018M09 [+], 4: 2014M03 [+], 2017M04ns,
ON 2020M11 [-] 2019M01ns, 2022M04ns
MIMAROPA 4: 2014M02 [+], 2015M11 [-], 3: 2014M03ns, 2017M10ns,
2018M08 [+], 2021M01 [-] 2019M07ns
Bicol Region 4: 2013M10 [+], 2015M10 [-], 4: 2013M10 [+], 2015M10ns,
2018M07 [+], 2020M11ns 2019M01ns, 2022M04ns
Western 3: 2014M05 [+], 2018M08 [+], 3: 2014M05ns, 2019M01ns,
Visayas 2020M12 [-] 2022M02ns
Central 3: 2013M10 [+], 2019M01 [+], 2: 2019M01ns, 2022M04ns
Visayas 2021M03ns
Eastern 4: 2013M10 [+], 2016M03 [-], 3: 2013M10 [+], 2019M01 [+],
Visayas 2019M01 [+], 2020M12ns 2022M04ns
Zamboanga 4: 2014M01 [+], 2016M04 [-], None with seasonal dummies
Peninsula 2018M09 [+], 2020M10 [-]
Northern 2: 2019M01 [+], 2020M10 [-] 4: 2013M10ns, 2017M04ns,
Mindanao 2019M01 [+], 2022M04ns
Davao Region None with seasonal dummies None with seasonal dummies
SOCCKSARG 2: 2018M03 [+], 2020M11ns 2: 2018M05ns, 2020M12 [-] with
EN seasonal dummies
CARAGA 1: 2021M01ns 2: 2018M09ns, 2022M04ns
BARMM 2: 2019M01 [+], 2020M10ns 2: 2019M01ns, 2022M04ns
ns = not significant at 5% level of probability based on OLS estimation that includes the break dates

Results of the causality and cointegration tests

The results of the Toda-Yamamoto causality test confirm that the final ARDL models include
the Bicol Region, Western Visayas, Caraga, and CALABARZON as peripheral markets, and
NCR, CAR, Cagayan Valley, Eastern Visayas, SOCCKSARGEN, Central Luzon, and Ilocos
Region as the reference markets (Table 5). NCR remains as a central market with at least one
regional market from Luzon, Visayas, and Mindanao that is spatially integrated with it. There
is no simultaneous causality bias, as indicated by the unidirectional relationship between
peripheral and reference market prices.

Table 5. Results of the Toda-Yamamoto causality test

Dependent Independent Pre- Post-ARDL Estimation Direction of Causality
Variable Variables ARDL
Estimati
on
Optimal Optima Break Dates
Lag l Lag
Length Length
Model:
Intercept
Bicol NCR, CAR, and 6 6 2: 2015M11, Bicol Region <- NCR,
Region Cagayan Valley 2018M02 CAR, and Cagayan Valley

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
11
 Dependent Independent Pre- Post-ARDL Estimation Direction of Causality
Variable Variables ARDL
Estimati
on
Optimal Optima Break Dates
Lag l Lag
Length Length
Western NCR, Eastern 2a 3 3: 2014M08, Western Visayas <- NCR,
Visayas Visayas, and 2018M05, Easterm Visayas, and
SOCCKSARGEN 2020M04 SOCCKSARGEN
Caraga NCR and Central 14 13 3: 2014M08, CARAGA <- NCR and
Luzon 2018M05, Central Luzon
2020M04
Model:
Intercept
and Trend
CALABA Ilocos Region and 2 6 1: 2018M01 CALABARZON <- Ilocos
RZON Cagayan Valley Region and Cagayan
Valley
a
Simultaneous causality bias is present in the VAR model. However, the inclusion of break dates results in
unidirectional relationship between Western Visayas and the three regional markets.

Appendix Table 2 provides a summary of the diagnostic and stability tests conducted on the
ARDL models. The Bicol Region and Western Visayas models followed a linear ARDL model,
while the Caraga and CALABARZON models utilized a non-linear ARDL (NARDL) model.

In terms of price adjustments, Bicol Region exhibited the fastest speed at 1.18 months, whereas
CALABARZON had the slowest at 3.60 months (Appendix Table 2). On the other hand, it
took Caraga 2.51 months and Western Visayas 2.90 months for retail prices to return to their
long-run equilibrium path after a shock (Appendix Table 2). Notably, both Bicol Region and
CALABARZON are classified as deficit areas, located near ports where imported rice is
unloaded. Despite this similarity, their speed of adjustments differs. Given speculation
regarding unfair rice trade practices in CALABARZON, further investigation and enhanced
price monitoring are necessary. In the case of Western Visayas and Caraga, which are situated
in Visayas and Mindanao, respectively, asymmetric price information is more persistent in
these areas than in Luzon.

Table 6 presents the results of the ARDL/NARDL estimation. Bicol Region and
CALABARZON show a high level of spatial market integration with Cagayan Valley, as
indicated by their close to one long-run elasticity of price transmission. This suggests that
changes in the retail price in Cagayan Valley have a proportional impact on the retail prices in
Bicol Region and CALABARZON, reflecting efficient price transmission between the
reference and peripheral markets.

In the long run, the spatial market integration between Western Visayas and SOCCKSARGEN
is stronger (0.58) to that with NCR (0.40), suggesting a more efficient trade flow between
Visayas and Mindanao than between Luzon and Visayas. On the other hand, a one percent
change in the retail price in Eastern Visayas leads to a 0.36 percent change in the retail price
in Western Visayas in the opposite direction. Given the higher per capita consumption but
lower share in paddy production in Eastern Visayas (104.03 kg and 5.04%) compared to
Western Visayas (102.03 kg and 11.62%), traders exploit the spatial arbitrage opportunity by

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
12
purchasing rice at a lower price in Western Visayas and selling it in Eastern Visayas at a higher
price. Consequently, this leads to an increased demand and higher prices in Western Visayas,
while Eastern Visayas experiences a decrease in price due to increased supply.

There is negative long-run asymmetric price transmission from Central Luzon to Caraga. A
one percent increase in the retail price in Central Luzon results in a 0.66 percent increase in the
retail price in Caraga. However, Caraga reacts more strongly to a decrease in price where a one
percent decrease in the retail price in Central Luzon results in a 0.79 percent decrease in the
retail price in Caraga. In the short run, however, the asymmetry is positive, with a positive
effect of contemporaneous price and a negative effect of the lagged one-month price.
There is also positive short-run asymmetric price transmission from the Ilocos Region to
CALABARZON. The effect of contemporaneous price is positive, while the effects of lagged
prices up to two periods are negative suggesting a dampening effect of past prices on the current
retail price in the short run.

Regarding the notable dates when significant structural breaks occurred in Western Visayas,
the coefficient was positive in May 2018 and negative in April 2020. These coefficients can be
attributed to the deficit in domestic rice stocks caused by the policy failures of the NFA in
2018, and the mandatory price freeze of basic necessities, including rice, which lasted for a
maximum of 60 days following the proclamation of the State of Public Health Emergency on
March 8, 2020 (Official Gazette 2020). On the other hand, additional research is needed to
clarify the reasons behind the positive structural break that occurred in Western Visayas in
August 2014 and in Caraga in April 2017, as well as the negative structural break in Caraga in
January 2021.

Table 6. Results of the ARDL/NARDL estimation

Dependent Variable / Lag Independent Variables Coeff. Standard t- Probabili
Structure Error Statistic ty
LDBICOL_REGION Constant 0.36 0.32 1.13 0.26
(9,8,5,12)
LDBICOL_REGION(- 0.61 0.11 5.44 0.00
1)
LDNCR 0.27 0.11 2.57 0.01
LDNCR(-1) -0.13 0.15 -0.89 0.38
LDNCR(-3) -0.34 0.14 -2.40 0.02
LDNCR(-5) 0.24 0.12 2.01 0.05
LDNCR(-7) -0.30 0.13 -2.40 0.02
LDNCR(-8) 0.22 0.08 2.63 0.01
LDCAR 0.13 0.13 1.03 0.30
LDCAR(-1) -0.19 0.22 -0.88 0.38
LDCAR(-3) 0.34 0.19 1.79 0.08
LDCAGAYAN_VALL 0.61 0.12 5.20 0.00
EY
LDCAGAYAN_VALL -0.32 0.14 -2.19 0.03
EY(-1)
LDCAGAYAN_VALL 0.12 0.05 2.58 0.01
EY(-12)
November 2015 -0.02 0.01 -1.83 0.07
February 2018 0.00 0.00 1.43 0.16

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
13
Dependent Variable / Lag Independent Variables Coeff. Standard t- Probabili
Structure Error Statistic ty
R2 0.98
2
Adjusted R 0.97
Long Run Effect
LDNCR 0.11 0.08 1.29 0.20
LDCAR -0.31 0.18 -1.78 0.08
LDCAGAYAN_VALL 1.11 0.16 6.92 0.00
EY
LDWESTERN_VISAYA Constant 0.48 0.11 4.24 0.00
S (1,0,1,1)
LDWESTERN_VISAY 0.66 0.05 12.27 0.00
AS(-1)
LEASTERJ_VISAYAS -0.12 0.03 -4.41 0.00
LDNCR 0.45 0.07 6.83 0.00
LDNCR(-1) -0.31 0.07 -4.33 0.00
LDSOCCKSARGEN 0.43 0.05 9.08 0.00
LDSOCCKSARGEN(- -0.23 0.05 -4.34 0.00
1)
August 2014 0.02 0.00 4.22 0.00
May 2018 0.02 0.00 3.67 0.00
April 2020 -0.03 0.01 -5.68 0.00
R2 0.97
Adjusted R2 0.97
Long Run Effect
LEASTERN_VISAYA -0.36 0.06 -6.27 0.00
S
LDNCR 0.40 0.08 5.27 0.00
LDSOCCKSARGEN 0.58 0.07 8.48 0.00
LDCARAGA (3,2,1,1) Constant 1.54 0.23 6.60 0.00
LDCARAGA(-1) 0.64 0.08 8.14 0.00
LDCARAGA(-2) 0.18 0.08 2.19 0.03
LDCARAGA(-3) -0.22 0.05 -4.60 0.00
LDNCR 0.31 0.06 5.03 0.00
LDNCR(-1) -0.11 0.09 -1.17 0.25
LDNCR(-2) -0.21 0.07 -2.86 0.01
LDCEN_LUZON_POS 0.72 0.07 10.45 0.00
LDCEN_LUZON_POS( -0.46 0.09 -4.85 0.00
-1)
LDCEN_LUZON_NEG 0.17 0.11 1.53 0.13
LDCEN_LUZON_NEG 0.14 0.11 1.32 0.19
(-1)
May 2014 -0.01 0.01 -0.69 0.49
April 2017 0.01 0.00 2.76 0.01
January 2019 0.00 0.01 -0.27 0.79

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
14
Dependent Variable / Lag Independent Variables Coeff. Standard t- Probabili
Structure Error Statistic ty
January 2021 -0.03 0.01 -4.42 0.00
R2 0.98
Adjusted R2 0.98
Long Run Effect
LDNCR -0.03 0.10 -0.28 0.78
LDCENTRAL_LUZON 0.66 0.11 5.77 0.00
_POS
LDCENTRAL_LUZON 0.79 0.10 7.88 0.00
_NEG
Wald Test F- df Probabi Remark
statisti lity
c
Long Run 5.38 (1, 128) 0.02 Asymme
tric
Short Run 111.51 (1, 128) 0.00 Asymme
tric
LDCALABARZON Constant 0.03 0.20 0.16 0.87
(1,5,9,2)
Trend 0.00 0.00 2.10 0.04
LDCALABARZON(-1) 0.72 0.06 11.93 0.00
LDILOCOS_REGION_ 0.28 0.07 4.15 0.00
POS
LDILOCOS_REGION_ -0.27 0.09 -3.10 0.00
POS(-1)
LDILOCOS_REGION_ -0.22 0.09 -2.55 0.01
POS(-2)
LDILOCOS_REGION_ -0.06 0.12 -0.46 0.64
NEG
LDILOCOS_REGION_ 0.19 0.15 1.26 0.21
NEG(-1)
LDILOCOS_REGION_ -0.34 0.15 -2.28 0.02
NEG(-3)
LDCAGAYAN_VALL 0.52 0.06 8.84 0.00
EY
LDCAGAYAN_VALL -0.34 0.07 -4.76 0.00
EY(-1)
January 2018 0.00 0.00 0.82 0.41
R2 0.97
Adjusted R2 0.98
Long Run Effect
LDCAGAYAN_VALL 0.95 0.20 4.79 0.00
EY
LDILOCOS_REGION_ -0.04 0.19 -0.20 0.85
POS
LDILOCOS_REGION_ 0.42 0.32 1.28 0.20
NEG

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
15
 Dependent Variable / Lag Independent Variables Coeff. Standard t- Probabili
Structure Error Statistic ty
Wald Test F- df Probabi Remark
statisti lity
c
Long Run 3.50 (1, 126) 0.06 Symmetr
ic
Short Run 14.16 (1, 126) 0.00 Asymme
tric
For brevity, short run estimates that are not significant at 5 percent level of probability were removed from the
table

Conclusion
The study identifies only four out of 17 regional markets that are spatially integrated. This is
done through a rather strict approach that includes conducting unit root tests accounting for
seasonality and structural breaks, establishing market pairs and groups based on the Toda-
Yamamoto causality test, and ensuring that the ARDL models do not suffer from simultaneous
causality bias.

Spatial price transmission from NCR, CAR, Eastern Visayas, and Ilocos Region to Bicol
Region, Western Visayas, Caraga, and CALABARZON, is not complete and rapid. In addition,
there is asymmetric price transmission from Central Luzon and Ilocos Region to Caraga and
CALABARZON. These can be explained by suspicion of rice hoarding and price manipulation,
policy failures of the NFA, and the presence of asymmetric price information.

The Department of Trade and Industry (DTI), Department of Agriculture (DA) – Agribusiness
and Marketing Assistance (AMAS), and local government unit’s price coordinating councils
should enhance their price monitoring efforts, especially in regions with slow speed of price
adjustments like CALABARZON. To address asymmetric price information, the dissemination
of price and market information should also be improved.

The Philippine Competition Commission (PCC) should continue its efforts to investigate
smuggling, hoarding, and price manipulation in the rice markets, prioritizing the regions with
low levels of spatial market integration. On the other hand, the NFA with its new mandate
under the Rice Tariffication Law, should ensure sufficient buffer stocks to stabilize prices in
case of a sudden shock in the rice market.

As regions in the Philippines exhibit varying levels of spatial market integration, targeted price
and market policies should be considered. A general policy, such as mandated price ceilings
on rice throughout the entire country, may lead to different outcomes–either effective or
ineffective, with larger or smaller impacts.

To gain provide further insights into the results of this study, it is important to correlate them
with spatial price transmission at the farm and wholesale levels, as well as with other types of
rice (i.e., well milled, special, and premium rice). Examining vertical price transmission in the
rice supply chain can also provide clarity regarding the behavior of retail markets. It is
suggested to employ mixed methods research to capture the qualitative aspects of this empirical
work.

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
16
References
Briones, R. M. (2019). Competition in the Rice Industry: An Issues Paper. PCC Issues Paper
1. Philippine Competition Commission. https://www.phcc.gov.ph/wp-
content/uploads/2019/03/Issues-Paper-Rice-030819-cover-1.pdf
Darné, O., & Diebolt, C. (2002). A Note on Seasonal Unit Root Tests. Quality &
Quantity 36: 305–310. https://doi.org/10.1023/A:1016032601197
Estadilla, R. J. C. (2022). Economic Impacts of Rice Tariffication Law on the Philippine Rice
Domestic Market. Journal of ISSAAS, 28(1): 76-92. The International Society for
Southeast Asian Agricultural Sciences. http://issaasphil.org/wp-
content/uploads/2022/06/7.-Estadilla-2022.-Rice-Tarification-Law-FINAL.pdf
Food and Agriculture Organization of the United Nations (FAO). (2021). Rapid Assessment
of the Impact of COVID-19 on Food Supply Chains in the Philippines.
https://www.fao.org/3/cb2622en/cb2622en.pdf
GMA News. (2013). Yolanda wiped out one-third of PHL’s rice crop – UN.
https://www.gmanetwork.com/news/money/content/336237/yolanda-wiped-out-one-
third-of-phl-s-rice-crop-un/story/
Intal, P. S., Jr. & Garcia, M. C. (2005). Rice and Philippine Politics. Philippine Institute for
Development Studies, Discussion Paper Series 13.
https://pidswebs.pids.gov.ph/CDN/PUBLICATIONS/ris-old-
backups/dps/pidsdps0513.pdf
Nkoro, E. & Uko, A. K. (2016). Autoregressive Distributed Lag (ARDL) Cointegration
Technique: Application and Interpretation. Journal of Statistical and Econometric
Models, 5(4): 63-91.
Meyer, J. & von Cramon-Taubadel, S. (2002), Asymmetric Price Transmission: A Survey,
No 24822, 2002 International Congress, August 28-31, 2002, Zaragoza, Spain,
European Association of Agricultural Economists.
Official Gazette. (2023). Executive Order No. 39. Imposition of Mandated Price Ceilings on
Rice. https://www.officialgazette.gov.ph/downloads/2023/08aug/20230831-EO-39-
FRM.pdf.
Official Gazette. (2023). Executive Order No. 42. Lifting Mandated Price Ceilings on Rice
under Executive Order No. 39.
https://www.officialgazette.gov.ph/downloads/2023/08aug/20230831-EO-39-
FRM.pdf.
Official Gazette. (2020). Proclamation No. 922. Declaring a State of Public Health
Emergency throughout the Philippines.
https://www.officialgazette.gov.ph/downloads/2020/02feb/20200308-PROC-922-
RRD-1.pdf
Pesaran, M.H. & Shin, Y. (1999). An Autoregressive Distributed Lag Modelling Approach to
Cointegration Analysis. In: Strom, S., Ed., Chapter 11 in Econometrics and Economic
Theory in the 20th Century the Ragnar Frisch Centennial Symposium, Cambridge
University Press, Cambridge, 371-413.
Pesaran, M. H., Shin, Y., &Smith R. J. (2001). Bounds Testing Approaches to the Analysis of
Level Relationships, Journal of Applied Econometrics, 16 (1): 289-326, May 2001.
Philippines Statistics Authority. (2024). Prices. https://openstat.psa.gov.ph/Database/Prices.
Presidential Communications Office. (2023). P20 rice still feasible – DA.
https://pco.gov.ph/news_releases/p20-rice-still-feasible-da/
Republic Act 11203. (2019). An Act Liberalizing the Importation, Exportation, and Trading
of Rice, Lifting for the Purpose the Quantitative Import Restriction on Rice, and for
Other Purposes.

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
17
 https://www.officialgazette.gov.ph/downloads/2019/02feb/20190214-RA-11203-
RRD.pdf
Rose, R. & Paparas, D. (2023). Price Transmission: The Case of the UK Dairy Market.
Commodities 2: 73-93. https://www.mdpi.com/2813-2432/2/1/4
Shin, Y., Yu, B., & Greenwood-Nimmo, M. (2014). Modeling Asymmetric Cointegration
and Dynamic Multipliers in a Nonlinear ARDL Framework. In: Horrace, W.C. and
Sickles, R.C., Eds., Festschrift in Honor of Peter Schmidt: Econometric Methods and
Applications, Springer Science & Business Media, New York, 281-314.
https://doi.org/10.1007/978-1-4899-8008-3_9.
Toda, H.Y., & Yamamoto, T., (1995). Statistical inference in vector autoregressions with
possibly integrated processes. Journal of Econometrics, 66 (1–2): 225–250.
Tolentino, V. B. J. & de la Pena, B. M. (2020). Deregulation and Tariffication at Last: The
Sage of Rice Sector Reform in the Philippines. Bangko Sentral ng Pilipinas Working
Paper Series 6. https://www.bsp.gov.ph/Media_And_Research/WPS/WPS202006.pdf
United States Department of Agriculture – Foreign Agricultural Service. (2023). Rice Export
Prices Highest in More Than a Decade as India Restricts Trade.
https://fas.usda.gov/data/rice-export-prices-highest-more-decade-india-restricts-trade.
United States Bureau of Labor Statistics (BLS). (2023). Consumer Price Index – Rebasing an
index. https://www.bls.gov/cpi/factsheets/rebasing.htm
Valera, H. G., Mayorga, J., Pede V. O., & Mishra, A. K. (2022). Estimating food demand and
the impact of market shocks on food expedintures: The case for the Philippines and
missing price data, Q Open, 2(2). https://doi.org/10.1093/qopen/qoac030

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
18
Appendix Table 1. Results of the unit root tests
Unit NCR CAR Ilocos Cagay Centr CALA MIM Bicol Wester Centra Easter Zamboan Norther Dava SOCC CAR AR
Root Regio an al BARZ ARO Regi n l n ga n o KSAR AG MM
Test n Valley Luzon ON PA on Visaya Visaya Visaya Peninsula Mindan Regio GEN A
s s s ao n
Model: Intercept
ADF I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1)
PP I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1)
I(0) I(0) I(0) I(0) I(0) I(0) I(0) I(0) I(0) I(1) I(0) I(0) I(0) I(1) I(0) I(1) I(1)
KPSS
None 0 None 0 0 0 0 None 0 0 None 0 and 𝜋 0 and 𝜋 0 0 0 0
HEGY
One Unknown Breakpoint
2013 2019 2019 2019 2013 2013M 2019 2019 2019M 2019 2013 2019M01 2019M0 2019 2019M 2019 2018
I(0) M06 M01 M01n M01ns M07n 07ns M01n M01 01ns M01ns M07ns ns 1ns M01n 01ns M01 M12
ns ns s s s ns s ns ns
2019 2019 2023 2019 2023 2019M 2019 2023 2019M 2019 2019 2019M01 2019M0 2023 2019M 2023 2019
I(1) M01 M01 M08 M01 M08 01 M01 M07 01 M01 M01 1 M09 01 M07 M01
Order Inco Inco Incon Inconc Incon Inconcl Incon Incon Inconcl I(1) Inconc Inconclus Inconclu I(1) Inconcl I(1) I(1)
of nclus nclus clusiv lusive clusiv usive clusiv clusi usive lusive ive sive usive
Integrati ive ive e e e ve
on
Model: Intercept and Trend
ADF I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1)
PP I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1) I(1)
I(1) I(1) I(1) I(0) I(1) I(1) I(1) I(0) I(1) I(1) I(0) I(0) I(0) I(0) I(1) I(1) I(0)
KPSS
0 0 0 0 0 and 0 0 0 0 0 0 0 and 𝜋 0 0 and 0 0 0
HEGY 𝜋 𝜋
One Unknown Breakpoint
2013 2013 2018 2013 2013 2013M 2018 2013 2018M 2018 2013 2018M06 2018M1 2022 2020M 2013 2018
I(0) M06 M07 M06n M03ns M07n 07ns M02n M06 12ns M12ns M07ns ns 2ns M07n 08ns M07 M12
ns ns s s s ns s ns ns
2019 2019 2019 2019 2023 2019M 2019 2019 2019M 2019 2019 2019M01 2019M0 2019 2019M 2023 2019
I(1) M01 M01 M01 M01 M09 01 M01 M01 01 M01 M01 1 M01 01 M07 M01
Order I(1) I(1) I(1) Inconc Incon I(1) I(1) Incon I(1) I(1) Inconc Inconclus Inconclu Incon I(1) I(1) Inco
of lusive clusiv clusi lusive ive sive clusiv nclus
Integrati e ve e ive
on

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
19
Appendix Table 2. Results of diagnostic and stability tests
Dependent Variable Independent Model Case [b] Break Dates [c] Lag Selection Lag
Variables [a] Criterion Structur
e
[1] Bicol Region NCR, CAR, and Linear Unrestricted constant and no 2: 2015M11ns, 2018M02ns Adjusted R2 (9,8,5,1
Cagayan Valley trend 2)
[2] Western Visayas NCR, Eastern Linear Unrestricted constant and no 3: 2014M08 [+], 2018M05 [+], SIC (1,0,1,1)
Visayas, and trend 2020M04 [-]
SOCCKSARGEN
[3] CARAGA NCR and Central Non- Unrestricted constant and no 4: 2014M05ns, 2017M04 [+], AIC (3,2,1,1)
Luzon linear trend 2019M01ns, 2021M01 [-]
[4] CALABARZON Ilocos Region and Non- Unrestricted constant and 1: 2018M01ns Adjusted R2 (1,5,9,2)
Cagayan Valley linear unrestricted linear trend

Dependent Variable F-Bounds [d] I(0) I(1) t-Bounds [e] I(0) I(1)
[1] 11.16 3.36 4.52 -6.68 -2.86 -3.78
[2] 10.44 3.36 4.52 -6.44 -2.86 -3.78
[3] 17.21 3.36 4.52 -7.36 -2.86 -3.78
[4] 7.21 4.20 5.32 -4.59 -3.41 -4.16

[a] Linear model means all the independent variables have a symmetric relationship with the dependent variable. Non-linear model means at least one of the independent variables has an asymmetric
relationship with the dependent variable.
[b] Case refers to the deterministic specifications of the model. This incudes five cases: (1) no constant and no trend, (2) restricted constant and no trend, (3) unrestricted constant and no trend, (4)
unrestricted constant and restricted linear trend, and (5) unrestricted constant and unrestricted linear trend.
[c] The break dates are significant at 5% level of probability based on the linear or non-linear ARDL estimation, except for those with ns that means not significant at 5% level of probability.
[d] The p-value of the F-Bounds test is less than 0.05. The critical values are taken from Pesaran, Shin and Smith (2001). The test statistic is greater than I(0) and I(1) critical values, implying the
presence of long-run relationship between the dependent and independent variable/s.
[e] The p-value of the t-Bounds test is less than 0.05. The critical values are taken from Pesaran, Shin and Smith (2001). The test statistic is greater than I(0) and I(1) critical values, implying that the
equilibrating relationship is non-generate or usual cointegration.

Dependent Variable ECM [f] p-value Speed [g] t-Bounds I(0) I(1)
[1] -0.85 0.00 1.18 -6.79 -2.86 -3.78
[2] -0.34 0.00 2.90 -6.53 -2.86 -3.78
[3] -0.40 0.00 2.51 -8.39 -2.86 -3.78
[4] -0.28 0.00 3.60 -5.44 -3.41 -4.16

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
20
 Dependent Variable Jarque-Bera [h] p-value LM(1) [i] p-value LM(p) [i] p-value
[1] 0.56 0.76 0.40 0.53 0.84 0.58
[2] 2.37 0.31 0.01 0.92 1.42 0.24
[3] 3.23 0.19 0.06 0.80 0.39 0.76
[4] 0.82 0.66 0.04 0.85 0.66 0.52

Dependent Variable ARCH (1) [j] p-value ARCH(p) [j] p-value

[1] 5.87 0.02 1.43 0.18
[2] 0.03 0.87 0.50 0.61
[3] 1.47 0.23 1.22 0.30
[4] 1.62 0.21 0.82 0.44

Dependent Variable Ramsey RESET F-statistic [k] p-value CUSUM [l] CUSUM of Squares [m]
[1] 1.66 0.20 Stable Stable
[2] 1.31 0.26 Stable Stable
[3] 0.64 0.42 Stable Stable
[4] 2.18 0.14 Stable Stable

[f] The error correction term (ECT) is significant at 1% significance level. This also implies the presence of cointegration between the dependent and independent variables.
[g] Speed of adjustment computed as -1/ECT denotes the length of time in months needed to return to long-run equilibrium path after a shock.
[h] The p-value of the Jarque-Bera (JB) normality test is greater than 0.05. This implies the presence of normality.
[i] The p-value of the Breusch-Godfrey (BG) Serial Correlation LM test is greater than 0.05 for order one and two or p (i.e., optimal lag length). This implies the absence of serial correlation.
[j] The p-value of the Autoregressive Conditional Heteroskedasticity (ARCH) test is greater than 0.05 for order one and two or p (i.e., optimal lag length). This implies the absence of heteroskedasticity.
The problem of heteroskedasticity in Central Visayas and Northern Mindanao was corrected using HAC covariances (Pre-whitening with lags = 1, Quadratic-Spectral kernel, Andres bandwidth).
[k] The p-value of the Ramsey RESET test is greater than 0.05. This implies that the model has no specification errors.
[l] CUSUM – Cumulative Sum of Recursive Residuals test and [m] CUSUM of Squares – Cumulative Square Sum of Recursive Residuals test show that the test statistics plotted at a 5% significance
level for the estimated coefficients of the model lay within the critical boundaries. This implies that the estimated coefficients of the model are stable.

ASEAN University Symposium for Sustainable Food System Faculty of Economics 18–19 April 2024
21