Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Impacts of Climate Shocks on Caloric Intake of Filipinos
Department of Science and Technology Food and Nutrition Research Institute
Nutritional Assessment and Monitoring Division
May 2015
2 | P a g e
Contents
List of Tables.............................................................................................................................................................................................. 3 List of Figures............................................................................................................................. ................................................................ 4 Abstract........................................................................................................................................................................................................ 5 Background................................................................................................................................................................................................. 6 Rationale....................................................................................................................................................................................................... 7 Methods
Data........................................................................................................................................................................................ 7
Approach.............................................................................................................................................................................. 7 Results and Discussion
Past and Future Climate Change in the Philippines.......................................................................................... 9
Main Findings of the Vulnerability Analysis
Descriptive Statistics............................................................................................................................... 10
Climate Shocks............................................................................................................................................ 12
Food Insecurity and Inadequacy of Energy Intake...................................................................... 12
The Determinants of Food Consumption......................................................................................... 14
How many and where are vulnerable?............................................................................................. 15
What are the characteristics of the most vulnerable households?...................................... 18
Simulating the Impacts of Climate shocks
Climate Shocks and Vulnerability............................................................................................. ......... 20 Policy Interventions................................................................................................................................. 20
Conclusion and Recommendations.................................................................................................................................................. 22 Annex........................................................................................................................ ..................................................................................... 24 References................................................................................................................... ................................................................................ 28
3 | P a g e
List of Tables Table 1. Number of sampled households by Region, 2008 NNS............................................................................. ...... 11 Table 2. Socio-Demographic Profile of Household Heads, 2008 NNS........................................................................ 11 Table 3. Household Participation to Government Programs of Households, 2008 NNS................................... 12 Table 4. Estimated Regression Coefficients using Increase in Maximum Temperature.................................... 26 Table 5. Estimated Regression Coefficients using Increase in Percent Changes in Rainfall............................. 27
Table 6. Estimated Regression Coefficients using Increase in Maximum Temperature (households
engaged in agriculture only) ................................................................................................ ...................................... 28 Table 7. Estimated Regression Coefficients using Increase in Percent Changes in Rainfall (households
engaged in agriculture only) ................................................................................................ ...................................... 29 Table 8. Summary of Food Security Status by Vulnerability Model............................................................................ 18 Table 9. Summary of Food Security Status by Vulnerability Model of Households engaged in
Agriculture(using subpopulation model)............................................................................................................. 18 Table 10. Simulated Impacts on Mean Vulnerability Index............................................................................................... 21
4 | P a g e
List of Figures Figure 1. Past and future climate patterns in Philippines - marked increases in maximum temperature
but normal rainfall patterns ....................................................................................................................................... 9 Figure 2. Past and future climate patterns in Philippines - marked increases in minimum temperature
but normal rainfall patterns ....................................................................................................................................... 10 Figure 3. Percent change in Maximum Temperature of 2008 relative to 1979-2006 average......................... 24 Figure 4. Percent change in Precipitation of 2008 relative to 1979-2006 average............................................... 25 Figure 5. Proportion of Food Insecure Households by Region using the Radimer-Cornell Tool..................... 13
Figure 6. Proportion of Food Insecure Households by Region based on the actual caloric intake............... 14 Figure 7. Proportion of Vulnerable to Food Insecurity Households by Province................................................... 16 Figure 8. Proportion of Vulnerable to Food Insecurity Households by Vulnerability Model............................ 17 Figure 9. Characteristics of Households Across Vulnerability Index............................................................................ 19
5 | P a g e
Impacts of Climate Shocks on Caloric Intake of Filipinos
Charmaine A. Duante, MSc., Carmina D. Cuarteros, Glen Melvin A. Gironella, Eldridge B. Ferrer, MSc., Cecilia S. Acuin, M.D., Ph.D., Mario Capanzana, Ph.D.,
Food and Nutrition Research Institute – Department of Science and Technology
Abstract
Background. The Philippines is among the top 10 countries worldwide at risk to climate change and disasters.
Forecasts revealed that there is an increase in the mean annual precipitation of the country, not only in magnitude
but also in variability. These changes may affect food security and may pose a big challenge for policy makers in
targeting the most vulnerable population for policy interventions. This paper examined the nature and extent of
vulnerability of households in the Philippines to the impacts of climate change on food security. Methodology. The
study utilized the 2008 National Nutrition Survey data on food consumption which was associated with level
changes in climate shocks from the downscaled climate scenarios provided by Philippine Atmospheric,
Geophysical and Astronomical Services Administration (PAGASA). A total of 4,854 households were included in the
study. Thirty-one percent of them are engaged in agriculture. Data on per capita caloric intake, information on
household heads, assets, and government program participation were analyzed. Vulnerability analyis was used to
allow the estimation of the probability of falling below the per capita caloric threshold of 2,000 kcal. Results.
Seven out of ten households were deemed food insecure as reported by the respondents. Based on the caloric
intake, the proportion of food insecure among Filipino households and among households engaged in agriculture
were 16% and 17% respectively. Regression analysis showed that there is no significant linear relationship
between caloric intake and climate shocks. Though, most of the predictors, specifically characteristics of the
household head performed as expected. These estimates were robust across different models. Though insignificant
at 5%, the effect of an inrease in maximum temperature on food consumption is negative. As maximum
temperature increases, food consumption is expected to decrease. A negative percent change in precipitation of at
most 20% is expected to increase food consumption, while a positive percent change up to 20% will bring a
negative effect. At the national level, 15% of the households are vulnerable to food insecurity. High proportions of
these households were noted in Abra, Ilocos Sur, Mt. Province, Quirino, Aurora, Taralac, Cavite, Marinduque,
Masbate, Camiguin, Zamboanga Del Sur, Zamboanga Sibugay, Lanao Del Sur and Saranggani. Among the
households engaged in agriculture, 16.3% are vulnerable to food insecurity. Simulation analysis showed that if the
maximum temperature is increased by 5% and 10%, the mean vulnerability is expected to increase by 13% and
19% respectively. Moreover, if the amount of rainfall is increased by 5% and 10%, the mean vulnerability is
expected to increase by about 10%. Simulating possible interventions showed that increasing the education of
household heads to at least high school or limiting household size into 4 is expected to decrease the mean
vulnerability by almost 5 percentage points. Conclusion. Understanding the overall effect of climate change in
food security is a complex process. This attempt may contribute in designing policy interventions by assessing
both sensitivity and adaptive capacity at the household level.
Keywords: vulnerability, food insecurity, inadequate caloric intake, climate shock, NNS
6 | P a g e
1. Background
Food security is a complex issue that all countries worldwide are dealing with especially with
climate change as the leading risk factor. As defined in the 1996 World Food Summit, food security
exists “when all people at all times have access to sufficient, safe, and nutritious food to maintain a
healthy and active life”. In addition, inclusion of economic access to food that meets one’s dietary
requirement defines the concept of food security (WHO, 2014).
According to the Food and Agriculture Organization (FAO) the four pillars of food security are
availability, access, utilization and stability. Availability relates to supply and agricultural
production. Moreover, this also plays an important role on food security through distribution and
exchange of products (Luther, 1999). Since it is crucial to produce quantities of food with sufficient
basic dietary requirement for daily life, this pillar is the most explored among the four. Access on
the other hand, refers to the affordability, allocation and food preference of an individual. This is
also where household income and other sources take their parts. That’s why generating
reasonable income for farmers, their communities and the overall rural population is very
important to cope with challenges on food access - since this population are those who are most
vulnerable to price fluctuations. The third pillar involves food use and metabolism. Food safety is
the primary key on this pillar and in order to achieve food security, the food utilized by a particular
household must be of good quality, safe and adequate for each of its members. Lastly, stability
refers to the capability to obtain food over time (FAO, 1997). Everyone should have access to food
at the right time, right form and right place. However, these four pillars of food security may be at
risk due to a single factor - climate change.
Climate change or erratic weather may lead to a decline in agricultural production (food
availability) which may cause increase in prices (food access) and eventually lower food
consumption or a shift in consumer’s preference (food security). Moreover, there could be a
possible decline in production, individuals depending mainly on agriculture are expected to have
higher vulnerability to food security (Orewa & Iyanbe, 2010). On the other hand, abrupt level
change in precipitation, and temperature may affect plant growth and results to a decrease in the
micronutrients available on the crops. With a shift in the demographic distribution of pests and
diseases due to climate change, water contamination and food safety (food utilization) may be
threatened (Lake et al, 2012). Lastly, the pillar which appears to be the most affected by climate
change is food stability wherein a small effect of climate shock may lead to a great variability in
prices which repeats the cycle.
With this information, an analysis of the impacts of climate change on vulnerability to food
insecurity of Filipino households was carried out. One of the main objectives of this study was to
quantify the extent to which climate change affect vulnerability to food insecurity with other
factors as either channels or confounders. To address these issues, a deeper understanding of
different household settings, their coping mechanisms and strategies that could minimize or cancel
out these climate shocks were carried out in this study.
7 | P a g e
Other objectives of this study include the identification and mapping of location of vulnerable
groups, determination of the extent or degree of vulnerability and formulation of policies that
would help communities or households adapt to climate shocks.
2. Rationale
Climate change affects food and nutrition security and further undermines current efforts to reduce hunger and protect nutrition. Furthermore, climate change exacerbates the current unacceptable high levels of food insecurity. Studies that may demonstrate the impacts of climate change particularly on food security not only in the present time but in the future as well are crucial in crafting policies and/or programs or policy directions to address the need on adapting and mitigating impacts of climate change in food security and hunger.
For example, the International Rice Research Institute (IRRI) conducted direct studies on the effects of observed climate change on crop growth and yield. Through this, more accurate information for assessing the impact of climate change on crop production and eventually, food security may be provided. According to this study, a 10% decline in grain yield for each 1°C increase in the growing-season minimum temperature was observed. Not only temperature, water availability is one of the most critical factors for sustaining crop productivity. Having said that there is an increase in mean temperature, there should also be an increase in water to support crop growth. However, extreme rain events can lead to future crop loss. In fact, NASA in 2002 stressed that global climate models project increasing trends in frequency of extreme precipitation events may continue and even strengthen over the coming decades. Furthermore, Rozenweig emphasized that “the impacts of excess soil moisture due to increased precipitation need to be taken into account because of associated crop losses and potential financial damages”. However, researchers argue that while droughts receive the most attention when it comes to assessing the impacts of climate change on agriculture, excess precipitation should also be a major concern. Rainfall variability from season to season greatly affects soil water availability to crop and thus pose crop production risks. Consequently, these impacts of climate change on crop production negatively affects food availability, conservation, access and utilization and exacerbates socioeconomic risks and vulnerabilities. Also, it is expected to further reduce food productivity and make production even more erratic in regions where agricultural productivity is already low. With local production declining and probable disruptions caused by climate hazards, income generating opportunities and purchasing power will decrease for vulnerable populations. With this, policy directions to address the need to adapt and mitigate impacts of climate change in food security and hunger were needed.
3. Methods
3.1 Data The study utilized the data of 7th National Nutrition Survey of the Philippines conducted in 2008,
a nationally representative survey by the Food and Nutrition Research Institute, Department of
Science and Technology. A total of 3,377 enumeration areas with 4,854 sample households were
included in the study. Among the variables included in this study were socio-demographic profile
8 | P a g e
of the household head (e.g. sex, age, education), household size, information, communication and
transportation infrastructures, access to electricity and government programs participated. Per
capita caloric intake based on consumption unit was used as the measure of food consumption.
The FNRI data was matched with the constructed shocks from the downscaled climate data
(rainfall, maximum and minimum temperature for each province) provided by the Philippine
Atmospheric Geophysical and Astronomical Services Administration (PAGASA) . An increase in
maximum temperature and decrease in minimum temperature as well as level changes in
precipitation, temperature in 2008 relative to its respective past 20 year average (1979-2006)
were expressed as shocks.
3.2 Approach
This study aims to assess the impacts of climate change on food consumption, measured through
per capita caloric intake (kcal). Per capita caloric intake was computed by dividing the household’s
total intake to the household’s aggregated consumption unit (cu). Caloric threshold was set to
2000 kcal. Caloric intake falling below this threshold means the household did not meet the
minimum energy requirement (under light activity).
Two climate variables were observed in the study: temperature (maximum and minimum) and
precipitation. The study considered to look at the effect of the changes in level/percent changes in
these variables. The formulas used in calculating the shocks were as follows:
∆ 𝐌𝐚𝐱𝐢𝐦𝐮𝐦 𝐓𝐞𝐦𝐩𝐞𝐫𝐚𝐭𝐮𝐫𝐞:
Maximum Temperature2008 − Mean Maximum Temperature1979−2006
∆ 𝐌𝐢𝐧𝐢𝐦𝐮𝐦 𝐓𝐞𝐦𝐩𝐞𝐫𝐚𝐭𝐮𝐫𝐞:
Minimum Temperature2008 − Mean Minimum Temperature1979−2006
% 𝐜𝐡𝐚𝐧𝐠𝐞 𝐢𝐧 𝐩𝐫𝐞𝐜𝐢𝐩𝐢𝐭𝐚𝐭𝐢𝐨𝐧:
Precipitation2008 − Mean Precipitation1979−2006Mean Precipitation1979−2006
x 100%
9 | P a g e
Linear Regression Analysis was used to determine which variables are significantly associated and
what is the direction of these associations with per capita caloric intake. Generalized Least Squares
was used to account for heteroskedasticity. Through this, estimates of the model of calorie
consumption were computed. Then, the vulnerability of each household to food security was
estimated.
Vulnerability Analysis was used to provide a quantitative measure of probability that a particular
household will be food insecure under different circumstances. Not only for its capability to
predict, this model identifies the risks that households are exposed to, while also estimating their
impacts to food security. Nevertheless, this analysis is also useful for preparedness planning since
it allows policymakers to prioritize and design interventions.
Vulnerability to food security was defined as:
𝐕𝐭,𝐚 = F(z) ∫(z − xt+1)a
z
a
f(Xt+1)
F(z)dxt+1
In simple terms, it is the normal probability that the household’s expected dietary energy
consumption 𝑥𝑡+1, measured in Kilocalories, will fall below a threshold z (Christiaensen &
Boisvert, 2000).
To compare the present state of food security and their vulnerability, profiling of the households
based on their present state of food security was done. The movement of the vulnerability index
along with the variables that has great impact on caloric intake were observed and assessed for
possible policies in the future.
4 Results and Discussion
4.1 Past and Future Climate Changes in the Philippines
The past and future climate patterns in the Philippines are shown in Figure 1. The climate data
from 1979 up to 2010 were hindcast while the data from 2011 up to 2050 were bias corrected
GCM scenario (MPEH5 A2).
10 | P a g e
As observed on the graph, there is an increasing trend in maximum temperature all throughout the
years. However, precipitation seemed to have normal fluctuations though, very slight increasing
trend can be observed. On the other hand, there is also an increase in the minimum temperature
through time (Figure 2).
0.00
1000.00
2000.00
3000.00
4000.00
5000.00
6000.00
21.00
21.20
21.40
21.60
21.80
22.00
22.20
22.40
22.60
22.80
1979 1984 1989 1994 1999 2004 2009 2014 2019 2024 2029 2034 2039 2044 2049
Figure 2: Past and future climate patterns in Philippines - marked increases in minimum temperature but normal rainfall patterns
Mean Minimum Temperature PrecipitationLinear (Mean Minimum Temperature) Linear (Precipitation)
0.00
1000.00
2000.00
3000.00
4000.00
5000.00
6000.00
28.5028.7028.9029.1029.3029.5029.7029.9030.1030.3030.5030.7030.9031.1031.3031.50
1979 1984 1989 1994 1999 2004 2009 2014 2019 2024 2029 2034 2039 2044 2049
Figure 1: Past and future climate patterns in Philippines - marked increases in maximum temperature but normal rainfall patterns
Mean Maximum Temperature PrecipitationLinear (Mean Maximum Temperature) Linear (Precipitation)
11 | P a g e
4.2 Vulnerability Analysis
Vulnerability analysis examines the probability of a particular household to fall below the caloric
threshold using variables that may serve as either measure of shock or coping mechanism. Some of
the variables used were information on household head and characteristics of household, climate
shocks and government programs. If there is a high risk of disaster, man-made or natural among a
community, it’s vulnerability tends to go up. Usually, the population groups that are highly
vulnerable to food insecurity include children, pregnant women, lactating mothers, female-headed
households, the elderly and disabled, households with limited ownership and those under threat of
drought and other disasters or hazards (WFP, n.d.). Basically, vulnerability analysis on food security
aims to determine the households who are chronically food insecure and temporarily food secure,
and on the other hand, households who are temporarily insecure and permanently food secure.
4.2.1 Descriptive Statistics
Table 1 shows the distribution of households by region. A total of 4,854 households were included in
the study. The total number of households per region ranges from 154 – 455 with an average of 256.
Among the households engaged in agriculture, more than 100 households were located in Cagayan
Valley, Visayas and SOCCSKARGEN. As expected, the least proportion was found in NCR (National
Capital Region).
Region All
households %
Households engaged in agriculture
%
Ilocos Region 291 6.00 74 4.94 Cagayan Valley 256 5.27 123 8.21 Central Luzon 385 7.93 88 5.87 Bicol 280 5.77 86 5.74 Western Visayas 346 7.13 105 7.01 Central Visayas 368 7.58 112 7.48 Eastern Visayas 268 5.52 115 7.68 Zamboanga Peninsula 224 4.61 90 6.01 Northern Mindanao 241 4.96 69 4.61 Davao Region 254 5.23 84 5.61 SOCCSKARGEN 259 5.34 123 8.21 NCR 454 9.35 5 0.33 CAR 193 3.98 89 5.94 ARMM 154 3.17 78 5.21 CARAGA 207 4.26 77 5.14 CALABARZON 455 9.37 88 5.87 MIMAROPA 219 4.51 92 6.14
Total 4854 100 1,498 100
Table 1. Number of sampled households by Region, 2008 NNS
12 | P a g e
The mean age of the household heads was 50 years, while the mean years of schooling was 8 or
approximately 2nd year high school. In addition, 17% of the households was female-headed.
Majority owns a television set and a telephone/cellphone. Eighty percent of the households had
access to electricity. The percentage of households which owned radio, motorcycle or bicycle and car
were 47%, 17% and 6% respectively. The mean years of schooling of household heads engaged in
agriculture is 2 years less (6-elementary) than that of the general population. Only 4% of these
households were female-headed and little proportion of these has communication/transportation
means.
Characteristics All Households
Households engaged in agriculture
Mean/Prevalence SD Mean/Prevalence SD
Age of Household Head 50 13.9 49 12.8
Years of Schooling of household Head 8 3.8 6 3.3
Household Size 5 0.4 5 2.5
Sex of Household head (female) 0.17 2.6 0.04 0.2
Ownership of radio 0.47 0.5 0.41 0.5
Ownership of television 0.65 0.5 0.46 0.5
Presence of electricity 0.80 0.4 0.64 0.5
Ownership of telephone/cellphone 0.54 0.5 .0.37 0.5
Ownership of car 0.06 0.2 0.02 0.1
Ownership of motor or bicycle 0.17 0.4 0.11 0.3
The proportions of household participating to specific government programs are shown in Table 3.
Less than 10% of the sampled households availed nutrition education, credit grant and fish
production. Meanwhile, more than 70% participated in food production and fruit trees/vegetables
production. Among the households engaged in agriculture, at least 76% participated in food
production, fruit trees/vegetables production and livestock production programs of the government.
Government Programs All Households
Households engaged in agriculture
% SD % SD Nutrition Education 6.5 25.7 7.5 27.6 Credit grant 4.7 31.1 7.4 32.7
Tinadahan Natin 18.1 38.5 16.5 37.1 Botika ng Baranggay 25.9 43.8 28.6 45.2 Free Medical and Dental 24.4 44.9 24.8 44.9
Food Production 77.7 41.6 92.3 26.6 Produce Fruit trees/Vegetables 70.9 45.4 85.9 34.8 Produce Livestock 54.6 49.8 76.8 42.2
Produce Fish 2.7 16.2 4.3 20.2
Table 2. Socio-Demographic Profile of Households, 2008 NNS
Table 3. Household Participation to Government Programs, 2008 NNS
13 | P a g e
4.2.2 Climate Shocks
About 8 out of 10 households nationwide or 68 out of 81 provinces experienced a decrease in maximum temperature relative to its 1979-2006 average (Annex: Figures 3 and 4). On the other hand, no household experienced a negative percent decrease in minimum temperature. More than half of the households experienced a positive percent change in rainfall, clustering up to 20%.
4.2.3 Food Insecurity and Inadequate Energy Intake
Figure 5 shows the 2008 regional estimates of food insecurity based on the Radimer-Cornell tool.
Seven in ten households were deemed food insecure, as reported by respondents to have
experienced anxiety (that food may run out before they can get money to buy more and/or food
bought did not last and they did not have enough money to get more) at least once in the past 3
months before the interview. High proportions of these households were noted in Bicol, Central
Visayas, SOCCSKSARGEN, ARMM and Nothern Mindanao (82.1, 80.7, 84.9, 84.5, 80.7 respectively).
Expectedly, prosperous regions like Central Luzon, NCR, CALABARZON and Davao recorded the least
number of food insecure households.
Figure 5. Proportion of Food Insecure Households by Region using the Radimer-Cornell Tool
14 | P a g e
Proportions of households with inadequate energy intake by region is presented in Figure 6. At the
national level, 16% or 4 in 25 households have inadequate energy intake.
Using the two measures of food security, Bicol Region, Central Visayas, ARMM and SOCCSKARGEN
can be considered as provinces with high prevalence of food insecurity.
Per capita caloric intake < 2,000 kCal
Figure 6. Proportion of Food Insecure Households by Region based on the actual caloric intake
15 | P a g e
4.2.4 The Determinants of Food Consumption
Two models of food consumption were constructed (Annex: Table 4 and 5). Each model included
the same variables but unique climate shock. Results showed that most of the predictors particularly
the characteristics of the household head performed as expected. These estimates were in fact
robust in both models. All of the socio-demographic characteristics of the household head were
highly associated with a higher level of food consumption. The age and years of schooling of the
household head has a monotonic relationship with the per capita caloric intake of the household. As
expected, as household size increases, the per capita intake decreases. Ownership of television set,
telephone/cellphone and good source of drinking water shows a highly significant association with
food consumption.
It is also shown in the models that all climate shocks remain insignificant. Though insignificant at
5%, the effect of an increase in maximum temperature on food consumption is negative. As
maximum temperature increases, food consumption is expected to decrease. Moreover, a negative
percent change in precipitation at most 20% is expected to increase food consumption. While a
positive percent change up to 20% will bring a negative effect, the sign reverses for percent change
greater than 20%.
To arrive at the potential impact, adaptive capacities should also be assessed together with the
sensitivity and hazards discussed above. Among the government programs participated in by the
households before or during the survey year included the free medical and dental check up and
livestock production. These factors showed a highly significant negative association with food
consumption. This result may have been due to the target population of these government
programs, where expected participants are from the poor sectors of the society.
Subpopulation analysis on households engaged in agriculture was carried out (Annex: Table 6 and
7). Results revealed that the effect of an increase in maximum temperature on food consumption
became negative, but still insignificant. Surprisingly, a percent decrease in rainfall between 10% and
20% showed a significant negative impact on food consumption. This means that if there is a percent
decrease in rainfall within the said range, there is an expected decrease in the food consumption of
those households engaged in agriculture. In both models, educational attainment and sex of the
household, ownership of television set and some sources of drinking water and participation in free
medical and dental check-up did not switch signs but became insignificant. These models also
showed less fit on the data but did not affect the over-all vulnerability of the population.
4.2.5 How many and where are the vulnerable?
Following the regression analysis, the vulnerability index for each household was computed using
the predicted caloric intake and its variance. The normal probability that the household’s percapita
16 | P a g e
caloric intake will fall below the threshold (2,000 kcal) was calculated. Figure 7 shows the
proportion of vulnerable households by province. Results showed that in 2008, about 16% of the
households have inadequate caloric intake. In the future period, we can expect 15% or 3 out of 20
Filipino households falling below the caloric threshold. On the provincial level, high proportions
were noted in Abra, Ilocos Sur, Mountain Province, Quirino, Aurora, Tarlac, Cavite, Marinduque,
Masbate, Capiz, Camiguin, Zamboanga del Sur, Zamboanga Sibugay, Lanao del Sur and Saranggani.
Figure 7. Proportion of Vulnerable to Food Insecurity Households by Province
17 | P a g e
Using the 2 climate shock models, temperature and rainfall, at the national level, the proportion of
vulnerable households across models were almost the same. However, across regions, the estimates
generated by the two models have discrepancies. Using the temperature model, some provinces like
Northern Mindanao, NCR and CARAGA have relatively high estimates while some provinces have
relatively low estimates.
The over-all profile of the households remained consistent in both models at the national level. On
the average, 80% of the households are permanently food secure. These households have an
average probability of being food insecure that is far below the vulnerability cut-off (0.5) which
implies that these households’ food security status is stable without any interventions. Three
percent are temporarily food secure. These are the households that are currently food secure but
likely to be food insecure in the future. Almost five percent are temporarily food insecure. Although
very close, the average vulnerability of the households belonging to this classification is still below
the cut-off point.
These households are more sensitive to exposures and sudden changes that make them shift food
security status. Nearly, twelve percent are chronically food insecure. These are the households that h
are currently food insecure and are likely to be food insecure in the future. This group is the most
important target on choosing policy instruments/interventions.
0
5
10
15
20
25
Figure 8. Proportion of Vulnerable to Food Insecurity Households by Vulnerability Model
Temperature Rainfall
18 | P a g e
Food Security Status
Future
Non-Vulnerable Vulnerable
Temperature Rainfall Temperature Rainfall
Current Food Secure 80.3 (3898) 80.4 (3903) 3.5 (170) 3.4 (165)
Food Insecure 4.7 (228) 4.6 (223) 11.5 (558) 11.6 (563)
*general population Meanwhile, using the temperature model specific for the households engaged in agriculture, there are more temporarily food insecure households and less chronically food insecure. On the other hand, using the rainfall model, there are less temporarily food secure households and more permanently food secure households. Though, the differences in proportion between the 2 climate shock models were found insignificant.
Food Security Status
Future
Non-Vulnerable Vulnerable
Temperature Rainfall Temperature Rainfall
Current Food Secure 79.4 (1190) 79.6 (1193) 3.7 (55) 3.5 (52)
Food Insecure 4.5 (67) 4.3 (64) 12.4 (186) 12.6 (189)
*subpopulation
4.2.6 What are the characteristics of the vulnerable households? To examine the movement of variables across different levels of vulnerability, the likelihood of being
food insecure in the future were categorized. The levels of vulnerability highlights the different food
security status - permanently food secure, transitorily food secure/insecure and chronically food
insecure. Each variable in the model was associated with the vulnerability across different
households.
Figure 9 shows the different household characteristics graphed with the vulnerability index
obtained using the rainfall model. The vulnerability index ranges from 0 to 1 and that as it
approaches 1, the household becomes vulnerable to food insecurity. Apparently, when the trend line
Table 8. Summary of Food Security Status by Vulnerability Model*
Table 9. Summary of Food Security Status of households engaged in
agriculture by Vulnerability Model*
19 | P a g e
tends to go down, households that are more likely to be vulnerable have less assets like television
set telephone/cellphone, radio, motorcycle/bicycle and car. However, for female headed
households, the trend line tends to go up. This means that higher proportion of female-headed
households are found to be vulnerable to food insecurity.
4.3 Simulating Impacts of Climate Change
The simulation conducted was not related to the dynamics of vulnerability rather comparative
statistics. Vulnerability under several assumptions were calculated using the estimates obtained in
the previous section. In this simulation, the process of computing the vulnerability was not repeated,
instead, the vulnerability indices obtained through the first analysis were compared to the
vulnerability indices under the different assumptions.
4.3.1 Climate Shocks and Vulnerability
Considering the individual impacts of maximum temperature and rainfall in food consumption, the
following scenarios were examined:
1. Increase maximum temperature by 5%
2. Increase maximum temperature by 10%
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
< 0.30 0.31 - 0.50 0.51 - 0.60 0.61 - 0.70 0.71 - 0.80 0.81 - 0.90 0.91 - 1.0
Female headed household Presence of Radio
Presence of TV Presence of Electricity
Presence of Telephone/Cellphone Presence of Car
Presence of Motorcyce/Bicycle
Figure 9. Characteristics of Households Across Vulnerability Index
20 | P a g e
3. Increase precipitation by 5%
4. Increase precipitation by 10%
Controlling for the effects of other factors, an additional 5% and 10% in the maximum temperature
increases the average vulnerability at the national level by 14% and 19% respectively. While,
increasing the amount of precipitation by 5% and 10% increases the average vulnerability both by
10%.
4.3.2 Policy Interventions
Considering the effects of the policy variables in food consumption, several alternative hypotheses
were formulated and tested:
1. Increase education of household head minimum to 6 years (elementary level)
2. Increase education of household head minimum to 10 years (high school level)
3. Limiting Household Size to 4 members
4. Limiting Household Size to 6 members
5. Access to credit
6. Produce fruit/trees or vegetables in garden
7. Produce Livestock
8. Produce Fish
Holding other factors constant, education has the largest impact on vulnerability. The education of
the household head contributes as asset and thus assumed to increase the adaptive capacity of
household, decreasing the potential impact of climate shocks. Simulation analysis showed that
increasing the education of household heads to at least high school or limiting household size into 4
is expected to decrease the mean vulnerability by almost 5 percentage points. This may imply that a
higher level of education than elementary education will decrease vulnerability by almost 4%
Furthermore, household’s initiative to produce fruit trees/vegetables, poultry/livestock or fish
production decreases vulnerability but very minimal.
21 | P a g e
Variable Condition
Simulated Actual
v1 temperature
v1 rainfall
v1
Climate Variable
adds 5% maximumtemperature 0.3837
0.2477
adds 10% maximum temperature 0.4416
less 5% rainfall
0.3491
less 10% rainfall 0.3491
Household Information
minimum education of 6 years 0.2463 0.2465
minimum education of 10 years 0.2055 0.2059
minimum education of 11 years 0.1871 0.1875
maximum of 4 members 0.2065 0.2068
maximum of 6 members 0.2387 0.2389
Government Program Participation
Nutrition Education 0.2511 0.2514
Credit Grant 0.2404 0.2410
Food Production 0.2500 0.2502
Tindahan Natin 0.2507 0.2511
Botika ng Baranggay 0.2487 0.2490
Free Medical/Dental Check up 0.2576 0.2576
Produce Fruit Trees/Vegetables 0.2455 0.2456
Produce Livestock 0.2437 0.2438
Produce Fish 0.2536 0.2533
*v1 – mean vulnerability As a matter of fact, the said factors may as well serve as channels for climate shocks and instead of
helping decrease the effect of climate change through its adaptive capacity, it may have increased the
sensitivity of a household thus canceling its negative effect on the the potential impact of climate
change. A household that chooses to raise fruit-bearing trees or vegetables mainly for household
consumption or for income generation, will surely be affected whenever climate shocks exists.
Though, the main reason of production is to support household consumption, in cases where the
supply is not enough in the market or if the goods are available but not affordable, small-scale and
large-scale production of fruit and vegetables may also be affected by the same climate shocks. When
this happens, the household will not have any other option but to become more vulnerable. This is
why households that rely only on agriculture are expected to be more vulnerable.
Table 10. Simulated Impacts on Vulnerability Index
22 | P a g e
5 Conclusion and Recommendations
Food security may change from time to time. For policy makers, this poses a big challenge in
targeting the population for formulating interventions that will aid households or individuals to
adapt or adjust to actual or expected climate shocks. This study considered three food security
status: chronically food insecure, permanently food secure and temporarily food secure/insecure.
The crucial part of the study is to correctly classify each household by their current and future food
security status. It is very important specifically for policy makers to correctly classify the population
in order to develop a well-targeted policy intervention in the future.
For further research and development, other factors that link household to agriculture should be
studied to better capture each household’s sensitivity. Moreover, the adaptive capacity of each
household plays a very important role to accurately assess the possible potential impact of erratic
weather on food security. It will also be beneficial for the simulation of policies and formulation of
interventions with significant effect on the target population.
Further studies on the possible food substitution using panel data, e.g. FIES and its relation to
climate shocks could also be done. In order to explore the channels through which climate change
impacts to consumption of Filipino households pass through, it is recommended to obtain data from
production or from market to establish a strong significant relationship. Other policy interventions
may be considered like fertilizer, irrigation, sprays and other factors related to production. Other
outcomes like proportion of households that meet 100% adequacy of energy requirement may also
be explored.
23 | P a g e
Annex
Figure 3. Change in Maximum Temperature of 2008 relative to 1979-2006 average.
24 | P a g e
Figure 4. Change in Precipitation of 2008 relative to 1979-2006 average.
25 | P a g e
Table 4. Estimated Regression Coefficients using Increase in Maximum Temperature
log (food consumption) Coef. Std. Err. t P>t [95% Conf. Interval]
Ho
use
ho
ld In
form
atio
n
log (household size)*** -1.035 0.014 -74.790 0.000 -1.062 -1.008
log (age of hh head)*** 0.103 0.027 3.850 0.000 0.051 0.156
log (educational attainment of hh head)*** 0.034 0.011 3.020 0.003 0.012 0.056
sex of hh head(female)*** -0.080 0.020 -3.990 0.000 -0.119 -0.041
ownership of radio -0.012 0.015 -0.790 0.432 -0.041 0.018
ownership of television*** 0.061 0.020 3.060 0.002 0.022 0.100
presence of electricity 0.000 0.023 0.000 0.999 -0.045 0.045
ownership of telephone/cellphone**** 0.105 0.017 6.230 0.000 0.072 0.138
ownership of car 0.062 0.031 1.980 0.048 0.001 0.123
ownership of motor or bicycle 0.032 0.020 1.600 0.111 -0.007 0.071
Source of drinking water
waterworks-public tapped/peddler/delivered 0.041 0.026 1.550 0.121 -0.011 0.093
deep well - tubed/piped within residential area*** 0.102 0.025 4.060 0.000 0.053 0.151
deep well - public use 0.014 0.023 0.620 0.538 -0.031 0.060
private well with faucet 0.015 0.047 0.330 0.744 -0.076 0.107
open dug well - owned -0.005 0.054 -0.100 0.922 -0.111 0.100
open dug well - public 0.060 0.041 1.460 0.145 -0.021 0.141
spring water piped into residential area*** 0.083 0.030 2.790 0.005 0.025 0.142
rainwater 0.147 0.095 1.550 0.122 -0.039 0.333
spring water -0.013 0.032 -0.390 0.699 -0.076 0.051
mineral/bottled water 0.048 0.025 1.930 0.054 -0.001 0.096
others 0.280 0.141 1.990 0.047 0.004 0.557
Cimate Shocks increase in maximum temperature -0.033 0.095 -0.350 0.723 -0.219 0.152
decrease in minimum temperature 0.000 (omitted)
Go
vern
me
nt
Pro
gram
Par
tici
pat
ion
Nutrition Education -0.016 0.028 -0.580 0.559 -0.071 0.039
Credit grant 0.033 0.034 0.980 0.326 -0.033 0.100
Food Production -0.046 0.035 -1.340 0.181 -0.114 0.022
Tinadahan Natin -0.016 0.019 -0.860 0.390 -0.054 0.021
Botika ng Baranggay -0.006 0.017 -0.390 0.699 -0.039 0.026
Free Medical and Dental*** -0.057 0.016 -3.520 0.000 -0.089 -0.025
Produce Fruit trees/Vegetables 0.031 0.029 1.090 0.276 -0.025 0.087
Produce Livestock* 0.036 0.018 2.010 0.045 0.001 0.072
Produce Fish -0.027 0.043 -0.620 0.534 -0.112 0.058
Constant 6.917 0.175 39.520 0.000 6.574 7.260
R-Squared 56.17
Adjusted R-Squared 55.97
26 | P a g e
Table 5. Estimated Regression Coefficients using Increase in Percent Changes in Rainfall
log (food consumption)
Coef. Std. Err. t P>t [95% Conf. Interval]
Ho
use
ho
ld In
form
atio
n
log (household size)*** -1.034 0.014 -74.640 0.000 -1.062 -1.007 log (age of hh head)*** 0.102 0.027 3.810 0.000 0.050 0.155 log (educational attainment of hh head)*** 0.034 0.011 3.020 0.003 0.012 0.056 sex of hh head(female)*** -0.080 0.020 -3.970 0.000 -0.119 -0.040 ownership of radio -0.011 0.015 -0.770 0.444 -0.041 0.018 ownership of television*** 0.061 0.020 3.060 0.002 0.022 0.100 presence of electricity -0.001 0.023 -0.050 0.958 -0.046 0.044 ownership of telephone/cellphone*** 0.104 0.017 6.170 0.000 0.071 0.137 ownership of car 0.061 0.031 1.950 0.051 0.000 0.122 ownership of motor or bicycle 0.032 0.020 1.590 0.113 -0.008 0.071 Source of drinking water
waterworks-public tapped/peddler/delivered 0.038 0.026 1.430 0.154 -0.014 0.090 deep well - tubed/piped within residential area*** 0.101 0.025 3.990 0.000 0.051 0.150 deep well - public use 0.012 0.023 0.510 0.607 -0.034 0.058 private well with faucet 0.013 0.047 0.270 0.785 -0.079 0.105 open dug well - owned -0.008 0.054 -0.160 0.876 -0.115 0.098 open dug well - public 0.060 0.041 1.440 0.150 -0.022 0.141 spring water piped into residential area*** 0.080 0.030 2.680 0.007 0.022 0.139 rainwater 0.153 0.095 1.610 0.108 -0.034 0.340 spring water -0.014 0.033 -0.430 0.668 -0.078 0.050 mineral/bottled water 0.048 0.025 1.950 0.052 0.000 0.097 others 0.279 0.141 1.980 0.048 0.002 0.556
Cimate Shocks
Percent Change in Rainfall < -20% 0.000 (omitted) -20% < Percent Change in Rainfall < -10% 0.001 0.002 0.290 0.772 -0.003 0.004
-10% < Percent Change in Rainfall < 0% 0.001 0.003 0.440 0.663 -0.005 0.008 0 < Percent Change in Rainfall < 10%" -0.003 0.004 -0.610 0.545 -0.011 0.006 10% < Percent Change in Rainfall < 20% -0.001 0.002 -0.910 0.363 -0.005 0.002 Percent Change in Rainfall > 20% 0.000 0.001 0.390 0.694 -0.002 0.003
Go
vern
men
t P
rogr
am
Par
tici
pat
ion
Nutrition Education -0.017 0.028 -0.590 0.554 -0.072 0.039 Credit grant 0.031 0.034 0.910 0.363 -0.036 0.098 Food Production -0.047 0.035 -1.350 0.178 -0.116 0.022 Tinadahan Natin -0.017 0.020 -0.880 0.381 -0.056 0.021 Botika ng Baranggay -0.007 0.017 -0.440 0.662 -0.040 0.025 Free Medical and Dental*** -0.056 0.016 -3.470 0.001 -0.088 -0.024 Produce Fruit trees/Vegetables 0.033 0.029 1.160 0.247 -0.023 0.089 Produce Livestock* 0.037 0.018 2.050 0.041 0.002 0.073 Produce Fish -0.025 0.043 -0.580 0.561 -0.110 0.060
Constant 6.930 0.176 39.450 0.000 6.586 7.275
R-Squared 56.45 Adjusted R-Squared 56.14
27 | P a g e
Table 6. Estimated Regression Coefficients using Increase in Maximum Temperature*
log (food consumption) Coef. Std. Err. t P>t 95% Conf Interval
Ho
use
ho
ld In
form
ati
on
log (household size)*** -1.017 0.026 -39.600
28 | P a g e
Table 7. Estimated Regression Coefficients using Increase in Percent Changes in Rainfall*
log (food consumption) Coef. Std. Err. t P>t 95% Conf Interval
Ho
use
ho
ld In
form
ati
on
log (household size)*** -1.014 0.026 -39.470 20% 0.003 0.002 1.890 0.060 0.000 0.007
Go
vern
me
nt
Pro
gram
Par
tici
pa
tio
n
Nutrition Education -0.045 0.049 -0.930 0.355 -0.141 0.051 Credit grant 0.080 0.059 1.350 0.177 -0.036 0.195 Food Production -0.098 0.078 -1.260 0.208 -0.251 0.055 Tinadahan Natin -0.007 0.038 -0.170 0.862 -0.081 0.068 Botika ng Baranggay 0.002 0.029 0.060 0.953 -0.056 0.059 Free Medical and Dental -0.042 0.030 -1.390 0.163 -0.101 0.017 Produce Fruit trees/Vegetables 0.127 0.054 2.370 0.018 0.022 0.232 Produce Livestock 0.101 0.037 2.760 0.006 0.029 0.173 Produce Fish -0.003 0.065 -0.050 0.961 -0.130 0.123
Constant 6.865 0.229 29.980
29 | P a g e
References: Behrman, J.R., & Deolalikar, A.B. 2012. The intrahousehold Demand for Nutrients in Rural South India: Individual Estimated, Fixed Effects and Permanent Income. The journal of human resources. 25 (4): 665-696. Retrieved November 26, 2013. Bhattacharya, J., DeLeire, T., and Haider, & S., Currie, J, 2003. Heat or Eat? Cold-Weather shocks and Nutrition in Poor American Families. Am J Public Health; 93(7): 1149-1154. Retrieved September 11, 2013 Bickel, G., Nord, M., Price, C., Hamilton W., & Cook, J., (2000). “Guide to measuring household food security”. Food and nutrition Service. USDA. Capaldo, J., karfakis, P., Knowles, M., Smulders, M., 2010. A model of vulnerability to food security. FAO, Agricultural and Development Economics Division, Rome. Christaensen, L.J. & Boisvert R.N. (2000). “On measuring houisehold food vulnerability: Case evidence from Northern Mali”. Working paper No. 2005-05, Ithaca, NY: Cornell University Claro, R.M., Levy, R.B., Bandoni, D.H., Mondini, L. (2010). “per Capita versus adult-equivalent estimates of calorie availability in household budget surveys”. 26(11):2188-2195. Retrieved October 5, 2013. FAO. (1997). “The food system and factors affecting household food security and nutrition” . Agriculture, food and nutrition for Affrica: a resource book for teachers of Agriculture. Rome: Agriculture and Consumer Protection Department. Retrieved September 10, 2013 FNRI. (2008). “Facts and Figures”. Departemnt of Science and technology, Taguig, Philippines FNRI. (2003). “Facts and Figures”. Departemnt of Science and technology, Taguig, Philippines Food Security. (n.d.) Retrieved Oct 3, 2013 from The World Bank Web site http://www.who.int /trade/glossary/story028/en/ Hannah L, Ikegami, > Hole DG, Seo C, et al. (2013) Global Climate Chnage Adaptation Priorities for Biodiversity and Food Security. PloS ONE 8(8): e72590. doi: 10.1371/journal.pone.0072590. Ingram G.P.T., & Brklacich JSI., (2005). “Climate change and food security.” Philosophical Transactions of the royal society B: Biological Sciences 360(1463): 2139-2148. Retrieved September 10, 2013. Kailash. (2010). “Climate change, mining and MIMAROPA.” Philippine Indigenous Peoples Links. Retrieved May 20, 2014 from http://www.piplinks.org/Cantilan+protests. Lake I.R., Hooper, L., Abdelhamid, A., et. Al. 2012. Climate Change and Food Security: Health Impacts in Developed Countries. Environmental Health Perspectives. 120(11): 1520-1526. Retrieved September 24, 2013. Luther, T. (1999). “The economics of global food security.” Review of Agricultural Economics. 21(2):473-488 Retrieved September 10, 2013. Mangyo, E., 2008. Who benefits more from higher household consumption? The intrahousehold allocation of Nutrients in China. Journal of Development Economics. 86 (2008): 296-312. Retrieved November 26, 2013.
30 | P a g e
Senate of the Philippines. (2013). “Natural Disasters at a glance.” Retrieved May 12, 2014 from http://senate.gov.ph/publications/AAG%202013-04%20-%20Natural%20Disasters_final.pdf Orewa, S.I., & Iyanbe, C.O. 2010. Determinants of Daily Food Calorie intake among Rural and Low-Income Urban Households in Nigeria. 3(4): 147-155. Retrieved September 12, 2013 Piza, S.F., Safir, A., & Skoufias E, 2013. “Disquiet on the Weather Front- The welfare Impacts of Climatic Variability in the Rural Philippines”. Policy Research Working Pare. World Bank, Poverty Reduction and Equity Department. WFP. (n.d.) “An overview of Vulnerability Analysis and mapping (VAM)”. Retrieved October. 6, 2013