Relationship between Inflation and Food Insecurity

Published

November 27, 2025

Introduction

Analysis of Indonesian household data reveals a significant correlation between rising food prices (Inflation) and increased FIES indicators, both within and between regions

Only data from the last 12 months is used (November 2024 to October 2025) for these analyses.

The visualization above illustrates how local cost of living affects food insecurity, controlling for household income. The y-axis shows the percentage increase in risk of experiencing each food insecurity outcome associated with a one-unit increase in the local cost of living index.

The statistical model takes the form:

\[\text{logit}(P(\text{food\_insecurity\_measure} = 1)) = \beta_0 + \beta_1 \cdot \text{ln\_pc\_def\_inc} + \beta_2 \cdot \text{pdef3} + \varepsilon\]

Where:

  • food_insecurity_measure is a binary indicator (0/1) for each food insecurity outcome
  • ln_pc_def_inc is the log of per capita deflated income (controlling for household resources)
  • pdef3 is the local cost of living index derived from Susenas data
  • Standard errors are clustered at the household level

The y-axis values represent the percentage increase in the probability of experiencing food insecurity, calculated as:

\[\text{Percentage Increase} = (\exp(\beta_2) - 1) \times 100\%\]

For example, a value of 350% for “Whole day without food” means that a one-unit increase in the local cost of living index is associated with a 3.5 times higher probability of going a whole day without food, after controlling for income. In other words, households in areas with higher costs of living are substantially more likely to experience severe food insecurity, even when comparing households with similar income levels.

Model Summary Relation Inflation and Food insecurity

To test this relationship a conditional logistic regression model is used, implemented through the survival::clogit function, which incorporates fixed effects by stratifying observations within groups. This approach is complemented by a linear fixed effects regression model using fixest::feols, allowing for comparison between binary and continuous outcome specifications while controlling for time-invariant unobserved heterogeneity.

tables

Household Fixed Effects Linear Models
Household Fixed Effects Linear Models
Food Insecurity Outcomes
Worried Unhealthy Low Diversity Skip Meal Eat Less Ran Out Went Hungry Whole Day Totfoodinsec
Consumer Price Index -0.001 (0.005) -0.000 (0.004) -0.005 (0.004) -0.003 (0.003) 0.002 (0.003) 0.000 (0.003) 0.001 (0.003) 0.001 (0.002) -0.005 (0.015)
Num.Obs. 12971 12971 12971 12970 12970 12970 12970 12970 12996
* p < 0.1, ** p < 0.05, *** p < 0.01
Standard errors clustered at household level
Effect of Consumer Price Index (CPI) on Food Insecurity: Household Fixed Effects Logistic Models
Food Insecurity Measure Odds Ratio (95% CI) Std. Error
Worried about food 0.99 (0.90-1.09) 0.049
Unable to eat healthy 1.00 (0.90-1.11) 0.055
Low food diversity 0.94 (0.85-1.05) 0.054
Skipped meals 0.93 (0.80-1.08) 0.076
Ate less than needed 1.05 (0.92-1.20) 0.068
Ran out of food 1.01 (0.87-1.17) 0.075
Went hungry 1.02 (0.86-1.22) 0.091
Whole day without food 1.07 (0.82-1.40) 0.135
Total food insecurity 0.96 (0.85-1.08) 0.061
* * p < 0.1, ** p < 0.05, *** p < 0.01
† Odds ratios represent the effect of a one-unit increase in CPI on the likelihood of experiencing each food insecurity outcome
‡ Standard errors are robust and clustered at household level
§ Models estimated using conditional logistic regression (clogit)

Relationship between Inflation and Income

These regression models show the relationship between income and the different FIES components, revealing how economic factors such as household income, local food costs, and agricultural participation influence various dimensions of food insecurity. The analysis examines eight distinct food insecurity indicators—from worrying about food to going whole days without eating—and explores how these experiences vary across income quintiles and local food cost distributions. By regressing each food insecurity measure against per capita income, local food prices, and agricultural household status, we can identify which economic factors most strongly predict different manifestations of food insecurity in Indonesia.

Model Summary

\[ \begin{align} \text{Model 1: } & y_i = \beta_0 + \beta_1 \text{pdef3}_i + \varepsilon_i \\ \text{Model 2: } & y_i = \beta_0 + \sum_{j=1}^{4} \beta_j \mathbb{1}(\text{loc\_cost\_quint}_i = j+1) + \varepsilon_i \\ \text{Model 3: } & y_i = \beta_0 + \beta_1 \text{ln\_pc\_def\_inc}_i + \beta_2 \text{pdef3}_i + \varepsilon_i \\ \text{Model 4: } & y_i = \beta_0 + \beta_1 \text{hhag}_i + \beta_2 \text{ln\_pc\_def\_inc}_i + \beta_3 \text{pdef3}_i + \varepsilon_i\\ \text{Model 5a: } & \log\left(\frac{P(y_i=1)}{1-P(y_i=1)}\right) = \beta_0 + \beta_1 \text{ln\_pc\_def\_inc}_i + \beta_2 \text{pdef3}_i \\ \text{Model 5b: } & \log(E(y_i)) = \beta_0 + \beta_1 \text{ln\_pc\_def\_inc}_i + \beta_2 \text{pdef3}_i \end{align} \]

\[\text{where } y_i \text{ represents household-level FIES (Food Insecurity Experience Scale) outcomes}\] \[\text{Model 5a is used for binary outcomes and Model 5b for count data (when totfoodinsec > 1)}\]

Coefficient Definitions:

\[ \begin{align} \text{Model 1: } & \beta_0 = \text{Constant} \\ & \beta_1 = \text{Spatial deflator} \\ \\ \text{Model 2: } & \beta_0 = \text{Constant} \\ & \beta_1 = \text{Location cost quintile 2} \\ & \beta_2 = \text{Location cost quintile 3} \\ & \beta_3 = \text{Location cost quintile 4} \\ & \beta_4 = \text{Location cost quintile 5} \\ \\ \text{Model 3: } & \beta_0 = \text{Constant} \\ & \beta_1 = \text{Log per-capita income} \\ & \beta_2 = \text{Spatial deflator} \\ \\ \text{Model 4: } & \beta_0 = \text{Constant} \\ & \beta_1 = \text{Agricultural household} \\ & \beta_2 = \text{Log per-capita income} \\ & \beta_3 = \text{Spatial deflator} \\ \\ \text{Model 5a/5b: } & \beta_0 = \text{Constant} \\ & \beta_1 = \text{Log per-capita income} \\ & \beta_2 = \text{Spatial deflator} \end{align} \]

\[\text{Note: All models are weighted using population weights (popw) to ensure representativeness.}\] \[\text{All models use heteroskedasticity-consistent standard errors}\]

tables

Comparison of m1 across different food insecurity outcomes
Food Insecurity Outcomes
Worried Unhealthy Low Diversity Skip Meal Eat Less Ran Out Went Hungry Whole Day Totfoodinsec
Constant 0.148*** (0.038) 0.011 (0.028) 0.116*** (0.034) −0.015 (0.020) 0.043* (0.025) −0.016 (0.020) −0.057*** (0.014) −0.039*** (0.009) 0.193 (0.136)
Spatial deflator 0.142*** (0.040) 0.154*** (0.030) 0.093*** (0.035) 0.089*** (0.021) 0.071*** (0.026) 0.096*** (0.022) 0.106*** (0.016) 0.059*** (0.010) 0.808*** (0.147)
Num.Obs. 12971 12971 12971 12970 12970 12970 12970 12970 12996
R2 0.002 0.004 0.001 0.002 0.001 0.003 0.005 0.004 0.004
R2 Adj. 0.002 0.003 0.001 0.002 0.001 0.003 0.005 0.004 0.004
* p < 0.1, ** p < 0.05, *** p < 0.01
Comparison of m2 across different food insecurity outcomes
Food Insecurity Outcomes
Worried Unhealthy Low Diversity Skip Meal Eat Less Ran Out Went Hungry Whole Day Totfoodinsec
Constant 0.215*** (0.012) 0.115*** (0.009) 0.154*** (0.011) 0.065*** (0.007) 0.092*** (0.009) 0.046*** (0.006) 0.022*** (0.003) 0.007*** (0.001) 0.715*** (0.044)
Location cost quintile 2 0.090*** (0.020) 0.038*** (0.014) 0.103*** (0.018) −0.005 (0.010) 0.024* (0.013) 0.027** (0.010) 0.007 (0.006) 0.005 (0.003) 0.290*** (0.067)
Location cost quintile 3 0.075*** (0.018) 0.024* (0.014) 0.006 (0.015) −0.026*** (0.009) −0.003 (0.012) 0.020** (0.010) 0.012** (0.006) 0.001 (0.002) 0.108* (0.062)
Location cost quintile 4 0.059*** (0.019) 0.060*** (0.016) 0.062*** (0.018) 0.016 (0.013) 0.029* (0.015) 0.045*** (0.012) 0.041*** (0.011) 0.024*** (0.009) 0.334*** (0.089)
Location cost quintile 5 0.111*** (0.019) 0.082*** (0.015) 0.081*** (0.017) 0.033*** (0.011) 0.041*** (0.013) 0.052*** (0.011) 0.045*** (0.008) 0.021*** (0.004) 0.465*** (0.069)
Num.Obs. 12971 12971 12971 12970 12970 12970 12970 12970 12996
R2 0.007 0.006 0.011 0.006 0.003 0.005 0.008 0.006 0.009
R2 Adj. 0.007 0.006 0.010 0.006 0.003 0.005 0.008 0.006 0.009
* p < 0.1, ** p < 0.05, *** p < 0.01
Comparison of m3 across different food insecurity outcomes
Food Insecurity Outcomes
Worried Unhealthy Low Diversity Skip Meal Eat Less Ran Out Went Hungry Whole Day Totfoodinsec
Constant 1.267*** (0.100) 0.835*** (0.079) 1.031*** (0.089) 0.425*** (0.047) 0.622*** (0.067) 0.348*** (0.047) 0.203*** (0.033) 0.057*** (0.017) 4.780*** (0.348)
Spatial deflator 0.078* (0.041) 0.107*** (0.030) 0.041 (0.035) 0.063*** (0.021) 0.038 (0.026) 0.075*** (0.021) 0.091*** (0.016) 0.054*** (0.010) 0.545*** (0.144)
Log per-capita income −0.079*** (0.006) −0.058*** (0.005) −0.064*** (0.006) −0.031*** (0.003) −0.041*** (0.004) −0.026*** (0.003) −0.018*** (0.002) −0.007*** (0.001) −0.323*** (0.024)
Num.Obs. 12967 12967 12967 12966 12966 12966 12966 12966 12992
R2 0.032 0.029 0.026 0.018 0.018 0.012 0.014 0.007 0.038
R2 Adj. 0.032 0.029 0.026 0.017 0.018 0.012 0.013 0.007 0.038
* p < 0.1, ** p < 0.05, *** p < 0.01
Comparison of m4 across different food insecurity outcomes
Food Insecurity Outcomes
Worried Unhealthy Low Diversity Skip Meal Eat Less Ran Out Went Hungry Whole Day Totfoodinsec
Constant 1.266*** (0.100) 0.838*** (0.079) 1.036*** (0.089) 0.435*** (0.048) 0.636*** (0.067) 0.361*** (0.048) 0.206*** (0.034) 0.057*** (0.017) 4.826*** (0.353)
Spatial deflator 0.079* (0.040) 0.106*** (0.030) 0.039 (0.035) 0.059*** (0.021) 0.031 (0.026) 0.069*** (0.022) 0.090*** (0.016) 0.054*** (0.010) 0.523*** (0.145)
Log per-capita income −0.079*** (0.006) −0.058*** (0.005) −0.064*** (0.006) −0.031*** (0.003) −0.041*** (0.004) −0.026*** (0.003) −0.018*** (0.002) −0.007*** (0.001) −0.324*** (0.024)
Agricultural household 0.002 (0.015) −0.004 (0.011) −0.008 (0.014) −0.015** (0.007) −0.021** (0.010) −0.020** (0.008) −0.004 (0.006) −0.001 (0.004) −0.069 (0.053)
Num.Obs. 12967 12967 12967 12966 12966 12966 12966 12966 12992
R2 0.032 0.029 0.026 0.018 0.019 0.013 0.014 0.007 0.038
R2 Adj. 0.032 0.029 0.026 0.018 0.018 0.013 0.013 0.007 0.038
* p < 0.1, ** p < 0.05, *** p < 0.01
Logistic and Poisson Regression Models for Food Insecurity Measures
Food Insecurity Outcomes
Worried Unhealthy Low Diversity Skip Meal Eat Less Ran Out Went Hungry Whole Day Totfoodinsec
Constant 4.115*** (0.740) 4.364*** (0.839) 4.463*** (0.802) 3.631*** (1.075) 4.428*** (0.986) 2.818*** (1.085) 2.585** (1.199) 0.925 (1.524) 4.362*** (0.576)
Spatial deflator −0.039 (0.236) 0.414 (0.254) 0.091 (0.258) 0.712** (0.310) 0.188 (0.314) 0.642** (0.308) 1.106*** (0.316) 1.461*** (0.395) 0.280 (0.195)
Log per-capita income −0.376*** (0.051) −0.478*** (0.058) −0.445*** (0.055) −0.516*** (0.078) −0.507*** (0.070) −0.449*** (0.078) −0.504*** (0.090) −0.479*** (0.116) −0.352*** (0.040)
Num.Obs. 12967 12967 12967 12966 12966 12966 12966 12966 12992
* p < 0.1, ** p < 0.05, *** p < 0.01

Figure comparing coefficient different models

Variables used on this page

Variables Used in Food Insecurity and Inflation Analysis

Food Insecurity Measures (FIES)

  • totfoodinsec: Total food insecurity score Individual FIES indicators:

  • worried: Worried about not having enough food

  • unhealthy: Unable to eat healthy/nutritious foods

  • low_diversity: Limited variety in diet

  • skip_meal: Skipped meals

  • eat_less: Ate less than should have

  • ran_out: Ran out of food

  • went_hungry: Went hungry

  • whole_day: Went whole day without eating

Price and Inflation Variables

  • CPI: Regional inflation measure
  • pdef3: Spatial deflator/local cost of living index
  • loc_cost_quint: Location cost quintiles (1-5)

Socioeconomic Variables

  • ln_pc_def_inc: Log of per capita deflated income
  • quint: Income quintile (1=Poorest to 5=Richest)
  • hhag: Agricultural household indicator
  • popw: Population weight for representative statistics