## _wp16247

## Source details

**Canonical URL:** [_wp16247](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16247.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16247.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16247.pdf.json)

---

### I. Introduction and motivation
- Quotation framing policy perspective: “Because shocks to commodity price inflation are typically beyond the control of policymakers, hard to predict, and often not sustained, central banks seeking to establish credibility are generally better off setting and communicating their monetary policy in terms of underlying inflation rather than headline inflation.”— IMF, 2011
- Literature consensus: central banks often “look through” supply shocks (food and energy) and rely on core inflation measures (Motley, 1997; Mishkin, 2007; Kiley, 2008; Mohanty, 2014).
- Concern motivating analysis: food price inflation may be structurally different in LICs, especially in SSA, due to a higher share of food in consumption baskets and differing persistence behavior (Walsh, 2011; Rangasamy, 2011; Durevall et al., 2013; Furceri et al., 2015).
- Consumption-basket shares reported: share of food is 40 percent on average in SSA, 15 percent in AEs, and 30 percent in EMs (Figure I.1).

### II. Research questions and approach
- Comparative and causal questions:
  - Compare average food inflation with non-food non-fuel (NF/NF) inflation in SSA.
  - Compare volatility and persistence of food vs. NF/NF inflation.
  - Test whether food inflation Granger-causes NF/NF inflation in SSA.
  - Assess pass-through from international food prices, fuel prices, and exchange rates to domestic food inflation.
  - Using disaggregated food items, evaluate heterogeneous effects of external drivers on different food items.
- Sample period: 2000 to 2016.

### III. Data and sample features
- Dataset 1 (41 SSA countries, 2000–2016):
  - Monthly headline CPI and 12 major subcomponents.
  - CPI subcomponents include: 1 Food and Nonalcoholic Beverages; 4 Housing, Water, Electricity, Gas, and other Fuels.
  - NF/NF computed by subtracting item 1 and item 4 from headline CPI after weighting and reweighing.
  - Average observations: 171 months.
  - Minimum observations: 114 (Republic of Congo).
  - Maximum observations: 197 (Benin, Botswana, Central African Republic, Ghana, Guinea, Kenya, Madagascar, Nigeria, and Togo).
- Dataset 2 (20 SSA countries, detailed food subcomponents):
  - Monthly detailed food subcomponents (up to 91 items) and CPI weights, aggregated into 9 food categories: bakery, coffee, fish, fruit, meat, milk, oils, sugar, and vegetables.
  - The 9 categories make up on average 87 percent of the food basket.
  - Average length: around 8.5 years of monthly data.
  - Minimum: just under 4 years (Ghana).
  - Maximum: nearly 15 years (Madagascar).
- Income classification (2015 per capita GDP (PPP)):
  - > US$5,000 = middle-income; otherwise low-income.
  - Dataset 1: 11 middle-income and 30 low-income countries.
  - Dataset 2: 5 middle-income and 15 low-income countries.

### IV. Inflation dynamics and stylized facts (2000–2016)
- Headline consumer price inflation in SSA trended down over two decades but remains above other regions; inflation dispersion in SSA is high, with about a quarter of countries experiencing increasing inflation trends (Figure III.1).
- Average annualized quarterly inflation (from monthly indices):
  - Food inflation greater than NF/NF inflation for all main sub-groupings.
  - Full sample (41 countries): annual average difference ~ 1.8 percentage points during 2000–16.
  - Cumulative effect over 16 years: food inflation cumulatively higher by over thirty percentage points relative to NF/NF inflation.
  - Largest differences: members of CEMAC and EAC, difference about 2½ percentage points per year on average.
  - LICs show higher food inflation relative to NF/NF by around 2.1 percentage points annually.
- Post-2007–08 dynamics:
  - The wedge between food and NF/NF inflation narrowed since 2009 to close to 1.4 percentage points annually.
  - Around sixty percent of the decline in the wedge has been driven by lower food inflation.
  - For 2009-16 the share of SSA countries with food inflation exceeding NF/NF inflation drops to just over 70 percent (from 90 percent in 2000-16).
  - Quarterly comparisons: about ¾ of countries have quarterly food inflation greater than NF/NF inflation more than half the time over 2000-16; on average quarterly food inflation is greater than quarterly NF/NF inflation about sixty percent of the time.

### V. Volatility of inflation
- Volatility measure: coefficient of variation (standard deviation divided by the mean multiplied by 100), computed quarterly; median volatility reported for groupings.
- Key findings:
  - The volatility of food inflation in SSA is over 50 percent higher than that of NF/NF inflation.
  - The difference in volatilities is driven by LICs; difference negligible for MICs.
  - Difference highest for WAEMU members; for CEMAC and the oil exporters food inflation is less volatile than NF/NF inflation on average.
  - Since 2009 the difference of median volatilities of food and NF/NF inflation has increased slightly in SSA, driven by a nearly 50 percent increase in the difference for LICs.
  - Heterogeneity: close to half of countries have broadly similar volatilities for food and NF/NF inflation across both 2000-16 and 2009-16.

### VI. Persistence of inflation (SARC estimates) and drivers
- Persistence measured using the sum of autoregressive coefficients (SARC) from country-by-country univariate autoregressions of annualized quarterly inflation (lags 1–4; lag selection by AIC), controlling for quarterly dummies.
- Main SARC findings:
  - Medians of SARC estimates across countries show over 2000-16 food inflation is more persistent than NF/NF inflation in SSA.
  - From 2009 onwards, food inflation is significantly less persistent than NF/NF inflation; decrease driven by LICs.
  - Distributional changes: food inflation SARC distribution shifted down and leftwards from 2000-16 to 2009-16; NF/NF SARC distribution flattened with mean broadly the same.
  - Country-level shares:
    - In 2000-16 around 60 percent of countries experienced food inflation more persistent than NF/NF inflation.
    - In 2009-16 close to 70 percent of countries witnessed a fall in the persistence of food inflation since 2000-16.
    - In 2009-16 food inflation is less persistent than NF/NF inflation in over 60 percent of countries.
- Cross-sectional regression (2009-16) on wedge between food and NF/NF persistence includes CPIA Macro index, inflation targeting dummy, GDP per capita, de jure central bank independence categorical variable, total droughts, logistics index, and share of China imports.
  - Results:
    - Having an inflation target and better macroeconomic management (higher CPIA Macro index) associated with a smaller wedge between persistence of food and NF/NF inflation.
    - Decrease in wedge driven by negative correlation between institutional variables and food persistence — improved monetary policy management appears to reduce food persistence via better anchoring of inflation expectations.
    - Two controls significant:
      - Higher share of imports from China associated with a lower wedge (driven by more persistent NF/NF inflation).
      - Having an exchange rate peg associated with a lower wedge (driven by less persistent food inflation).
    - Remaining structural variables not statistically significant.

### VII. Transmission between food and NF/NF inflation (Granger causality)
- Method: Granger-causality tests using VAR with six lags and monthly time dummies.
- Findings:
  - Second-order effects from food inflation on NF/NF inflation are limited.
  - Only a fifth of the sample shows one-way Granger-causality from food to NF/NF inflation.
  - In another fifth of the sample Granger causation runs both ways.
  - Monetary policy framework does not make a clear difference: food inflation Granger-causes NF/NF inflation in both forward-looking policy framework countries and in countries with exchange rate pegs.
- Conclusion: Neither food inflation nor NF/NF inflation consistently drives the other in SSA; second-order effects are limited.

### VIII. External drivers of domestic food inflation: pass-throughs (aggregate sample)
- Estimation: country-by-country quarterly regressions of domestic food inflation on current and 4 lags of world food inflation, world fuel inflation, and nominal exchange rate depreciation, controlling for 4 lags of domestic food inflation, time trend, and quarterly dummies. Pass-through computed as sum of coefficients on each foreign variable.
- Average pass-throughs (whole sample):
  - World food inflation: 32 percent.
  - World fuel inflation: 2 percent.
  - Exchange rate depreciation: 17 percent.
  - Interpretation: a world food inflation pass-through of 32% indicates that for a 100% increase in world food prices in a given year, domestic food prices would increase by 32% within that year.
- Country-specific prevalence:
  - Over one third of countries have statistically significant pass-throughs from world food inflation.
  - Over one third of countries have statistically significant pass-throughs from exchange rate depreciation.
  - Less than ten percent have statistically significant pass-throughs from world fuel inflation.
- Distributional patterns:
  - World food pass-through coefficients broadly centered between 0 and 40 percent, with many countries higher.
  - World fuel pass-throughs nearly all below 20 percent.
  - Exchange rate pass-throughs broadly centered between 0 and 40 percent.
- Interpretation: Incomplete pass-through in SSA from world food and fuel prices, and exchange rates to domestic food prices.

### IX. Disaggregated pass-through analysis (20-country detailed food sample)
- Method: pass-through regression reduced to 1 lag for degrees of freedom; nine food categories represent about 87% of the food basket on average.
- Estimated average pass-through coefficients (simple averages across the 20 countries, excluding outliers as noted):
  - World food inflation → domestic food inflation: 19 percent.
  - World fuel inflation → domestic food inflation: 3 percent.
  - Nominal exchange rate depreciation → domestic food inflation: 14 percent.
- Fresh vs non-fresh foods:
  - Pass-through to fresh food: 10 percent.
  - Pass-through to non-fresh food: 24 percent.
  - Statistical significance:
    - Only 15% of the pass-through coefficients for fresh food were statistically significant (3 countries).
    - Over a third of countries had a statistically significant pass-through to non-fresh foods.
    - Both fresh and non-fresh categories had close to half of the countries possessing a statistically significant pass-through from exchange rates.
- Additional finding: pass-through from world fuel prices is broadly negligible in the disaggregated sample.

### X. Structural/domestic drivers and food composition
- Key findings from 20-country sample:
  - Higher domestic food inflation is broadly driven by fresh (non-tradable) food items across SSA, in both MICs and LICs.
  - The difference between fresh food inflation and non-fresh food inflation is greatest for LICs at 2.7 percentage points.
  - Around three quarters of countries in the 20-country sample have fresh food inflation greater than non-fresh food inflation.
  - Fresh food inflation is consistently less volatile than non-fresh food inflation; for three quarters of the sample fresh food inflation is less volatile than non-fresh.
- Implication: The wedge between food and NF/NF inflation is driven to a great extent by fresh food items, pointing to domestic supply, storage, transport, and market structure factors.

### XI. Policy implications and recommendations
- On inflation measurement and monetary stance:
  - Measures of inflation that exclude food prices would likely underestimate underlying price pressures in SSA and other LICs where food inflation remains significantly high and persistent.
  - Central banks and fiscal authorities in SSA and other LICs should take food prices into consideration when assessing the appropriate stance of monetary and fiscal policy.
  - Using measures of core inflation that exclude food prices to assess the monetary stance would not be appropriate in many SSA countries, especially where persistence of food prices is as high as that of NF/NF prices (currently over a third of SSA countries).
- Monetary policy actions:
  - Continue to monitor food inflation and seek to reduce the second-round impact of food price shocks on NF/NF prices.
  - Focus on establishing credibility, as credibility may help reduce the persistence of food inflation.
- Structural and trade policies:
  - Improve the tradability of food by reducing tariffs.
  - Improve transport and storage infrastructure to reduce perishability-driven price pressures.
  - Implement structural measures to increase productivity in the agricultural sector to decrease food price pressures and improve welfare.
- Institutional implication:
  - Given heterogeneity across countries and items, central banks in SSA should consider country-specific characteristics (e.g., share of food in consumption basket, tradability of food items) when interpreting core inflation measures.

### XII. Suggestions for further research
- Investigate the relationship between improved central bank credibility and the persistence of food inflation, given rapid modernization of central banks in SSA and other LICs.
- Explore demand and supply determinants of food inflation in LICs to support more precise policy prescriptions for SSA and other LICs.

### XIII. Summary conclusions for SSA (2000–16)
- Average food inflation is consistently higher than NF/NF inflation during 2000-16.
- Food inflation is on average more volatile than NF/NF inflation throughout 2000-16.
- Food price shocks have been at least as persistent as NF/NF price shocks on average during 2000-16. However, since 2009 the persistence of food price shocks has declined markedly for the majority of countries in SSA; findings suggest improved central bank credibility and monetary policy management have played an important role.
- Pass-through from international prices and exchange rates to domestic food prices is incomplete; domestic drivers — especially fresh food components — play a major role.
- Granger-causality analysis indicates limited second-order effects from food to NF/NF inflation; neither series robustly drives the other.

*Source: _wp16247 - Box 3.3 in IMF (2016).*

### References ___________________________________________________________________________  25

### _wp16247 - References ___________________________________________________________________________  25

### I. Introduction and motivation
- Quotation framing policy perspective: “Because shocks to commodity price inflation are typically beyond the control of policymakers, hard to predict, and often not sustained, central banks seeking to establish credibility are generally better off setting and communicating their monetary policy in terms of underlying inflation rather than headline inflation.”— IMF, 2011
- Literature consensus: central banks often “look through” supply shocks (food and energy) and rely on core inflation measures (Motley, 1997; Mishkin, 2007; Kiley, 2008; Mohanty, 2014).
- Recent concern: food price inflation may be structurally different in LICs, especially in SSA, due to a higher share of food in consumption baskets and differing persistence behavior (Walsh, 2011; Rangasamy, 2011; Durevall et al., 2013; Furceri et al., 2015).
- Consumption-basket shares: share of food is 40 percent on average in SSA, 15 percent in AEs, and 30 percent in EMs (Figure I.1).

### II. Research questions and approach
- Compare average food inflation with non-food non-fuel (NF/NF) inflation in SSA.
- Compare volatility and persistence of food vs. NF/NF inflation.
- Test whether food inflation Granger-causes NF/NF inflation in SSA.
- Assess pass-through from international food prices, fuel prices, and exchange rates to domestic food inflation.
- Using disaggregated food items, evaluate heterogeneous effects of external drivers on different food items.
- Sample period: 2000 to 2016.

### III. Key data features
- Dataset 1: Monthly headline CPI and 12 major subcomponents for 41 SSA countries (2000–2016).
  - CPI subcomponent definitions include: 1 Food and Nonalcoholic Beverages; 4 Housing, Water, Electricity, Gas, and other Fuels; NF/NF computed by subtracting item 1 and item 4 from headline CPI after weighting and reweighing.
  - Average observations: 171 months (over 14 years).
  - Minimum observations: 114 (Republic of Congo).
  - Maximum observations: 197 (Benin, Botswana, Central African Republic, Ghana, Guinea, Kenya, Madagascar, Nigeria, and Togo).
- Dataset 2: Monthly detailed food subcomponents (up to 91 items) and CPI weights for 20 SSA countries.
  - Aggregated into 9 food categories for comparability: bakery, coffee, fish, fruit, meat, milk, oils, sugar, and vegetables.
  - These 9 categories make up on average 87 percent of the food basket.
  - Average length: around 8.5 years of monthly data.
  - Minimum: just under 4 years (Ghana).
  - Maximum: nearly 15 years (Madagascar).
- Income classification: 2015 per capita GDP (PPP) > US$5,000 = middle-income; otherwise low-income.
  - In Dataset 1: 11 middle-income and 30 low-income countries.
  - In Dataset 2: 5 middle-income and 15 low-income countries.
- Other macroeconomic and structural indicators used (Annex 2 Table A.1).

### IV. Inflation dynamics and stylized facts (2000–2016)
- Headline consumer price inflation in SSA has trended down over two decades but remains above other regions; inflation dispersion in SSA is high, with about a quarter of countries experiencing increasing inflation trends (Figure III.1).
- Average annualized quarterly inflation (derived from monthly indices) shows food inflation greater than NF/NF inflation for all main sub-groupings.
  - Full sample (41 countries): annual average difference ~ 1.8 percentage points during 2000–16.
  - Cumulative effect over 16 years: food inflation cumulatively higher by over thirty percentage points relative to NF/NF inflation.
  - Largest differences: members of CEMAC and EAC, difference about 2½ percentage points per year on average.
  - MICs similar to AEs; LICs show higher food inflation relative to NF/NF by around 2.1 percentage points annually.

### V. Principal empirical findings
- Comparative measures (2000–2016):
  - Food inflation has been higher than, more volatile than, and at least as persistent as NF/NF inflation, especially in LICs.
  - After 2007–2008 international food price surge, from 2009 onwards, food inflation on average remained higher than NF/NF inflation by 1.4 percentage points per year and more volatile.
  - On average, however, food inflation is significantly less persistent than NF/NF inflation.
- Drivers of persistence:
  - Regression results suggest improvements in macroeconomic management across SSA, and adoption of forward-looking monetary policy frameworks in some countries, are associated with less persistent food inflation relative to NF/NF inflation.
- Granger causality:
  - Tests indicate that, across the majority of SSA countries, food inflation does not cause NF/NF inflation, nor does NF/NF inflation always cause food inflation.
- Pass-through to domestic food prices:
  - Over a third of SSA countries experienced statistically significant pass-through to domestic food prices from changes in world food prices and exchange rates.
  - Average pass-throughs: 32 percent from world food prices and 17 percent from exchange rates.
- Heterogeneity by item and tradability:
  - Higher food inflation in SSA is most prominent in LICs.
  - Food inflation is higher in fresh (non-tradable) food items.
  - More countries show statistically significant pass-through from world food prices and exchange rates to non-fresh food prices.

### VI. Policy implications and recommendations
- Measures of inflation that exclude food prices would likely underestimate underlying price pressures in SSA and other LICs where food inflation remains significantly high and persistent.
- Monetary policy should:
  - Continue to monitor food inflation and seek to reduce the second-round impact of food price shocks on NF/NF prices.
  - Focus on establishing credibility, as credibility may help reduce the persistence of food inflation.
- Structural and trade-related policies:
  - Policies that increase tradability of food items—such as reducing tariffs and improving storage and transportation infrastructure—could potentially reduce food inflation pressures in LICs.
- Institutional implications:
  - Given heterogeneity across countries and items, central banks in SSA should consider country-specific characteristics (e.g., share of food in consumption basket, tradability of food items) when interpreting core inflation measures.

*Source: _wp16247 - References ___________________________________________________________________________  25*

### Box 3.3 in IMF (2016).

### Box 3.3 — Food Inflation, NF/NF Inflation, Volatility, Persistence, and Drivers in SSA (2000–16)

### A. Food vs NF/NF inflation: wedge and prevalence
- The wedge between food and NF/NF inflation has narrowed since 2009 (i.e., following the international food price surge of 2007-08), but is still close to 1.4 percentage points annually.
- Around sixty percent of this decline in the wedge has been driven by lower food inflation.
- Inflation for most food sub-groupings experienced a reduction since 2009 with the exception of EAC countries where the difference between food and NF/NF inflation actually increased to just over 3 percentage points per year.
- Country-specific averages (Annex 2 Table A.5):
  - 90 percent of SSA countries from 2000 to 2016 experienced food inflation exceeding NF/NF inflation.
  - For the 2009-16 period the figure drops to just over 70 percent of SSA countries witnessing food inflation exceeding NF/NF inflation.
- Quarterly comparisons:
  - About ¾ of countries in SSA have quarterly food inflation greater than NF/NF inflation more than half the time over the period 2000-16.
  - On average in SSA, quarterly food inflation is greater than quarterly NF/NF inflation about sixty percent of the time.
- Interpretation:
  - Food inflation was an important contributor to headline inflation in the vast majority of SSA countries from 2000 to 2016.
  - Since 2009 food inflation appears less significant, likely reflecting exclusion of the 2007-08 surge and factors such as increased resilience to food price shocks and more modernized central banks and improved macroeconomic management.

### B. Volatility of inflation
- Volatility measure: coefficient of variation (standard deviation divided by the mean multiplied by 100), computed quarterly for each country; median volatility reported for groupings.
- Key findings:
  - The volatility of food inflation in SSA is over 50 percent higher than that of NF/NF inflation.
  - The difference in volatilities of food and NF/NF inflation is driven by LICs; the difference is negligible for MICs.
  - The difference is highest for WAEMU members; for CEMAC and the oil exporters food inflation is less volatile than NF/NF inflation on average.
  - The difference of the median volatilities of food inflation and NF/NF inflation has increased slightly relative to NF/NF inflation since 2009 in SSA, driven by a nearly 50 percent increase in the difference for LICs.
- Heterogeneity:
  - Country-specific inflation volatility (Annex 2 Table A.6) reveals considerable heterogeneity; across both 2000-16 and 2009-16 close to half of countries have broadly similar volatilities for food and NF/NF inflation.
- Conclusion:
  - Food inflation in SSA is, on average, more volatile than NF/NF inflation during both the 2000-16 and 2009-16 periods, but heterogeneity across countries is considerable.

### C. Persistence of inflation (SARC estimates)
- Method:
  - Persistence measured using the sum of autoregressive coefficients (SARC) from country-by-country univariate autoregressions of annualized quarterly inflation (lags 1–4; lag selection by AIC), controlling for quarterly dummies.
- Main findings:
  - Medians of SARC estimates across countries show that over 2000-16 food inflation is more persistent than NF/NF inflation in SSA.
  - From 2009 onwards (i.e., after stripping out the 2007-08 international food price shock), food inflation is significantly less persistent than NF/NF inflation; this is driven by decreasing persistence of food inflation in LICs.
  - Distributional changes:
    - The distribution of food inflation SARC estimates shifted down and leftwards from 2000-16 to 2009-16 (decreasing persistence broadly across SSA).
    - NF/NF SARC distribution flattened with mean broadly the same.
  - Country-level shares:
    - In 2000-16 around 60 percent of countries experienced food inflation more persistent than NF/NF inflation.
    - In 2009-16 food inflation has become much less persistent, with close to 70 percent of countries in SSA witnessing a fall in the persistence of food inflation since 2000-16.
    - In 2009-16 food inflation is less persistent than NF/NF inflation in over 60 percent of countries.
- Drivers of the change in persistence (2009-16 cross-sectional analysis):
  - Regression on the wedge between food and NF/NF persistence includes: CPIA Macro index (macroeconomic management), inflation targeting dummy, GDP per capita, de jure central bank independence categorical variable, total droughts, logistics index, and share of China imports.
  - Results:
    - Having an inflation target and better macroeconomic management (higher CPIA Macro index) is associated with a smaller wedge between the persistence of food and NF/NF inflation.
    - The decrease in the wedge is driven by a negative correlation between these institutional variables and food persistence — i.e., improved monetary policy management appears to reduce food persistence via better anchoring of inflation expectations.
    - Two controls statistically significant:
      - Higher share of imports from China associated with a lower wedge (driven by more persistent NF/NF inflation).
      - Having an exchange rate peg associated with a lower wedge (driven by less persistent food inflation).
    - Remaining structural variables not statistically significant, possibly due to small sample size or collinearity.
- Conclusion:
  - Shocks to food prices were at least as persistent as shocks to NF/NF prices in SSA during 2000-16, but since 2009 shocks to food prices are less persistent for the majority of countries; improved policies promoting price stability and anchoring inflation expectations reduce the persistence of food inflation relative to NF/NF inflation.

### D. Transmission between food and NF/NF inflation (Granger causality)
- Method:
  - Granger-causality tests using VAR with six lags and monthly time dummies (Annex 2 Table A.12).
- Findings:
  - Second-order effects from food inflation on NF/NF inflation are limited.
  - Only a fifth of the sample shows one-way Granger-causality from food to NF/NF inflation.
  - In another fifth of the sample Granger causation runs both ways.
  - Monetary policy framework does not make a clear difference: food inflation Granger-causes NF/NF inflation in both forward-looking policy framework countries (e.g., Kenya, South Africa) and in countries with exchange rate pegs (e.g., Cameroon, Gabon, Mali).
- Conclusion:
  - Neither food inflation nor NF/NF inflation consistently drives the other in SSA; second-order effects are limited.

### E. External drivers of domestic food inflation: pass-throughs
- Estimation:
  - Country-by-country quarterly regressions of domestic food inflation on current and 4 lags of world food inflation, world fuel inflation, and nominal exchange rate depreciation, controlling for 4 lags of domestic food inflation, time trend, and quarterly dummies.
  - Pass-through computed as sum of coefficients on each foreign variable.
- Average pass-throughs (whole sample):
  - World food inflation: 32 percent.
  - World fuel inflation: 2 percent.
  - Exchange rate depreciation: 17 percent.
  - (Interpretation given in source: A world food inflation pass-through of 32% indicates that for a 100% increase in world food prices in a given year, domestic food prices would increase by 32% within that year.)
- Country-specific findings:
  - Over one third of countries have statistically significant pass-throughs from world food inflation.
  - Over one third of countries have statistically significant pass-throughs from exchange rate depreciation.
  - Less than ten percent have statistically significant pass-throughs from world fuel inflation.
- Distributional patterns:
  - World food pass-through coefficients broadly centered between 0 and 40 percent, with many countries higher.
  - World fuel pass-throughs nearly all below 20 percent.
  - Exchange rate pass-throughs similar to world food, broadly centered between 0 and 40 percent.
- Interpretation:
  - There is incomplete pass-through in SSA from world food and fuel prices, and exchange rates to domestic food prices.

### F. Structural/domestic drivers and food composition
- Data: 20 SSA countries with detailed weights for nine food subcomponents.
- Key findings:
  - Higher domestic food inflation is broadly driven by fresh (non-tradable) food items across SSA, in both MICs and LICs.
  - The difference between fresh food inflation and non-fresh food inflation is greatest for LICs at 2.7 percentage points.
  - Around three quarters of countries in the 20-country sample have fresh food inflation greater than non-fresh food inflation (Annex 2 Table A.9).
  - Fresh food inflation is consistently less volatile than non-fresh food inflation (Table IV.3); for three quarters of the sample fresh food inflation is less volatile than non-fresh (Annex 2 Table A.10).
- Implication:
  - The wedge between food and NF/NF inflation is driven to a great extent by fresh food items, pointing to domestic supply, storage, transport, and market structure factors.

### G. Summary conclusions for SSA (2000–16)
- Average food inflation is consistently higher than NF/NF inflation during 2000-16.
- Food inflation is on average more volatile than NF/NF inflation throughout 2000-16.
- Food price shocks have been at least as persistent as NF/NF price shocks on average during 2000-16. However, since 2009 the persistence of food price shocks has declined markedly for the majority of countries in SSA; findings suggest improved central bank credibility and monetary policy management have played an important role.
- Pass-through from international prices and exchange rates to domestic food prices is incomplete; domestic drivers — especially fresh food components — play a major role.
- Granger-causality analysis indicates limited second-order effects from food to NF/NF inflation; neither series robustly drives the other.

*Source: Box 3.3 in IMF (2016), _wp16247 - Box 3.3 in IMF (2016)._

### Box 3.3, IMF (2016).

### Box 3.3, IMF (2016)

### Disaggregated pass-through analysis (20-country detailed food sample)
- Sample and methodology:
  - Regression: pass-through regression estimated as in equation (2) but with fewer lags (reduced number of lags to 1 to ensure sufficient degrees of freedom).
  - Sample: nine food categories (plus fresh and non-fresh foods composite) for the 20 SSA countries in the second dataset.
  - Weights: country-specific food weights used; domestic food inflation based solely on the combined inflation of the nine food categories.
  - Representativeness: the nine food categories represent about 87% of the food basket on average (Annex 2 Table A.3).

- Estimated average pass-through coefficients (simple averages across the 20 countries, excluding outliers as noted):
  - World food inflation → domestic food inflation: 19 percent.
  - World fuel inflation → domestic food inflation: 3 percent.
  - Nominal exchange rate depreciation → domestic food inflation: 14 percent.

- Fresh vs non-fresh foods:
  - Pass-through to fresh food: 10 percent.
  - Pass-through to non-fresh food: 24 percent.
  - Statistical significance:
    - Only 15% of the pass-through coefficients for fresh food were statistically significant (3 countries).
    - Over a third of countries had a statistically significant pass-through to non-fresh foods.
    - Both fresh and non-fresh categories had close to half of the countries possessing a statistically significant pass-through from exchange rates.

- Additional notes:
  - The pass-through from world fuel prices is broadly negligible.
  - Summary results in Table IV.4 are based on simple averages for the 20 countries excluding the outliers highlighted in the table notes.

### Key empirical findings (aggregated results and contrasts)
- Differences in inflation levels, volatility, and persistence:
  - Over the period 2000 to 2016, food inflation has been on average 1.8 percentage points per year higher than NF/NF inflation.
  - Food inflation has been more volatile than NF/NF inflation.
  - Food inflation has been at least as persistent as NF/NF inflation.
  - Excluding the 2007-08 global food price surge (i.e., since 2009) food inflation has, on average, become less persistent.
  - Heterogeneity: despite aggregate patterns, many individual SSA countries show different dynamics; for many countries food inflation has become more persistent even though on average persistence fell since 2009.

- Pass-through magnitudes and cross-country prevalence:
  - Over a third of SSA countries have experienced statistically significant pass-through to domestic food prices from world food prices and exchange rates.
  - Average pass-through (aggregate statement): world food prices → domestic food prices: 32 percent; exchange rates → domestic food prices: 17 percent.
  - Higher food inflation is largely driven by fresh (and non-tradable) food items.
  - More countries show statistically significant pass-through from world food prices and exchange rates to non-fresh food prices than to fresh food prices.

- Causality:
  - Granger causality tests do not yield a definitive conclusion on whether food inflation causes NF/NF inflation across SSA.
  - In around a fifth of the sample of 41 countries, there is a one-way causality from food inflation to NF/NF inflation.

### Interpretation and structural drivers
- External drivers alone cannot explain all variation in domestic food inflation in SSA; structural and domestic drivers also play an important role.
- Evidence indicates food inflation has been driven by food that is largely non-tradable (i.e., fresh food).
- Greater pass-through to non-fresh food than fresh food suggests incomplete/limited tradability for perishable food items in SSA.
- Possible structural constraints behind limited tradability and higher fresh-food-driven inflation:
  - Shortage of storage capacity for perishable food.
  - Inadequate transport infrastructure.

### Policy implications and recommendations
- Monetary and fiscal policy stance:
  - Central banks and fiscal authorities in SSA and other LICs should take food prices into consideration when assessing the appropriate stance of monetary and fiscal policy.
  - Using measures of core inflation that exclude food prices to assess the monetary stance would not be appropriate in many SSA countries, especially where persistence of food prices is as high as that of NF/NF prices (currently over a third of SSA countries).

- Structural and trade policies to reduce food price pressures:
  - Improve the tradability of food by reducing tariffs.
  - Improve transport and storage infrastructure to reduce perishability-driven price pressures.
  - Implement structural measures to increase productivity in the agricultural sector to decrease food price pressures and improve welfare.

### Suggestions for further research
- Investigate the relationship between improved central bank credibility and the persistence of food inflation, given rapid modernization of central banks in SSA and other LICs.
- Explore demand and supply determinants of food inflation in LICs to support more precise policy prescriptions for SSA and other LICs.

*Source: IMF staff calculations and Box 3.3, IMF (2016).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16247.pdf_