key: cord-0945270-osdzr92z
authors: Nechifor, Victor; Ramos, Maria Priscila; Ferrari, Emanuele; Laichena, Joshua; Kihiu, Evelyne; Omanyo, Daniel; Musamali, Rodgers; Kiriga, Benson
title: Food security and welfare changes under COVID-19 in Sub-Saharan Africa: Impacts and responses in Kenya
date: 2021-03-03
journal: Glob Food Sec
DOI: 10.1016/j.gfs.2021.100514
sha: 3670b0a74ee8d931fa3315cc5f8071c707571e66
doc_id: 945270
cord_uid: osdzr92z

The COVID-19 pandemic has affected all Sub-Saharan economies through a multitude of impact channels. The study determines the medium-term macroeconomic outcomes of the pandemic on the Kenyan economy and links the results with a detailed food security and nutrition microsimulation module. It thus evaluates the effectiveness of the adopted government measures to reduce the negative outcomes on food security and to enable economic recovery at aggregate, sectoral and household levels. Through income support measures, the food sector and food demand partially recover. However, 1.3% of households still fall below calorie intake thresholds, many of which are in rural areas. Results also indicate that the state of food security in Kenya remains vulnerable to the evolution of the pandemic abroad.

compensate for this reduced spending on transportation and hospitality. In the DEMETRA model, this was implemented by recalibrating the technical shares of the individual consumer commodity groups entering the household demand functions. (5) Remittances: a reduction in remittances to households by 23% from baseline values -this reduction is in line with remittance reduction for Sub-Saharan Africa in World Bank (2020b).

Based on data from the KHIBS 2015/2016, the national average of daily DEC per capita is 1,970 kcal, which is quite similar across regions (but rural with a greater dispersion). Moreover, 46% of households at National level have a balanced diet in terms of the three macronutrients recommended goal for energy-supplying macronutrients (i.e., fats between 10-15%, proteins between 15-30% and carbohydrate between 55-75% of calories daily consumed) while 2.7% of them does not meet any of these goals. Unbalanced diet does not mean necessarily being below the minimum bound but also to be above the maximum threshold (Ramos et al., 2020).

Kenyan households (National level) spend on average more a than 50% of their budget in food and display a quite diversified diet consuming 11 of 16 food groups accounted for the HCE-DDS. However, disparities rise across regions. For instance, in rural areas households spend on average 65% of their income in food and a lower HCE-DDS (10), while in Nairobi and Mombasa (Metropolis) households spend 40% of their budgets in food and display a best diversified diet (HCE-DDS of 12) (Ramos et al., 2020) . Income tax 100 % tax relief for persons earning gross monthly income of up to KSh 24,000.

Reduction of resident Personal Income Tax Rate (Pay-As-You-Earn) top rate from 30% to 25%.

Due to data limitations on the distribution of households across tax brackets, in the current version of study the income tax was reduced by 16.6% across all household groups represented in the model. This corresponds to the effective reduction of personal income tax obtained for the top rate from 30% to 25% Turnover tax Reduction from 3% to 1% with taxable turnover thresholds increased from an income of between KSh 1 million (US$ 10,000) to KSh 50 million (US$ 500,000) for MSMSEs.

A two-thirds reduction in the model indirect tax rate.

Corporate tax Reduction in corporate tax from 30% to 25% for residents and increase; withholding tax rate on dividend payable to non-residents from 10% to 15%.

A 16.6% reduction in enterprise direct tax rate of the model. * For Nairobi and Mombasa, transfers where distributed across the lower 3 income quintiles with the following shares: 10% to quintile 3, 25% to quintile 4 and 65% to quintile 5. Outside Nairobi and Mombasa, household groups in the economic model are not disaggregated across income groups; total transfers were allocated proportionally based on the pre-COVID-19 government transfer patterns in the Social Accounting Matrix.

** Other stimulus spending such as liquidity support in the form of loans to banks and enterprises, early VAT refunds and government pending bill payments were not included in the analysis since these measures could not be captured by the modelling framework.

*** The increase in government expenditure on goods and services was implemented in the DEMETRA model by increasing the government demand for the corresponding commodities in the SAM at pre-COVID-19 prices for 2020.

-5.6% -6.8% -7.0% -7.3% -8.3%

-11.6%

-15.0%

-13.0%

-11.0%

-9.0%

-7.0%

-5.0% Note: National sample is 21625 households; where 5% is from Metropolis, 35% from peri-urban regions and 60% from rural regions. According to this sample from the KIHBS 2015/2016 2233 households has at least 1 stunted child and this is the size of the sample represented in the household category "Min HAZ <=-2". The full macro-micro commodity mapping is presented in Annex 7 of Ramos et al. (2020) .

The assessment of government measures in the article is complemented by a sensitivity analysis regarding the scale of the household cash transfers as a means to address income reduction. In the base case a total of KSh 10 billion (USD 92 million) is allocated by the government as cash payments to vulenerable households distributed across household groups as described in Table A7 .

We seek to assess the implications of both increasing and decreasing the amount of transfers.

Considering the large impacts of the base government measures on government deficit (an 89% increase in public deficit in All Measures V-V scenario and 110% in All Measures W-W), we appraise as reasonable a doubling of cash payments (a KSh 10 billion in cash transfers, representing 0.1% of the GDP) which would not put too much additional presure on the public deficit. This step up in transfers is applied as an addition to the All Measures V-V scenario (labeled as All Measures V-V Plus senstivity scenario below). We also assess the implications of a scalling down of the cash payments by reducing these by 50% from the base value (KSh 5 billion on total transfers, labeled as All Measures V-V Minus). It is assumed that in both these new sensitivity analysis scenarios no additional funds from abroad are made available, hence the government covers the corresponding increase in spending through internal borrowing.

The CGE modelling results show that, at a macro-economic level, there is a trade-off between aggregate economic output (measured in GDP, employment and supply) on one hand, and consumer demand and food-related imports on the other (Table A13 ). An increase in cash transfers (All Measures V-V Plus) leads to a decrease in output but also determines an improvement in consumer demand. Converesely, a decrease in cash transfers (All Measures V-V Minus), improves GDP outcomes but determines a reduction in consumer demand. Neverthless, the differences between the sensitivity analysis scenarios are small since the changes in cash transfers imply a redistribution of internal resources from investment (though government borrowing) to consumption. They represent a small share in the overall GDP, although have a visible impact on government deficit, and are not accompanied by an increase in inflows from abroad in the form of loans or grants. In welfare terms, the scaling up to household transfers benefits mostly the rural households, while the aggregate welfare of peri-urban and metropolitan households marginally decreases due to income effects of lower employment and investment.

For food security, the scaling up of cash transfers determines an increase in aggregate food demand which is translated into higher DEC/capita levels for the bottom and the top quintiles ( Figure A5 ), mainly driven by the increase in DEC/capita in rural areas (Table A14 ). The lower quintiles that benefit are those households with a calorie intake of below the 2250 calories/capita reference value (see Table A9 ). For the top income quintiles, the higher effective income in rural areas determines an increase in rural economic activity and consequently an improvement in food consumption of the richer rural households as well. The same trends can be observed in terms of macronutrient intake changes ( Figure A6 ) -the increase in cash transfers determines a higher intake for all three macronutrient types at the extremes of the income distribution. At an aggregate level, the 10 billion KSh increase in cash transfers determines an expansion of households with food sufficiency by 0.1 percentage points and 0.15 percentage points in metropolitan and rural areas respectively, but also a slight reduction of 0.03 percentage points in peri-urban areas.

For a decrase in cash transfers (All Measures V-V Minus), the DEC/capita and macronutrient intakes decrease for lower income quintiles households. For the top income household quintiles, food security metrics do not deteriorate as income levels benefits from lower government spending.

The sensitivity analyis thus illustrates that an increase cash payments to support the declining household income is positive for poorer households, also with positive secondary effects for top quintiles. The negative welfare impact on urban households of an increase in government spending suggest that the scaling up of these transfers should be done by seeking external support in the shape of foreign loans or grants from international donors. 

Accelerating the Pace of Interventions to Address the Impact of Covid-19 on Businesses and the Broader Economy. Kenya Private Sector Alliance

Coordinating private sector efforts and complementing GoK Efforts on the COVID-19 response. Kenya Private Sector Alliance

Flowers of Hope to Save Thousands of Farm Jobs. Kenya Private Sector Alliance

KNBS, 2020. KNBS Survey on Socio Economic Impact of COVID-19 on Households Report

Key staples food balance sheet as at end of