Dear Tamar and Tamila,

We’d like to follow up on our meeting yesterday and share preliminary findings of some analytical work we have been doing in the past couple of days to help predict the impact of the crisis on labor markets and social benefits. The analysis is based on nationally representative surveys collected by GEOSTAT (HIES 2018 and LFS 2018), which represents the full country (as opposed to the social registry which represents only a third of the country, about 300,000 households as we understand).

We hope you will find the analysis a useful complement to the modeling you are working on. We are available any time to answer any clarification questions you may have and/or adapt the analysis based on your specific requests (on assumptions and parameters). Please take this as WB technical assistance to design response measures and predict financing needs, regardless of whether response policies will be funded by the State, the WB, other donors.

Few key messages:

1. TSA program expansion. As discussed yesterday, we expect that the TSA (the existing program) coverage will naturally expand. We understand that the regular application process is open and application requirements have been relaxed during the lockdown, in alignment with good practice in many countries. The revenue index (in particular the cash flows for the second month before application) is the only shock-responsive element of the current PMT scoring formula. Which means that even without changing  anything in the current design (keeping the eligibility threshold for TSA at 65,000), some of the households who were not eligible until March (and for instance did not even register/apply) will be eligible if they apply in May for example, under the assumption that many wage workers will lose their job and will declare zero income taxes (hence their revenue index will lower and their score too).

We had run simulations under alternative scenarios in the LM (attached file). For example, if we assume that 20% of wage workers (in the overall population) will lose jobs in the next 2 months, 10,183 new households would be eligible to TSA (with score <65,000). If we assume that not all of them will apply (because of lack of information, mobility issues…) but only 80% of them will apply, then about 8,146 new households would apply and qualify under the TSA (with score below 65,000), implying an additional cost of 8.8 million GEL. If in addition to the 20% workers layoffs we assume that those wage workers who keep the job will see they salaries reduced by 20 percent, for example, the additional monthly cost of TSA will be 9.2 million GEL. Table 1 in the attached file provides alternative scenarios.

2. Temporary benefits for the poor – using the current PMT scoring formula to select beneficiaries. By replicating the PMT score in the HIES 2018, we estimate that about 225,231 households in the entire population have a score between 65,000 and 100,000 in normal times (as of 2018 data).  94,946 of them have children with age <16.  Of these 94,946 households, 38,594 have no wage worker in the household (so would not be entitled to the new temporary unemployment benefits either). 94,053 would be households with children and at least one work-able member in working age.  If we assume that only 80% of them would apply to the emergency temporary benefit, the cost for one-off benefit (of an amount equal to 200 GEL) would be would be about 15 million GEL, for a temporary 4 months payment would be 60 million GEL (this assuming that the 200 GEL are in addition to the child benefit some households are already receiving).

We also looked at the characteristics of households with score between 65,000 and 100,000 and we see that even households that are without children below 16 are more disadvantaged in terms of probability of informal work and per capita consumption: for instance 77% of households without children in that score interval do not have any source of wage income as opposed to 41% of households with children in the same score interval. The monthly consumption per equivalent adult is 136GEL in households without children versus 151GEL in household with children (Table 3 in the attached).  

3. Temporary benefits for informal workers – using income tax database and public benefit receipt to select beneficiaries. A quicker way to target informal workers (as the computation and verification of the PMT for new households takes several days), could be based on simple criteria easy to verify. The emergency benefit would still be on-demand, targeted to households with no members in formal employment since March 15, 2020 (for example, TBC) which you can verify in the income tax database, who declare to have no source of income, not receiving TSA, not receiving pensions and other form of social assistance. Audits and sanctions could be announced for those who will be identified as not declaring the true.  Based on preliminary estimates on LFS 2018 (where formal wage workers can be more precisely estimated as those paying taxes and having sick leave and annual leave benefits), about 260,000 households would be eligible (i.e. these are households with no formal wage worker, not receiving TSA, pension or other form of assistance) and about half of them are with children (135,869, table 7 in the attached). If we assume that only 80% of them would apply to the emergency temporary benefit, the cost for one-off benefit (of an amount equal to 200 GEL) would be about 52 million GEL. If we restrict to households with children only and assume that only 80% of them would apply, the cost for one-off benefit (of an amount equal to 200 GEL) would be about 27 million GEL.

Warm regards,

Maddalena, on behalf of the team