Evidence-Based Revenue Forecasting FY 2021-22

Evidence-Based Revenue Forecasting FY 2021-221

(FBR Taxes)

June 2021

Strategic Planning Reforms & Statistics Wing

1 Revenue Forecasting Report has been prepared in consultation with team of experts from Fiscal Policy Center, Pakistan Institute Development Economics (PIDE), Islamabad

List of Acronyms

Contents

1. FBR Collection Profile

Major part of government tax revenues is collected by FBR. A substantial increase in the tax collections has been witnessed during last two decades. FBR tax collection was just around Rs.392 billion in 2000-01 which has jumped to around Rs.4 trillion in 2019-20. The average growth till FY 2017-18 remained 14.1%, however, due to Covid-19 pandemic the yearly growth has plummeted, thus affecting slightly the average annual growth for the period 2001-2020 as well (13.9%).

1.1 Five Years FBR Collection Trend

The last five years’ collection trend is shown in the following table. The 5 years average growth was 9.3%, however, by excluding the last two years, average growth for normal years i.e. FY 2013-14 to FY 2017-18 stands at 14.6% (Table 1).

Table 1: FBR Collection Trend

(*) for the Years 2013-14 to 2017-18

Tax-GDP Ratio

Similarly, in the tax-GDP ratio, similar trend has been noticed and it kept on rising till FY 2017-18 and reached 11.2 (Graph 1), but later on this healthy trend couldn’t continue during FY 2018-19 and FY 2019-20, which is mainly attributed to Covid-19 pandemic related economic challenges.

<!-- Start of picture text --> Graph 1: FBR: Tax-GDP Ratio
12.0 11.1
10.7 10.6
10.1
10.0 9.4 9.58
9.0
8.0
6.0
4.0
2.0
0.0
2013-14 2014-15 2015-16 2016-17 2017-18 2018-19 2019-20
<!-- End of picture text -->

1.2 CFY Collection: July-March 2020-21

After a slow growth in previous two years now the collection is picking up, which is evident from table 2. This performance is very encouraging as compared to previous couple of years. The nine months target has been achieved to the extent 103.4%. Table 2: CFY Collection: July-March 2020-21

Rs. Million

With the recovery of economy the revenue collection is also improving, and with further improvements in coming months hope the tax collection shall also increase, thus enabling FBR to achieve its revenue FY 2021-22 target.

2. Evidence-Based Revenue Forecasting (EBRF) FY 2021-22

FBR collects Direct Taxes (DT), Sales Tax (Domestic & Imports), Federal Excise Duty (FED) and Customs Duties (CD). In this study tax-wise buoyancy estimates have been used to forecast the FBR head-wise revenues for FY 2021-22. The tax buoyancy is an indicator to measure efficiency and responsiveness of revenue mobilization in response to growth in the GDP or national income. A buoyant tax means the tax revenues increase more than proportionately in response to a rise in national income/GDP/base.

2.1 Methodology

The SPRS Wing before budget, projects the head-wise revenue estimates based on buoyancy estimates² from previous years on rolling basis. Tax-wise buoyancy estimates are calculated by using historical collection and respective bases data. The current tax-wise buoyancy estimates are estimated from data for the years 1999-2000 to 2019-20 in respect of economic indicators i.e. GDP, Non-agri GDP, LSM and Imports.³

The respective proxy bases of head-wise FBR taxes are as under:

2 Buoyancy is relationship between changes in actual revenues and changes in proxy tax base from National Accounts data.

3 Annex-I and Annex II

Since these tax bases have a dynamic relationship with GDP, in the second step we have calculated the buoyancies of these tax bases with GDP =.

Finally these two are multiplied to arrive at respective tax to GDP buoyancies=

Using these buoyancy estimates autonomous growth for each tax is estimated by using the projections provided by Finance Division for each proxy base.⁴ It provides for the GDP Growth and inflation forecast for upcoming year. GDP value used for this projection of growth is Gross Value Addition of Sectors at Constant Factor Costs. Using the inflation forecasts these GDP growth estimates are converted into GDP growth estimates at Current Factor Cost. Further, GDP-Gross value addition at Current Factor Cost is used for measuring the Buoyancy estimates to be used for revenue forecasts.

In respect of Direct Taxes (DT) Non-Agri GDP has been taken as its proxy base. The buoyancy of direct taxes has been estimated in two steps: in the first step Direct Tax to Non-Agri GDP and in second step Non-Agri-GDP to GDP (Gross Value AdditionCurrent Factor Cost) is estimated. Then two buoyancy estimates have been multiplied to estimate the Direct Taxes to GDP Buoyancy estimates.

Direct taxes to Non-Agri GDP Buoyancy (B1) =

4 Finance Division provides the macroeconomic framework for the next fiscal year along with rolling estimates for next two years in their Medium Term Budget Strategy Paper (2021-22 to 2023-24).

Non-Agri GDP to GDP Buoyancy (B2) =

% Change in Non-Agri GDP % Change in GDP

Direct taxes to GDP Buoyancy = B1 * B2

Similarly for Customs (CD) the proxy base is dutiable imports, whereas the projections as provided by Finance Division are for the total imports. Therefore, the buoyancy has been estimated in two steps: first Customs to dutiable imports and in second step Dutiable Imports to total Imports and then these two have been multiplied to get the CD buoyancy estimates.

For Federal Excise (FED) and Sales Tax (Domestic) the base is Large Scale Manufacturing (LSM) and for Sales Tax (Imports) the base is imports. Therefore, for these taxes the buoyancy values are directly estimated from proxy bases. Following table reflects the respective buoyancies which are estimated using simple log-log regression method in excel sheet by using log values of actual data for taxes.⁵

Table 3: Buoyancy Results

5 Source: FBR Annual Reports and Economic Survey of Pakistan

Calculating Autonomous Growth

In the second step autonomous growth has been estimated by multiplying buoyancy estimates (Table 3) to projected growths of respective bases (Table 4):⁶

Table 4: Growth Assumptions for FY 2021-22

Table 5 provides an autonomous growth parameters for each tax head. These are based on the buoyancy estimates from the last 20 years actual revenue collection data and the macroeconomic proxy bases provided by Pakistan Bureau of Statistics/Annual Economic Surveys and the growth projections published in Budget Strategy Paper / Economic Adviser’s Wing, Finance Division.

Table 5: Autonomous Growth

6 Autonomous growth depends on macroeconomic indicators and assumption of tax structure being able to capture revenue from these as per past performance. A higher growth in base will lead towards higher independent growth in respective taxes and thus higher projected collection.

7 Source Medium Term Budget Strategy Paper 2021-22 to 2023-24 / Economic Adviser’s Wing, Finance Division

These autonomous growths will be used to project the autonomous growth estimates of respective revenue heads.

To be sure of the results robustness a small exercise was also done by using the reduced and increased nominal growths by 10%. The resulting change was insignificant.

2.2 Head-wise Revenue Projections FY 2021-22

The autonomous growth (Table 5) has been applied on base year’s expected collection (i.e. 2020-21) for each respective head to project an increase of Rs. 636 billion . This addition has been added in the expected collection of FY 2020-21 thus, the revenue forecast for FY 2021-22 without additional Policy/Admn measures has been obtained to the tune of Rs. 5,336 billion. The required growth over the expected collection of FY 2020-21 i.e. Rs. 4,700 billion would be 13.5% in FY 2021-22. However, with the addition of Policy/Admn measures the target for FY 2021-22 would be Rs. 5,829 billion , which is near to IMF suggestion i.e. requiring a growth of about 24.0%.

Table 6: Revenue Projections for FY 2021-22 (Rs. Billion)

(*) This excludes Policy/Admn Measures and consists of projection due to autonomous growth

Concluding Remarks

It has been found that the overall FBR taxes are buoyant (1.04 overall buoyancy) and there is a potential for increase in tax revenues provided that macroeconomic indicators are doing well. It is evident from last 20 years data that FBR revenues increased substantially with around 14% average growth. The head-wise break-up reveals that the direct taxes are most buoyant with I.13 buoyancy value, followed by sales tax (domestic) with 1.11 buoyancy estimates. On the other hand, sales tax (imports), customs and FED have relatively lesser buoyancies. In this regard, addressing the issues of narrow base, unnecessary exemptions and valuation problems at import stage can be instrumental for making these tax heads more buoyant, thus enabling the revenue organization to fetch more tax revenues.

The revenue forecast for FY 2021-22 by applying the buoyancy estimates and projected respective macroeconomic indicators would be Rs.5,336 billion without Policy/Admn measures. After adding Policy/Admn measures the target would be Rs.5,829 billion. However the revenue collection and achieving of target would largely depend on the performance of the economy against the targets.

Annex-I

Net Tax collection 2000-01 to 2019-20 (Rs. Million)

Annex-II

Tax Bases Data (Rs. Million)

Annex-III Log Values

Buoyancy Estimates based on 20 years data