Methodology of Examining the Nexus between Trade Liberalization, Growth and Poverty: Some Thoughts Dr. Selim Raihan Assistant Professor Department of Economics.

Methodology of Examining the Nexus between

Trade Liberalization, Growth and Poverty:

Some Thoughts

Dr. Selim RaihanAssistant Professor

Department of EconomicsUniversity of Dhaka

Presented at the Inception Workshop on

Mainstreaming International Trade into National Development Strategy: A Pilot Project in Bangladesh and India

7-8 February, 2007, Kolkata

2

For the analysis of the relationship between trade liberalisation, growth and poverty we need to explore the mechanism of the relationship.

3

If there are absence of any structural rigidities in the economy…

4

5

6

Impacts on Growth

• Short run impact: Factors of production may be unable to move to the expanding sectors from the contracting sectors. Therefore, the loss in growth due to the contraction of the import competing sectors may be higher than the gain in growth due to the expansion of the export-oriented sectors. As a result, a negative impact on economic growth is a possibility.

• Long run impact: In the long run, if resources can be reallocated to the more efficient sectors, then there might be a positive impact on economic growth. There may also be some gains trough productivity increase.

7

Poverty Impact (Head-Count)

• Household Income

• Poverty Line income

Influenced by the changes in the prices of goods and services

8

Households’ Initial Endowments of Labour, Capital, Land, Transfer, Remittances are very important.

9

Income Composition of the Households

Percentage Contributions to the Household Income from Household Categories

Skilled labour

Unskilled labour

Non-agricultural

capital

Land Dividends Intra-household transfers

Public transfers

Remittances Total

Rural Landless 3.19 90.63 0.00 0.00 - 5.30 0.37 0.51 100.00

Marginal farmers 4.73 59.16 24.80 2.01 -. 8.38 0.35 0.57 100.00

Small farmers 17.07 37.67 24.57 15.67 - 4.26 0.10 0.66 100.00

Large farmers 9.88 5.28 34.43 49.74 - 0.41 0.01 0.24 100.00

Non-agriculture 23.01 40.45 27.79 4.79 - 2.96 0.38 0.61 100.00

Urban

Illiterate 1.69 67.41 28.79 0.00 - 1.66 0.05 0.40 100.00

Low education 7.31 41.07 41.27 6.69 - 2.94 0.26 0.45 100.00

Medium education 30.82 1.20 58.75 7.88 0.06 0.37 0.74 0.18 100.00

High education 20.08 0.26 59.72 14.95 0.20 1.14 3.43 0.21 100.00

All 16.06 35.08 35.00 10.32 0.02 2.52 0.53 0.43 100.00

Source: Household Income and Expenditure Survey 2000 for Bangladesh.Note: ‘-’ denotes not applicable to this household category.

10

If there are significant structural rigidities in the economy…

11

• Trade liberalization in this case will not necessarily lower the prices of imports.

• As resources are not sufficiently freely mobile among sectors, the loss in growth will be likely to be higher than the gain in growth.

• Impacts on factor prices will not be straightforward.

12

Different Methods

• Econometric Studies: Ex Post Analysis

• CGE Studies: Ex Ante Analysis

13

Econometric Studies

• Cross country Studies: Generalization for all countries. Fail to show the differences in the country contexts. No conclusive evidences.

• Single country studies: Use of macro time series data. Problem of time series data. No conclusive evidences.

• Studies on manufacturing sectors: No conclusive evidences.

• Theoretical problem: In most cases ad hoc model.

Raihan (2005), “Trade Liberalisation and Growth: A Dynamic Panel Econometric Study on Bangladesh Manufacturing Industries”.

15

Dependent variable: lnYit

Explanatory variables

Model includes TRLIB1

Model includes TRLIB2

Model includes TRLIB3

Model includes TRLIB4

Model includes TRLIB5

Constant 0.611 (0.44)

-0.294 (0.61)

-0.137 (0.53)

0.256 (0.52)

-0.294 (0.61)

lnYit-1 -0.347*** (0.06)

-0.337*** (0.06)

-0.335*** (0.07)

-0.337*** (0.05)

-0.347*** (0.06)

lnKit 0.142*** (0.04)

0.145*** (0.04)

0.116*** (0.03)

0.148*** (0.04)

0.142*** (0.04)

lnLit 0.337*** (0.11)

0.335*** (0.11)

0.275*** (0.08)

0.317*** (0.101)

0.337*** (0.11)

lnKit-1 0.107*** (0.03)

0.109*** (0.08)

0.088*** (0.02)

0.106*** (0.03)

0.107*** (0.03)

lnLit-1 0.209** (0.09)

0.206** (0.09)

0.193** (0.08)

0.198** (0.08)

0.209** (0.09)

lnTRLIB1it -0.374 (0.40)

- - - -

lnTRLIB1it-1 0.061 (0.14)

- - - -

lnTRLIB2it - -0.002 (0.02)

- - -

lnTRLIB2it-1 - -0.365 (0.24)

- - -

lnTRLIB3it - - -0.194*** (0.03)

- -

lnTRLIB3it-1 - - -0.137 (0.05)

- -

lnTRLIB4it - - - 0.041 (0.01)

-

lnTRLIB4it-1 - - - 0.005 (0.01)

-

TRLIB5

- - - - -0.157* (0.09)

R2 (adjusted)

0.36 0.36 0.38 0.38 0.37

Observations

567 567 567 567 567

16

CGE Studies

• Policy simulation

• Useful for ex ante analysis

• If carefully designed – theoretically consistent

17

• Most of the static CGE models in the context of Bangladesh has shown negative impact of trade liberalization on economic growth and poverty (e,g. Khondker and Raihan (2004), Raihan and Khondker (2005), Khondker, Raihan and Mujeri (2005), etc.).

• It has, however, been argued that the impacts of trade liberalisation are rather dynamic in nature and, thus the medium and long-run impacts are likely to be different from the short-run impacts.

CGE Studies (cont..)

18

• Dynamic CGE model for Bangladesh

Raihan (2005): First intertemporal dynamic CGE model for Bangladesh economy.

Annabi, Khondker, Raihan, Cockburn and Decaluwe (2006): First recursive dynamic CGE model for Bangladesh economy.

CGE Studies (cont..)

19

Some thoughts on the Nexus between Trade Liberalization and Poverty

• One of the principal mechanism through which trade liberalisation operates is the price mechanism.

• The consumers and producers are affected by the changes in the prices of the commodities they buy from and sell in the market.

• We can examine whether trade liberalisation has any effect on the changes in these prices.

• From the Household Expenditure Surveys (HES) we can calculate the unit prices of the products the consumers buy from and the sellers sell in the market.

20

• To perform the poverty analysis we can identify the bottom 10 percent of the households and the basket of commodities they consume and sell.

• We can examine the changes in the unit prices of this basket.

• We can also disaggregate the households between moderate and extreme poor, between rural and urban households, between agricultural and non-agricultural households, etc.

Some thoughts (cont..)

21

It is very plausible that the changes in the prices of the commodities are influenced by not only the trade policy but also by a number of governmental policy and structural variables. In order to isolate the actual effect of trade liberalisation on the changes in unit prices we need to perform some type of decomposition analysis.

Some thoughts (cont..)

22

• Trade liberalisation also affects poverty through its impact on government revenue. If liberalisation of trade is accompanied by a decline in government tariff revenue then it may reduce the level of government expenditure for health, education and other poverty alleviation programmes.

• It is, therefore, important to examine the trend and pattern of government revenue and expenditure after the liberalisation of trade.

• However, the pattern of government expenditure is affected by not only the availability of funds but also some other factors. It is thus required to isolate the actual effect of trade liberalisation on the level and pattern of government expenditure. A type of decomposition analysis is required in this regard.

Some thoughts (cont..)

23

• In Bangladesh and India most of the poor rely on labour markets for the bulk of their income. Thus the effects of trade liberalisation on wages and employment are important, especially those of unskilled workers.

• If reform boosts the demand for labour-intensive products, it boosts the demand for labour and either (or both) of wages or employment will increase.

• However, if the poor are mostly in completely unskilled families, while it is semi-skilled labour that receives the boost, poverty will be unaffected – or, possibly, worsened.

Some thoughts (cont..)

24

• It is also important where the various wage rates lie relative to the poverty line. If wages are pushed up from subsistence to higher levels, or if the expanding sectors offer above poverty-line wages, then poverty will be alleviated.

• On the other hand, if poverty is measured by counting individuals below the poverty line - the headcount index - and wages do not cross critical thresholds, recorded poverty could be unaffected, despite changes in welfare.

• The effects of trade liberalisation on wages and employment can be examined by estimating the labour demand function using the national and sectoral data.

Some thoughts (cont..)

25

• Linking WTO issues and RTAs with poverty:

– Agricultural liberalization and poverty: Raihan et al (2007), Sahay (2006)

– DFQF market access for LDCs: Raihan et al (2007).

– NAMA and poverty: Raihan et al (2007).

– MFA phase out: Raihan and Khondker (2005)

– SAFTA: Raihan et al (2007)

Some thoughts (cont..)