## Methods used for economic data analysis

## Empirical consequences

This chapter will concentrate on the consequences of the information analysis. The first subdivision will discourse the descriptive statistics and in the 2nd subdivision the consequences of the Heckman two-step attack will be discussed.

## Descriptive statistics

The descriptive statistics of the study informations will be discussed by comparing and qualifying the families that affected and do non affected by the clime alteration. The sample size used for the analysis is hence 420 respondents.

An uneven distribution of land ownership exists in coastal part of Bangladesh, with a important proportion of land being owned by big landholders ( Alauddin and Hamid 1997 ) . Agricultural study ( 1996 ) shows that 54 % of households in coastal countries hold merely 17 % of the entire agricultural land ( PDO-ICZMP-2003 ) . The bulk of the rural population is either landless husbandmans ( who sell their labour or cultivate other ‘s land ) or fringy husbandmans ( who have less than.2 hour angle of belongings ) ( Opstal 2006 ) . Over the past decennary the husbandmans are declined. Now a twenty-four hours in the coastal Bangladesh fishing is one of the most of import economic activities. They are largely landless or have a little secret plan of land to utilize for life intent.

In the survey country entire land size is changed due to climate alteration. It appears from the given tabular array which shows the comparative analysis of land form before and after Aila. In 2008 the mean sum is 157.02 hectare/year and in 2009 it is 99.89 hectare/year. Land is used for different intents. In 2008, 159 respondents used their land for cultivation i.e they are the agricultural land proprietor and due to climate alteration merely 75 respondents are the proprietor of the agricultural land. This sum is diminishing.

Comparison of entire land usage in 2008 and 2009:

In last 5 old ages 62 families lost their land in the survey country. The entire sum of damaged land is 36911.58 hectares. Most of the people depend on agribusiness so this is a great loss for their survive. For this their income is decreased, outgo is decreased and they have no adequate money to purchase the agricultural land. From this it is concluded that they live below poorness line. Harmonizing to a recent ( Oct’09 ) survey done by the South Asia Association of Poverty Eradication, each affected family has seen their income lessening by about 44 % as a consequence of Cyclone Aila.

The chief independent variable is expenditures by family for a basket of basic demands, which is considered as a measuring of ‘poverty. ‘ This outgo measuring really represents a poorness threshold value, which is derived from HIES ( Household Income-Expenditure Survey 2009 ) by BBS and is tantamount to US $ 208/capita/year ( BBS, 2008 ) . It is referred as ‘Basic Need Cost ‘ in the theoretical account.

In 2009 we get merely 84 respondents out of 420 bashs non populate below poorness line. It is estimated by utilizing our outgo informations from primary study analysis. So due to climate alteration most of the families live below poorness line.

## Econometric Analysis

Now we would wish to go on with calculating out the nature and extent of relationship between agricultural land ownership form and poorness of Koyra. Hence, in this chapter we conduct econometric analysis.

## Variables used in econometric theoretical accounts

With a position to placing the relationship form between agricultural land ownership form and poorness we ran a figure of econometric theoretical accounts. But before we proceed to the operation with econometric theoretical accounts, allow us hold a expression at the variables used in the theoretical account.

## Dependent variable

The dependant variable is entire land owned by, which is considered to be affected by clime alteration. This variable indicates how much land was owned by the family in 2009. The values were taken in hectares for the full family.

## Independent variables

Below we have mentioned the independent variables, with short account, that we used in theoretical accounts. Variable ‘household size ‘ refers to the entire figure of members in a family. ‘Education ‘ refers to household ‘s mean aggregative academic schooling twelvemonth. It is the figure obtained by summing up of formal schooling old ages of all members in a family and so spliting it with the figure of entire family members. This variable is considered as a placeholder for capacity of families. The variable ‘Duration with community ‘ refers to the figure of old ages the respondent family life with the current community.

Along with the above-named dependant and independent variables, we used the undermentioned two independent variables for building correlativity and arrested development.

## Econometric Methodology:

We used a Heckman Two Step Model for dependent variable ‘land ownership ‘ in order to happen out if there is any sample choice prejudice in the theoretical account. This theoretical account consists of two procedures that are addressed by two different equations: a choice equation and a conditional equation. The first probit equation is a choice procedure for the families holding land-ownership or non. In the 2nd equation the effects of independent variables on ‘land ownership ‘ are examined.

These procedures are related to each other through their mistake footings which contain the unobservable. If there is no correlativity between the mistake footings of the two equations, there is no demand to execute a Heckman two measure attack as there is no sample choice prejudice and an OLS arrested development provides the indifferent consequence ( Dow and Norton, 2003 ) .

The Heckman two-step attack is based on the premise that the choice equation and the conditional equation are related to each other through their mistake footings. When there is no relation between the mistake footings there is no demand to execute a Heckman two measure attack as there is no sample choice prejudice and an OLS arrested development will give indifferent calculators. For such a theoretical account, the bottom line in STATA end product gives a value for ? ( rho ) with associated p-value. This ? is a likeliness ratio bespeaking the correlativity between the mistake footings of the equations in Heckman theoretical account.

The correlativity between the mistake footings is indicated in tabular array ( Annex ) by the selectivity parametric quantity, ? . The Heckman ‘s lambda is included in the arrested development to command for the influence of unseen features of the variables. The arrested development coefficient of the control factor is an index for the covariance of the mistake footings. In the theoretical account the control factor is non-significant.

The losing information job can originate in a assortment of signifiers. We can see that there are losing informations in the sample. The figure of losing informations in is 3, but the job is more terrible for, where the figure of losing informations is 80. Since the information is losing chiefly on the dependant variable, a nonrandom sample choice exists in this instance. There is a possibility that due to some common form, the respondents did non supply any informations. If that has happened, prejudices could ever happen in OLS in gauging the population theoretical account. As a consequence, we use here the Heckman theoretical account.

Our theoretical account is

## Empirical consequences

This chapter will concentrate on the consequences of the information analysis. The first subdivision will discourse the descriptive statistics and in the 2nd subdivision the consequences of the Heckman two-step attack will be discussed.

## Descriptive statistics

The descriptive statistics of the study informations will be discussed by comparing and qualifying the families that affected and do non affected by the clime alteration. The sample size used for the analysis is hence 420 respondents.

An uneven distribution of land ownership exists in coastal part of Bangladesh, with a important proportion of land being owned by big landholders ( Alauddin and Hamid 1997 ) . Agricultural study ( 1996 ) shows that 54 % of households in coastal countries hold merely 17 % of the entire agricultural land ( PDO-ICZMP-2003 ) . The bulk of the rural population is either landless husbandmans ( who sell their labour or cultivate other ‘s land ) or fringy husbandmans ( who have less than.2 hour angle of belongings ) ( Opstal 2006 ) . Over the past decennary the husbandmans are declined. Now a twenty-four hours in the coastal Bangladesh fishing is one of the most of import economic activities. They are largely landless or have a little secret plan of land to utilize for life intent.

In the survey country entire land size is changed due to climate alteration. It appears from the given tabular array which shows the comparative analysis of land form before and after Aila. In 2008 the mean sum is 157.02 hectare/year and in 2009 it is 99.89 hectare/year. Land is used for different intents. In 2008, 159 respondents used their land for cultivation i.e they are the agricultural land proprietor and due to climate alteration merely 75 respondents are the proprietor of the agricultural land. This sum is diminishing.

Comparison of entire land usage in 2008 and 2009:

## Measurements

## entire land usage in 2008 ( hectare/year )

## entire land usage in 2009 ( hectare/year )

Average

157.02

99.89

Median

24

15

Manner

5

5

Std. Deviation

313.35

219.44

Minimum

0

1

Maximum

2409

1320

Sum Used Lands by the respondents

65946

33864Source: Field Survey.

In last 5 old ages 62 families lost their land in the survey country. The entire sum of damaged land is 36911.58 hectares. Most of the people depend on agribusiness so this is a great loss for their survive. For this their income is decreased, outgo is decreased and they have no adequate money to purchase the agricultural land. From this it is concluded that they live below poorness line. Harmonizing to a recent ( Oct’09 ) survey done by the South Asia Association of Poverty Eradication, each affected family has seen their income lessening by about 44 % as a consequence of Cyclone Aila.

The chief independent variable is expenditures by family for a basket of basic demands, which is considered as a measuring of ‘poverty. ‘ This outgo measuring really represents a poorness threshold value, which is derived from HIES ( Household Income-Expenditure Survey 2009 ) by BBS and is tantamount to US $ 208/capita/year ( BBS, 2008 ) . It is referred as ‘Basic Need Cost ‘ in the theoretical account.

In 2009 we get merely 84 respondents out of 420 bashs non populate below poorness line. It is estimated by utilizing our outgo informations from primary study analysis. So due to climate alteration most of the families live below poorness line.

## Econometric Analysis

Now we would wish to go on with calculating out the nature and extent of relationship between agricultural land ownership form and poorness of Koyra. Hence, in this chapter we conduct econometric analysis.

## Variables used in econometric theoretical accounts

With a position to placing the relationship form between agricultural land ownership form and poorness we ran a figure of econometric theoretical accounts. But before we proceed to the operation with econometric theoretical accounts, allow us hold a expression at the variables used in the theoretical account.

## Dependent variable

The dependant variable is entire land owned by, which is considered to be affected by clime alteration. This variable indicates how much land was owned by the family in 2009. The values were taken in hectares for the full family.

## Independent variables

Below we have mentioned the independent variables, with short account, that we used in theoretical accounts. Variable ‘household size ‘ refers to the entire figure of members in a family. ‘Education ‘ refers to household ‘s mean aggregative academic schooling twelvemonth. It is the figure obtained by summing up of formal schooling old ages of all members in a family and so spliting it with the figure of entire family members. This variable is considered as a placeholder for capacity of families. The variable ‘Duration with community ‘ refers to the figure of old ages the respondent family life with the current community.

Along with the above-named dependant and independent variables, we used the undermentioned two independent variables for building correlativity and arrested development.

## Econometric Methodology:

We used a Heckman Two Step Model for dependent variable ‘land ownership ‘ in order to happen out if there is any sample choice prejudice in the theoretical account. This theoretical account consists of two procedures that are addressed by two different equations: a choice equation and a conditional equation. The first probit equation is a choice procedure for the families holding land-ownership or non. In the 2nd equation the effects of independent variables on ‘land ownership ‘ are examined.

These procedures are related to each other through their mistake footings which contain the unobservable. If there is no correlativity between the mistake footings of the two equations, there is no demand to execute a Heckman two measure attack as there is no sample choice prejudice and an OLS arrested development provides the indifferent consequence ( Dow and Norton, 2003 ) .

The Heckman two-step attack is based on the premise that the choice equation and the conditional equation are related to each other through their mistake footings. When there is no relation between the mistake footings there is no demand to execute a Heckman two measure attack as there is no sample choice prejudice and an OLS arrested development will give indifferent calculators. For such a theoretical account, the bottom line in STATA end product gives a value for ? ( rho ) with associated p-value. This ? is a likeliness ratio bespeaking the correlativity between the mistake footings of the equations in Heckman theoretical account.

The correlativity between the mistake footings is indicated in tabular array ( Annex ) by the selectivity parametric quantity, ? . The Heckman ‘s lambda is included in the arrested development to command for the influence of unseen features of the variables. The arrested development coefficient of the control factor is an index for the covariance of the mistake footings. In the theoretical account the control factor is non-significant.

The losing information job can originate in a assortment of signifiers. We can see that there are losing informations in the sample. The figure of losing informations in is 3, but the job is more terrible for, where the figure of losing informations is 80. Since the information is losing chiefly on the dependant variable, a nonrandom sample choice exists in this instance. There is a possibility that due to some common form, the respondents did non supply any informations. If that has happened, prejudices could ever happen in OLS in gauging the population theoretical account. As a consequence, we use here the Heckman theoretical account.

Our theoretical account is

We assumed that is observed if

Where and hold correlativity

## Consequences:

The consequences of our Heckman theoretical account are provided in Table ( Annex ) . Using as a dependant variable in Heckman arrested development, we find and the changeless term are important while is undistinguished. We besides find positive relationship for and with. Sing the absolute values of the coefficients ( tabular array ) , the consequence shows that is the most influential between the two variables.

A typical usage of a logarithmic transmutation variable is to draw outlying informations from a positively skewed distribution closer to the majority of the informations in a pursuit to hold the variable be usually distributed. In arrested development analysis the logs of variables are routinely taken, non needfully for accomplishing a normal distribution of the forecasters and/or the dependant variable but for interpretability.

The standard reading of coefficients in a arrested development analysis is that a one unit alteration in the independent variable consequences in the several arrested development coefficient alteration in the expected value of the dependent variable while all the forecasters are held changeless. Interpreting a log transformed variable can be done in such a mode ; nevertheless, such coefficients are routinely interpreted in footings of per centum alteration ( Introductory Econometricss: A Modern Approach by Woolridge for treatment and derivation ) .

We ‘ll research the relationship between the landownership form and the per capita ingestion outgo. In this theoretical account we are traveling to hold the dependant variable in its original metric and the independent variable log-transformed. Similar to the anterior illustration the reading has a nice format, a one per centum addition in the independent variable additions ( or decreases ) the dependant variable by ( coefficient/100 ) units. In this peculiar theoretical account we take log with PCE and the coefficients on and stand for the estimated fringy effects of the regressors in the implicit in arrested development equation. So, an addition in the family size by one member additions land ownership by 6.30 hectares and an addition in the family ingestion outgo by one per centum additions land ownership by 0.613 hectares.

On the other manus, family size is the least influential variable. It is positively related with landownership form. So these two variables have greater influence on poorness. We used the Heckman two measure theoretical accounts while taking land ownership as a dependant variable in the conditional equation of this theoretical account, along with other independent variables, consequence in theoretical account shows that PCE is positively related with landownership.

The p value of lambda is 0.193 i.e. 19 % . So this is non important for the theoretical account i. e. there is no correlativity between the mistake footings of the two equations in Heckman theoretical account. The lambda term is positively signed – which suggests that the mistake footings in the choice and primary equations are positively correlated. So ( unseen ) factors that make more discernible tend to be associated with higher values of our independent variables in the choice equation. However, since the lambda term is non important, we can non come to any such decision and hence we conducted OLS.

But if we use the OLS we get the followers

Table 1: OLS Consequence

## — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —

lnd_owners~p | Coef. Std. Err. T P & A ; gt ; |t| [ 95 % Conf. Interval ]

## — — — — — — -+ — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —

lnpce | 58.21023 18.98437 3.07 0.002 20.86622 95.55423

hh_size | 4.660069 6.495749 0.72 0.474 -8.117666 17.4378

_cons | -204.742 97.52465 -2.10 0.037 -396.5819 -12.90203

## — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —

We present the usual OLS arrested development in Table 1. As we can see from Table 1, and is both positive, while the former is non important and the latter is important. Similarly, the changeless term is negative but important.

Table 2

From the above OLS tabular array we consider the independent variables are per capita outgo, instruction degree, during with the community, family size and plus 2008 and the dependant variable is land ownership form of the respondents. In this analysis the theoretical account is important in instance of plus 2008 for dependant variable land ownership because in this instance the value of P is 0 % . We know if the value of P is less than 5 % so the theoretical account is important. From the arrested development we get per capita outgo, instruction degree, during with the community and plus 2008 is positive. But without plus 2008 all other variables are non important. Similarly the changeless term is besides positive but non important.

Consequences from assorted OLS arrested development theoretical accounts are shown in Table 1 and.2. The former shows consequences when theoretical account is run with and while the latter shows consequences when land ownership is incorporated with other independent variables. Valuess of coefficient are different for the independent variables in the consequence tabular arraies. Using land ownership ( i.e. our step of poorness ) as a dependant variable in OLS arrested development, we found without one, all the explanatory variables are non important ( Table 2 ) . We besides found important positive relationship per capita outgo, instruction degree, during with the community and plus 2008 with land ownership whereas it is significantly negative for family size.

Annex

. heckman lnd_ownership lnpce hh_size, twostep select ( lnpce edulevel duringwithcomty hh_size asst2008 ) rhosigma

Heckman choice theoretical account — two-step estimations Number of obs = 417

( arrested development theoretical account with sample choice ) Censored obs = 80

Uncensored Obs = 337

Wald chi2 ( 4 ) = 9.83

Prob & A ; gt ; chi2 = 0.0434

## — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —

| Coef. Std. Err. omega P & A ; gt ; |z| [ 95 % Conf. Interval ]

## — — — — — — -+ — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —

lnd_owners~p |

lnpce | 61.28878 20.67387 2.96 0.003 20.76873 101.8088

hh_size | 6.303549 7.203314 0.88 0.382 -7.814687 20.42179

_cons | -286.9731 123.3481 -2.33 0.020 -528.731 -45.21517

## — — — — — — -+ — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —

select |

lnpce | .0682579.1348031 0.51 0.613 -.1959514.3324671

edulevel | .0096151.025462 0.38 0.706 -.0402896.0595197

duringwith~y | .0161874.005286 3.06 0.002.005827.0265477

hh_size | .007615.046654 0.16 0.870 -.0838252.0990552

asst2008 | -1.13e-06 7.34e-07 -1.53 0.125 -2.57e-06 3.12e-07

_cons | -.0686488.6543009 -0.10 0.916 -1.351055 1.213757

## — — — — — — -+ — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —

Millss |

lambda | 181.4302 139.4798 1.30 0.193 -91.94525 454.8057

## — — — — — — -+ — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —

rho | 0.74328

sigma | 244.09453

lambda | 181.43021 139.4798