Introduction

Exploratory Spatial Data Analysis

Basic spatial econometrics
Advanced spatial econometrics
Summary

Spatial Econometrics Methods using Stata

1
Marcos Herrera

1 CONICET-IELDE
National University of Salta (Argentina)

Luxembourg Institute of Socio-Economic Research

Belval, 15th February 2017

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis
Basic spatial econometrics
Advanced spatial econometrics
Summary

General index
1 Introduction

2 Exploratory Spatial Data Analysis

Loading data and choropleth maps
Matrices and spatial tests
Spatial local analysis

3 Basic spatial econometrics

Taxonomy of spatial models
Spatial detection using OLS residual
Methods of estimation

4 Advanced spatial econometrics

Basic panel data models
Static spatial panel models
Dynamic spatial panel models

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis
Basic spatial econometrics
Advanced spatial econometrics
Summary

What is Spatial Econometrics?

Denitions of Spatial econometrics:

Jean Paelinck introduced the term Spatial Econometrics in

1974 to designate: a combination of economic theory,
mathematical formalization and statistics with: role of spatial
interdependence, importance of factors in other places, explicit
modelling of space".

Luc Anselin (1988) dened the Spatial Econometrics as:

econometric branch dealing with spatial interaction and
spatial structure in cross-sectional models and data panel
(separating spatial dependence and spatial heterogeneity).

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis
Basic spatial econometrics
Advanced spatial econometrics
Summary

Why is Spatial Econometrics important nowadays?

Theory-driven
From individual decision to social-spatial interaction.
Common shocks.
Peer-eects, contextual eects, neighbourhood eects.

Data-driven
Geo-referenced information.

Technology
Geographical Information Systems.
Capability of statistical software.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis
Basic spatial econometrics
Advanced spatial econometrics
Summary

Types of spatial data and Objective of workshop

Types of spatial data:

Geostatistic data: continuous spatial eld (noise surface,
pollution surface).
Aerial (Lattice or Regional): discrete spatial data, xed
polygons or points (counties, provinces, countries).
Point pattern: location as a random event (crimes, accidents).

This workshop proposes a data-driven analysis using lattice data:

Specifying the structure of spatial dependence.

Testing for the presence of spatial eects.

Estimating and interpreting models with spatial eects.

M. Herrera Spatial econometrics using Stata

Example used: Impact of net migration on unemployment
Hot topic in economics.

Competitive theories:
Orthodox theory: net migration causes more unemployment
(positive relationship).
New Economic Geography theory: net migration causes less
unemployment (negative relationship).
Denition of variables:
UNEMPLOYMENT RATE as the number of people
unemployed as a percentage of the labour force.
NET MIGRATION RATE as the ratio of net migration during
the year to the average population in that year. The value is
expressed per 1,000 persons. Net migration is the dierence
between immigration into and emigration from the area during
the year (net migration is therefore negative when the number
of emigrants exceeds the number of immigrants).
Level of analysis:
NUTS 2 (Europe 15), 164 regions from 2007 to 2012.
 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

General Index
1 Introduction

2 Exploratory Spatial Data Analysis

Loading data and choropleth maps
Matrices and spatial tests
Spatial local analysis

3 Basic spatial econometrics

Taxonomy of spatial models
Spatial detection using OLS residual
Methods of estimation

4 Advanced spatial econometrics

Basic panel data models
Static spatial panel models
Dynamic spatial panel models

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

From shape to dta

First we need a le of administrative areas:
http://www.diva-gis.org/, http://www.gadm.org/.

Georeferencing information (lattice data) usually is stored in a

shapele (a collection of les with a common lename with at
least three connected les):

.shp is the le that store geometric objects.

.shx is an index le of geometric objects.
.dbf is the database le, in dBASE format, and
contains information of attributes of the objects.

Shape les cannot be read directly in Stata.

However, shp2dta command can import shapeles and convert

them in Stata format.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

From shape to dta

First we need a le of administrative areas:
http://www.diva-gis.org/, http://www.gadm.org/.

Georeferencing information (lattice data) usually is stored in a

shapele (a collection of les with a common lename with at
least three connected les):

.shp is the le that store geometric objects.

.shx is an index le of geometric objects.
.dbf is the database le, in dBASE format, and
contains information of attributes of the objects.

Shape les cannot be read directly in Stata.

However, shp2dta command can import shapeles and convert

them in Stata format.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

From shape to dta

First we need a le of administrative areas:
http://www.diva-gis.org/, http://www.gadm.org/.

Georeferencing information (lattice data) usually is stored in a

shapele (a collection of les with a common lename with at
least three connected les):

.shp is the le that store geometric objects.

.shx is an index le of geometric objects.
.dbf is the database le, in dBASE format, and
contains information of attributes of the objects.

Shape les cannot be read directly in Stata.

However, shp2dta command can import shapeles and convert

them in Stata format.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

From shape to dta

First we need a le of administrative areas:
http://www.diva-gis.org/, http://www.gadm.org/.

Georeferencing information (lattice data) usually is stored in a

shapele (a collection of les with a common lename with at
least three connected les):

.shp is the le that store geometric objects.

.shx is an index le of geometric objects.
.dbf is the database le, in dBASE format, and
contains information of attributes of the objects.

Shape les cannot be read directly in Stata.

However, shp2dta command can import shapeles and convert

them in Stata format.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

sh2dta command

Syntax:

shp2dta using shp.lename, database(lename)

coordinates(lename) [options]

Example:

shp2dta using nuts2_164, database(data_shp) coordinates(coord)

genid(id) genc(c)

The shp2dta command generates two les:

data_shp.dta: contains information from .dbf le, id, latitude

(y_c) y longitude (x_c).

coord.dta: contains geometric information from .shp le.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

sh2dta command

Syntax:

shp2dta using shp.lename, database(lename)

coordinates(lename) [options]

Example:

shp2dta using nuts2_164, database(data_shp) coordinates(coord)

genid(id) genc(c)

The shp2dta command generates two les:

data_shp.dta: contains information from .dbf le, id, latitude

(y_c) y longitude (x_c).

coord.dta: contains geometric information from .shp le.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

Merging data sets

The new database (data_shp.dta) does not contain

information about economics variables.

Using the index of geometric objects has been generated a

excel le with variables from Eurostat: unemployment rate and
net migration rate.

Both dataset are easily jointed using POLY_ID as link variable:

import excel "C:\.\.\.\nuts2_164.xls", firstrow

save "C:\.\.\.\migr_unemp07_12.dta"
use data_shp, clear
merge 1:1 POLY_ID using migr_unemp, gen(union) force

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

Merging data sets

The new database (data_shp.dta) does not contain

information about economics variables.

Using the index of geometric objects has been generated a

excel le with variables from Eurostat: unemployment rate and
net migration rate.

Both dataset are easily jointed using POLY_ID as link variable:

import excel "C:\.\.\.\nuts2_164.xls", firstrow

save "C:\.\.\.\migr_unemp07_12.dta"
use data_shp, clear
merge 1:1 POLY_ID using migr_unemp, gen(union) force

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

Merging data sets

The new database (data_shp.dta) does not contain

information about economics variables.

Using the index of geometric objects has been generated a

excel le with variables from Eurostat: unemployment rate and
net migration rate.

Both dataset are easily jointed using POLY_ID as link variable:

import excel "C:\.\.\.\nuts2_164.xls", firstrow

save "C:\.\.\.\migr_unemp07_12.dta"
use data_shp, clear
merge 1:1 POLY_ID using migr_unemp, gen(union) force

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

Visualizing in maps
A choropleth is a map in which each area is coloured with an
intensity proportional to the value of a quantitative variable. Some
classical maps:

Quantiles: class breaks correspond to quantiles of the distribution

of variable (each class includes approximately the same number of
polygons).
Equal Intervals: class breaks correspond to values that divide the
distribution of variable attribute into k equal-width intervals.
Boxplot: the distribution of variable attribute is divided into 6
classes dened as follows: [min, p 25 1.5 iqr ],
(p 25 1.5 iqr , p 25], (p 25, p 50],(p 50, p 75], (p 75, p 75 + 1.5 iqr ]
and (p 75 + 1.5 iqr , max], where iqr is the interquartile range.
Standard Deviates: the distribution of variable attribute is divided
into k classes (2 k 9) whose width is dened as a fraction p of
its standard deviation sd .
M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

spmap command

Syntax:

spmap [attribute] [if] [in] using basemap [,basemap_options]

Details: basemap_options

polygon(polygon_suboptions)
line(line_suboptions)
point(point_suboptions)
diagram(diagram_suboptions)
arrow(arrow_suboptions)
label(label_suboptions)
scalebar(scalebar_suboptions)
graph_options]

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

spmap command

Syntax:

spmap [attribute] [if] [in] using basemap [,basemap_options]

Details: basemap_options

polygon(polygon_suboptions)
line(line_suboptions)
point(point_suboptions)
diagram(diagram_suboptions)
arrow(arrow_suboptions)
label(label_suboptions)
scalebar(scalebar_suboptions)
graph_options]

M. Herrera Spatial econometrics using Stata

Quantile map

spmap U2012 using coord, id(id) clmethod(q) title("Unemployment rate")

legend(size(medium) position(5)) fcolor(Blues2) note("Europe, 2012" "Source:
Eurostat")
Equal intervals map

spmap NM2012 using coord, id(id) clmethod(e) title("Net migration rate")

legend(size(medium) position(5)) fcolor(BuRd) note("Europe, 2012" "Source:
Eurostat")
Box map

spmap U2012 using coord, id(id) clmethod(boxplot) title("Unemployment rate")

legend(size(medium) position(5)) fcolor(Heat) note("Europe, 2012" "Source:
Eurostat")
Deviation map

spmap NM2012 using coord, id(id) clmethod(s) title("Net migration rate")

legend(size(medium) position(5)) fcolor(BuRd) note("Europa, 2012" "Source:
Eurostat")
Combine map

spmap U2012 using coord, id(id) fcolor(RdYlBu) cln(8) point(data(migr_unemp_shp)

xcoord(x_c) ycoord(y_c) deviation(NM2012) sh(T) fcolor(dknavy) size(*0.3))
legend(size(medium) position(5)) legt(Unemployment) note(" " "Solid triangles
indicate values over the mean of net-migration." "Europa, 2012. Source: Eurostat")
Combine map

spmap NM2012 using coord, id(id) fcolor(RdYlBu) cln(8) diagram(var(U2012)

xcoord(x_c) ycoord(y_c) fcolor(gs2) size(1.7)) legend(size(medium) position(5))
legstyle(3) legt(Net migration) note("Boxes indicate values of unemployment."
"Europe, 2012. Source: Eurostat")
 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

General Index
1 Introduction

2 Exploratory Spatial Data Analysis

Loading data and choropleth maps
Matrices and spatial tests
Spatial local analysis

3 Basic spatial econometrics

Taxonomy of spatial models
Spatial detection using OLS residual
Methods of estimation

4 Advanced spatial econometrics

Basic panel data models
Static spatial panel models
Dynamic spatial panel models

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

Centrality of spatial W

We show spatial concentration in previous maps, in a formal way:

0

yi ij ik yi ui
yj = ji 0 jk yj + uj , (6) (1)
yk ki kj 0 yk uk
y = Ay + u, (2)
Strategy of identication:
0 0

ij ik wij wik
A = ji 0 jk = wji 0 wjk = W .
ki kj 0 wki 0 wkj
We transform a non-identied model in other that contains only one
parameter: .
W captures `who is the neighbour of whom': must be EXOGENOUS!

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

Criteria used to create W

Usually, the building of W is an ad-hoc procedure of the researcher.
Common criteria are:
1 Geographical:

Distance functions:
inverse
inverse with threshold
Contiguity:
Rook
Queen
K nearest neighbours.
2 Socio-economic:

Similarity degree in economic dimensions (or social networks).

3 Combinations between both criteria.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

Generating W using Stata

In Stata there are (at least) three commands to generate W:

spatwmat:
Distance criterion.
Used for spatial univariate analysis.
Format le no compatible with spmat.
spwmatrix:
Generate W using geographic criteria (no contiguity).
Generate W under socio-economic criteria.
Import, export and manipulate from GeoDa.
Compatible format le with spatwmat.
spmat:
Generate W using geographic criteria (no under knn).
Import, export and read matrices from GeoDa.
Format le no compatible with spatwmat.
M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

Generating W using Stata

We will use a geographic criterion:

spwmatrix: for example 5-nn.

. spwmatrix gecon y_c x_c, wn(W5st) knn(5) row connect

Nearest neighbor (knn = 5) spatial weights matrix (164 x 164)
calculated successfully and the following action(s) taken:
- Spatial weights matrix created as Stata object(s): W5st.
- Spatial weights matrix has been row-standardized.
Connectivity Information for the Spatial Weights Matrix
- Sparseness: 3.049%
- Neighbors: Min : 5
Mean : 5
Median: 5
Max : 5

It is not advisable to work with units without neighbours.

In addition, it is usual to standardize W (usually row-standardize).
M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

Univariate spatial tests

The following statistics provide a measure of global spatial autocorrelation and
allow us to know its signicance.
Moran I test (1950):
(yi y )wij (yj y )
n i j
I= S0 N .
2
(yi y )
i=1

Geary c test (1954):

n n
2
wij (yi yj )
c = n1 i=1j=1
.
2S0 n
2
(yi y )
i=1

Getis-Ord G test (1992):

n n
wij yi yj
i j6=i
G= n n .
yi yj
i j6=i

Null hypotheses of tests: No spatial autocorrelation.

M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

Global spatial tests in Stata

. spatgsa U2012, w(W5st) moran geary two
Measures of global spatial autocorrelation
--------------------------------------------------------------
Moran's I
--------------------------------------------------------------
Variables | I E(I) sd(I) z p-value*
--------------------+-----------------------------------------
U2012 | 0.767 -0.006 0.045 17.084 0.000
--------------------------------------------------------------
Geary's c
--------------------------------------------------------------
Variables | c E(c) sd(c) z p-value*
--------------------+-----------------------------------------
U2012 | 0.228 1.000 0.054 -14.282 0.000
--------------------------------------------------------------
*2-tail test

. spatgsa U2012, w(W5bin) go two

Measures of global spatial autocorrelation
--------------------------------------------------------------
Getis & Ord's G
--------------------------------------------------------------
Variables | G E(G) sd(G) z p-value*
--------------------+-----------------------------------------
U2012 | 0.039 0.031 0.001 11.864 0.000
--------------------------------------------------------------
*2-tail test
M. Herrera Spatial econometrics using Stata
Moran's I scatterplot

splagvar U2012, wname(W5st) wfrom(Stata) ind(U2012) order(1) plot(U2012)

moran(U2012)
Moran's I scatterplot

splagvar NM2012, wname(W5st) wfrom(Stata) ind(NM2012) order(1) plot(NM2012)

moran(NM2012)
 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

General Index
1 Introduction

2 Exploratory Spatial Data Analysis

Loading data and choropleth maps
Matrices and spatial tests
Spatial local analysis

3 Basic spatial econometrics

Taxonomy of spatial models
Spatial detection using OLS residual
Methods of estimation

4 Advanced spatial econometrics

Basic panel data models
Static spatial panel models
Dynamic spatial panel models

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Loading data and choropleth maps
Basic spatial econometrics Matrices and spatial tests
Advanced spatial econometrics Spatial local analysis
Summary

Local indicators of spatial association

A version of Moran I test is used to detect spatial clusters in local
dimension:
n
(xi x)
Ii (d) = n wij (d) (xj x) , (3)
1 2
n (xi x) j=1,j6=i
i=1
where wij (d) is a weighting distance.
Null hypotheses is no spatial autocorrelation and the signicance of
Ii could be contrasted using normal distribution:

[Ii E [Ii ]]
z [Ii ] = p .
Var [Ii ]
This test allows grouping observations in 4 categories (see scatter
Moran): High-High (H-H), Low-Low (L-L), Low-High (L-H) and
High-Low (H-L).
M. Herrera Spatial econometrics using Stata
Local Moran's I scatterplot

genmsp_v0 U2012, w(W5st)

graph twoway (scatter Wstd_U2012 std_U2012 if pval_U2012>=0.05, msymbol(i) mlabel (id)
mlabsize(*0.6) mlabpos(c)) (scatter Wstd_U2012 std_U2012 if pval_U2012<0.05, msymbol(i) mlabel
(id) mlabsize(*0.6) mlabpos(c) mlabcol(red)) (lt Wstd_U2012 std_U2012), yline(0, lpattern(--))
xline(0, lpattern(--)) xlabel(-1.5(1)4.5, labsize(*0.8)) xtitle("{it:z}") ylabel(-1.5(1)3.5, angle(0)
labsize(*0.8)) ytitle("{it:Wz}") legend(o) scheme(s1color) title("Local Moran I of Unemployment
rate") splagvar NM2012, wname(W5st) wfrom(Stata) ind(NM2012) order(1) plot(NM2012)
moran(NM2012)
Local Moran's I map

spmap msp_U2012 using coord, id(id) clmethod(unique) title("Unemployment rate")

legend(size(medium) position(4)) ndl("No signif.") fcolor(blue red) note("Europe,
2012" "Source: Eurostat")
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

General Index
1 Introduction

2 Exploratory Spatial Data Analysis

Loading data and choropleth maps
Matrices and spatial tests
Spatial local analysis

3 Basic spatial econometrics

Taxonomy of spatial models
Spatial detection using OLS residual
Methods of estimation

4 Advanced spatial econometrics

Basic panel data models
Static spatial panel models
Dynamic spatial panel models

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Alternatives of specication
General Cli-Ord model (Manski model)

y = Wy + X + WX + u,
u = Wu + .

Imposing restrictions in , and we can obtain the following

models:

= 0, = 0, 6= 0 SLM .
= 0, 6= 0, = 0 SEM .
= 0, 6= 0, 6= 0 SARAR .
6= 0, = 0, 6= 0 SDM .

M. Herrera Spatial econometrics using Stata

Alternatives of specication
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

General Index
1 Introduction

2 Exploratory Spatial Data Analysis

Loading data and choropleth maps
Matrices and spatial tests
Spatial local analysis

3 Basic spatial econometrics

Taxonomy of spatial models
Spatial detection using OLS residual
Methods of estimation

4 Advanced spatial econometrics

Basic panel data models
Static spatial panel models
Dynamic spatial panel models

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Residual tests
The rst step (under the specic to general strategy M8) is to estimate a
regression model and obtain the residuals.
In our case our initial model is

U 2012 = 1 + 2 NM 2012 + u.
This equation is estimated under OLS:
. reg U2012 NM2012
Source | SS df MS Number of obs = 164
-------------+------------------------------ F( 1, 162) = 56.32
Model | 1453.94714 1 1453.94714 Prob > F = 0.0000
Residual | 4182.20231 162 25.8160636 R-squared = 0.2580
-------------+------------------------------ Adj R-squared = 0.2534
Total | 5636.14945 163 34.577604 Root MSE = 5.081
------------------------------------------------------------------------------
U2012 | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
NM2012 | -.7011928 .0934347 -7.50 0.000 -.8856998 -.5166859
_cons | 11.43504 .4697136 24.34 0.000 10.50749 12.36259
------------------------------------------------------------------------------

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Residual tests

There are a set of tests that allow the detection of spatial autocorrelation:

Parameters in H1
Null hypotheses Test
Spatial lag Error lag

- yes LMERROR
=0
yes yes
LMERROR
yes - LMLAG
=0
yes yes
LMLAG
No spatial
Moran s I
autocorrelation

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Spatial residual test in Stata

reg U2012 NM2012

spatdiag, weights(W5st)

Diagnostic tests for spatial dependence in OLS regression

------------------------------------------------------------
Diagnostics
------------------------------------------------------------
Test | Statistic df p-value
-------------------------------+----------------------------
Spatial error: |
Moran's I | 12.703 1 0.000
Lagrange multiplier | 148.081 1 0.000
Robust Lagrange multiplier | 0.750 1 0.386
|
Spatial lag: |
Lagrange multiplier | 182.220 1 0.000
Robust Lagrange multiplier | 34.889 1 0.000
------------------------------------------------------------

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Conclusion of spatial residual tests

Our rst step was a simple OLS:

U 20012 = 1 + 2 NM 2012 + u.

Now, according to the evidence reected in the tests, an SLM

should be estimated:

U 20012 = (W U 20012) + 1 + 2 NM 2012 + u.

Problem: How do we estimate this model?

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Estimation methods from spatial models

The estimation of models with spatial components can not be

performed by OLS (except simple cases such as SLX: WX).
The estimation needs alternative methods: ML or IV / GMM.

Syntax:

spreg estimator depvar [indepvars] [if] [in],

id(varname) [options]

Option estimator allows:

ml: maximum likelihood (ML).

gs2sls: generalized spatial two-stage least squares (IV/GMM).

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Estimation methods from spatial models

The estimation of models with spatial components can not be

performed by OLS (except simple cases such as SLX: WX).
The estimation needs alternative methods: ML or IV / GMM.

Syntax:

spreg estimator depvar [indepvars] [if] [in],

id(varname) [options]

Option estimator allows:

ml: maximum likelihood (ML).

gs2sls: generalized spatial two-stage least squares (IV/GMM).

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

General Index
1 Introduction

2 Exploratory Spatial Data Analysis

Loading data and choropleth maps
Matrices and spatial tests
Spatial local analysis

3 Basic spatial econometrics

Taxonomy of spatial models
Spatial detection using OLS residual
Methods of estimation

4 Advanced spatial econometrics

Basic panel data models
Static spatial panel models
Dynamic spatial panel models

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Maximum Likelihood: spreg ml

SLM

spreg ml U2012 NM2012, id(id) dlmat(W5_st)

SEM

spreg ml U2012 NM2012, id(id) elmat(W5_st)

SARAR

spreg ml U2012 NM2012, id(id) dlmat(W5_st) elmat(W5_st)

SDM

spreg ml U2012 NM2012 wx_NM2012, id(id) dlmat(W5_st)

Cli-Ord

spreg ml U2012 NM2012 wx_NM2012, id(id) dlmat(W5_st) elmat(W5_st)

M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Maximum Likelihood: spreg ml

SLM

spreg ml U2012 NM2012, id(id) dlmat(W5_st)

SEM

spreg ml U2012 NM2012, id(id) elmat(W5_st)

SARAR

spreg ml U2012 NM2012, id(id) dlmat(W5_st) elmat(W5_st)

SDM

spreg ml U2012 NM2012 wx_NM2012, id(id) dlmat(W5_st)

Cli-Ord

spreg ml U2012 NM2012 wx_NM2012, id(id) dlmat(W5_st) elmat(W5_st)

M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Maximum Likelihood: spreg ml

SLM

spreg ml U2012 NM2012, id(id) dlmat(W5_st)

SEM

spreg ml U2012 NM2012, id(id) elmat(W5_st)

SARAR

spreg ml U2012 NM2012, id(id) dlmat(W5_st) elmat(W5_st)

SDM

spreg ml U2012 NM2012 wx_NM2012, id(id) dlmat(W5_st)

Cli-Ord

spreg ml U2012 NM2012 wx_NM2012, id(id) dlmat(W5_st) elmat(W5_st)

M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Maximum Likelihood: spreg ml

SLM

spreg ml U2012 NM2012, id(id) dlmat(W5_st)

SEM

spreg ml U2012 NM2012, id(id) elmat(W5_st)

SARAR

spreg ml U2012 NM2012, id(id) dlmat(W5_st) elmat(W5_st)

SDM

spreg ml U2012 NM2012 wx_NM2012, id(id) dlmat(W5_st)

Cli-Ord

spreg ml U2012 NM2012 wx_NM2012, id(id) dlmat(W5_st) elmat(W5_st)

M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Maximum Likelihood: spreg ml

SLM

spreg ml U2012 NM2012, id(id) dlmat(W5_st)

SEM

spreg ml U2012 NM2012, id(id) elmat(W5_st)

SARAR

spreg ml U2012 NM2012, id(id) dlmat(W5_st) elmat(W5_st)

SDM

spreg ml U2012 NM2012 wx_NM2012, id(id) dlmat(W5_st)

Cli-Ord

spreg ml U2012 NM2012 wx_NM2012, id(id) dlmat(W5_st) elmat(W5_st)

M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Maximum Likelihood: spreg ml

Variable SLM SEM SARAR SDM CLIFF-ORD

NM 2012 0.19 0.15 0.16 0.16 0.15
W NM 2012 0.10 0.03
const 2.41 10.97
1.74 2.82 1.87

b 0.82 0.88 0.80 0.87
b 0.86 0.33 0.31
Nota: p < 0.05.

What is the best model?

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Maximum Likelihood: selecting the best model

From specic to general strategy:

Using LM tests: spatial lag model (SLM).

But there are other posible models: SDM, SARAR and

Cli-Ord.

Between SDM and SEM: LRCOMFAC .

From general to specic strategy:

Start using Cli-Ord model (M1) and to eliminate sequentially

non-signicant variables.

M. Herrera Spatial econometrics using Stata

Maximum Likelihood: Common factor test
Assuming the SDM model has been estimated:

y = Wy + X + WX + u.

H0 : + = 0,
The null and alternative hypotheses are:
H1 : + 6= 0.
Under H0 , = , and replacing into the SDM model:

y = Wy + X + WX ( ) + u = Wy + X WX + u,
(I W ) y = (I W ) X + u.

The last expression is summarized in SEM : y = X + (I W )1 ,

where has been replaced by .
Under null hypothesis, we have an SEM and, under alternative hypothesis, an
SDM :

LRCOMFAC = 2 l|H1 l|H0 q2

h i
as
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Maximum Likelihood: Alternative models

Variable SLM SEM SARAR SDM CLIFF-ORD

NM 2012 0.19 0.15 0.16 0.16 0.15
W NM 2012 0.10 0.03
const 2.41 10.97
1.74 2.82 1.87

b 0.82 0.88 0.80 0.87
b 0.86 0.33 0.31
LRCOMFAC 6.81

Nota: p < 0.05.

The nal model is the SLM.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

IV-GMM: spivreg command

SLM

spivreg U2012 NM2012, dl(W5_st) id(id)

SEM

spivreg U2012 NM2012, el(W5_st) id(id)

SARAR

spivreg U2012 NM2012, dl(W5_st) el(W5_st) id(id)

SDM

spivreg U2012 NM2012 wx_NM2012, dl(W5_st) id(id)

Cli-Ord

spivreg U2012 NM2012 wx_NM2012, dl(W5_st) el(W5_st) id(id)

M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

IV-GMM: spivreg command

SLM

spivreg U2012 NM2012, dl(W5_st) id(id)

SEM

spivreg U2012 NM2012, el(W5_st) id(id)

SARAR

spivreg U2012 NM2012, dl(W5_st) el(W5_st) id(id)

SDM

spivreg U2012 NM2012 wx_NM2012, dl(W5_st) id(id)

Cli-Ord

spivreg U2012 NM2012 wx_NM2012, dl(W5_st) el(W5_st) id(id)

M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

IV-GMM: spivreg command

SLM

spivreg U2012 NM2012, dl(W5_st) id(id)

SEM

spivreg U2012 NM2012, el(W5_st) id(id)

SARAR

spivreg U2012 NM2012, dl(W5_st) el(W5_st) id(id)

SDM

spivreg U2012 NM2012 wx_NM2012, dl(W5_st) id(id)

Cli-Ord

spivreg U2012 NM2012 wx_NM2012, dl(W5_st) el(W5_st) id(id)

M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

IV-GMM: spivreg command

SLM

spivreg U2012 NM2012, dl(W5_st) id(id)

SEM

spivreg U2012 NM2012, el(W5_st) id(id)

SARAR

spivreg U2012 NM2012, dl(W5_st) el(W5_st) id(id)

SDM

spivreg U2012 NM2012 wx_NM2012, dl(W5_st) id(id)

Cli-Ord

spivreg U2012 NM2012 wx_NM2012, dl(W5_st) el(W5_st) id(id)

M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

IV-GMM: spivreg command

SLM

spivreg U2012 NM2012, dl(W5_st) id(id)

SEM

spivreg U2012 NM2012, el(W5_st) id(id)

SARAR

spivreg U2012 NM2012, dl(W5_st) el(W5_st) id(id)

SDM

spivreg U2012 NM2012 wx_NM2012, dl(W5_st) id(id)

Cli-Ord

spivreg U2012 NM2012 wx_NM2012, dl(W5_st) el(W5_st) id(id)

M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

IV-GMM: Alternative models

Variable SLM SEM SARAR SDM CLIFF-ORD

NM 2012 0.14 0.20 0.15 0.14 0.16
W NM 2012 0.02 0.06
const 1.57 10.50
1.56 1.40 2.16

b 0.89 0.90 0.91 0.85
b 0.83 0.27 0.09
Nota: p < 0.05.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Correction by heteroskedasticity

Variable SLM SEM SARAR SDM CLIFF-ORD

NM 2012 0.14 0.21 0.15 0.14 0.16
W NM 2012 0.02 0.06
const 1.57 10.50
1.57 1.40 2.18

b 0.89 0.90 0.91 0.85
b 0.82 0.18 0.02
Nota: p < 0.05.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

Interpretation of estimated parameters

In SLM, SARAR or SDM models, a change of the variable xk

in region i will aect the region itself and aects potentially
the other regions indirectly through the spatial multiplier
mechanism ((I W )1 ).
In a linear model, the marginal eect is:

E (yi ) b E (yj )
= k =0
xik xik

but in spatial models with Wy and/or Wx , the second eect is

not zero.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Taxonomy of spatial models
Basic spatial econometrics Spatial detection using OLS residual
Advanced spatial econometrics Methods of estimation
Summary

SLM. Direct and indirect eects

The marginal eect of the explanatory variable xk on the dependent variable is:
y1 y1
x1k
xnk
.. .. ..
h i
y y

x 1k
... xnk
=
. . .
,

yn yn
x1k
x 1k
0 0

k
0 k 0
(In W )1

= .. .. .. .. ,

. . . .
0 0 k
1
= (In W ) [k In ] , (4)
Direct eect: average of the elements of principal diagonal of
(In W )1 [k In ].
Indirect eect: (spatial spillover) average of sum of rows, without of elements
of principal diagonal of (In W )1 [k In ].
M. Herrera Spatial econometrics using Stata
Example under SLM in Stata
If we apply the above expression to our example (SLM using MLE):
. mata:
---------------------- mata (type end to exit)-------------------
: b = st_matrix("e(b)")
: b
1 2 3 4
+-------------------------------------------------------------+
1 | -.1898462498 2.40790211 .8174714987 8.182474325 |
+-------------------------------------------------------------+
: lambda = b[1,3]
: lambda
.8174714987
: S = luinv(I(rows(W))-lambda*W)
: end
-----------------------------------------------------------------
. * Total effects
. mata: (b[1,1]/rows(W))*sum(S)
-1.040090991
* Direct effects
. mata: (b[1,1]/rows(W))*trace(S)
-.2414164099
. * Indirect effects (spatial spillovers)
. mata: (b[1,1]/rows(W))*sum(S) - (b[1,1]/rows(W))*trace(S)
-.7986745812
 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

General Index
1 Introduction

2 Exploratory Spatial Data Analysis

Loading data and choropleth maps
Matrices and spatial tests
Spatial local analysis

3 Basic spatial econometrics

Taxonomy of spatial models
Spatial detection using OLS residual
Methods of estimation

4 Advanced spatial econometrics

Basic panel data models
Static spatial panel models
Dynamic spatial panel models

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

Fixed eect and random eects

Consider a linear model with k independent variables xit :

yit = xit + uit , (5)

where i = 1, . . . , n, t = 1, . . . , T and uit is a random error term.
This model doesn't control by heterogeneity: specic temporal variables
could be aect on dependent variable.
Solution:
uit = i + t + it ,
where i is a common region-specic eect and t is a common
time-specic eect for all regions.
These eects could be treated as xed or random.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

Fixed or random eects

Hausman test (1978) is computed as:

0
H = (fe re ) (Vfe Vre )1 (fe re ) ,
where fe is the vector of coecients of the consistent estimator
fe, re is the vector of coecients of the ecient estimator re,
with Vfe and Vre as the variance-covariance matrix of fe and re,
2
respectively. This statistic is distributed as q , with q degrees
(number of common coecients in both models).
Hausman test can be consider as a statistic of validation of re
estimator, null hypotheses.

M. Herrera Spatial econometrics using Stata

Detection of spatial dependence
To incorporate spatial eects we must have some evidence of their presence. A
possible test that can be used is CD test (Pesaran, 2004):
!
n1 n
2T
q
CD = n(n1) bij ,
i=1 j=i+1

where bij is the correlation coecient in the residuals between i and j :

T
ubit ubjt
t=1
bij = bji =  1/2  1/2 , (1) (6)
T T
ubit2 ubjt2
t=1 t=1
Null hypothesis: no autocorrelation in cross-section dimension.
In Stata:
. xtreg U NM, fe
(ommitted product)
. xtcsd, pes abs
Pesaran's test of cross sectional independence = 60.169, Pr = 0.0000
Average absolute value of the off-diagonal elements = 0.464
 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

General Index
1 Introduction

2 Exploratory Spatial Data Analysis

Loading data and choropleth maps
Matrices and spatial tests
Spatial local analysis

3 Basic spatial econometrics

Taxonomy of spatial models
Spatial detection using OLS residual
Methods of estimation

4 Advanced spatial econometrics

Basic panel data models
Static spatial panel models
Dynamic spatial panel models

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

Spatial lag model

The SLM with xed eects is:
yt = Wyt + Xt + + t ,
(7)
t N 0, 2 In ,
 

where

y1t x11t x21t xk 1t 1
y2t x12t x22t xk 2t 2
yt = . , Xt = . .. .. .. , = . .
.. .. ..

. . .
ynt x1nt x2nt xknt n
Under random eects, this model can be written as:
yt = Wyt + Xt + + t ,
(8)
| {z }
ut
t N 0, 2 In , N 0, 2 In .
  

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

SLM. Direct and indirect eects

The partial eect of one unit increases on the SLM model is as follows:
y1 y1
x1k
xnk
.. .. ..
h i
y y

x1k
... xnk
=
. . .
,

yn yn
x1k
x1k
0 0

k
0 k 0
(In W )1

= .. .. .. .. ,

. . . .
0 0 k
1
= (In W ) [k In ] , (9)
Direct eect: average of the elements of principal diagonal of
(In W )1 [k In ].
Indirect eect: (spatial spillover) average of sum of rows, without of elements
of principal diagonal of (In W )1 [k In ].
M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

Spatial Error Model

The SEM model with xed eects is:

yt = Xt + + t

t = W t + t (10)

t N 0, 2 In
 

and the version of SEM model with xed eects is:

yt = Xt + + t ,
| {z }
ut
 t = W t + t , 
t N 0, 2 In , N 0, 2 In ,


M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

Spatial Durbin Model

SDM specication:

yt = Wyt + Xt + WXt + t , (11)

with direct-indirect eects:

y1 y1

x1k xnk
.. .. ..
h i
y y
.

x1k ... xnk =
. . ,

yn yn
x1k x1k

k w12 k w1n k
w21 k k w2n k
= (In W )1 .. .. .. .. ,

. . . .
wn1 k wn2 k k
= (In W )1 [k In + k W ] , (12)
M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

command xsmle

SLM

xsmle U NM, fe wmat(W5_st) mod(sar) hausman

SEM

xsmle U NM, fe emat(W5_st) mod(sem) hausman

SARAR

xsmle U NM, fe wmat(W5_st) emat(W5_st) vce(r) mod(sac)

SDM

xsmle U NM, fe wmat(W5_st) mod(sdm) hausman

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

command xsmle

SLM

xsmle U NM, fe wmat(W5_st) mod(sar) hausman

SEM

xsmle U NM, fe emat(W5_st) mod(sem) hausman

SARAR

xsmle U NM, fe wmat(W5_st) emat(W5_st) vce(r) mod(sac)

SDM

xsmle U NM, fe wmat(W5_st) mod(sdm) hausman

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

command xsmle

SLM

xsmle U NM, fe wmat(W5_st) mod(sar) hausman

SEM

xsmle U NM, fe emat(W5_st) mod(sem) hausman

SARAR

xsmle U NM, fe wmat(W5_st) emat(W5_st) vce(r) mod(sac)

SDM

xsmle U NM, fe wmat(W5_st) mod(sdm) hausman

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

command xsmle

SLM

xsmle U NM, fe wmat(W5_st) mod(sar) hausman

SEM

xsmle U NM, fe emat(W5_st) mod(sem) hausman

SARAR

xsmle U NM, fe wmat(W5_st) emat(W5_st) vce(r) mod(sac)

SDM

xsmle U NM, fe wmat(W5_st) mod(sdm) hausman

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

Alternative Models

Variable SLM SEM SARAR SDM

NM 0.16 0.16 0.13 0.15
W NM 0.02
b 0.79 0.84 0.78

b 0.87 0.29
COMFAC 88.41

Spatial effects (long run)

Directs 0.19 0.17 0.18
Indirects 0.54 0.66 0.58
Totals 0.74 0.84 0.76
AIC 2352.79 2733.18 2629.62 2642.88

Nota: p < 0.05.

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

General Index
1 Introduction

2 Exploratory Spatial Data Analysis

Loading data and choropleth maps
Matrices and spatial tests
Spatial local analysis

3 Basic spatial econometrics

Taxonomy of spatial models
Spatial detection using OLS residual
Methods of estimation

4 Advanced spatial econometrics

Basic panel data models
Static spatial panel models
Dynamic spatial panel models

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

Dynamic SLM models

Dynamic versions of Spatial Lag Model (dynSLM ):

yt = yt1 + Wyt + Xt + + t , (13)

yt = Wyt1 + Wyt + Xt + + t , (14)
yy = yt1 + Wyt1 + Wyt + Xt + + t , (15)
These model give us the option to obtain direct and indirect eects in
short and long run:
Short run (assuming = = 0,):

= (In W )1 [k In ] .
h i
y
x1k ... y
xnk (16)
t
Long run (assuming yt = yt1 = y ):

= [(1 ) In ( + ) W ]1 [k In ] .
h i
y
x1k ... y
xnk (17)
t

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

Dynamic SDM models

Dynamic versions of Spatial Durbin Model (dynSDM ):
yt = yt1 + Wyt + Xt + WXt + + t , (18)
yt = Wyt1 + Wyt + Xt + WXt + + t , (19)
yt = yt1 + Wyt1 + Wyt + Xt + WXt + + t , (20)
These model give us the option to obtain direct and indirect eects in
short and long run:
Short run (assuming = = 0,):

= (In W )1 [k In + k W ] .
h i
y
x1k ... y
xnk (21)
t
Long run (assuming yt = yt1 = y ):

= [(1 ) In ( + ) W ]1 [k In + k W ] . (22)
h i
y y
x1k ... xnk t

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

Example of unemployment-migration. Serial correlation

Wooldridge (2002) develops a simple statistic for autocorrelation

detection.
Drukker (2003) implements it in Stata, xtserial command:

. xtserial U NM
Wooldridge test for autocorrelation in panel data
H0: no first-order autocorrelation
F( 1, 163) = 110.869
Prob > F = 0.0000

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

xsmle command

SLM 1 (eq. 13)

xsmle U NM, dlag(1) fe wmat(W5_st) type(ind) mod(sar)

effects nsim(499)

SLM 2 (eq. 14)

xsmle U NM, dlag(2) fe wmat(W5_st) type(ind) mod(sar)

effects nsim(499)

SLM 3 (eq. 15)

xsmle U NM, dlag(1) fe wmat(W5_st) type(ind) mod(sar)

effects nsim(499)

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

xsmle command

SLM 1 (eq. 13)

xsmle U NM, dlag(1) fe wmat(W5_st) type(ind) mod(sar)

effects nsim(499)

SLM 2 (eq. 14)

xsmle U NM, dlag(2) fe wmat(W5_st) type(ind) mod(sar)

effects nsim(499)

SLM 3 (eq. 15)

xsmle U NM, dlag(1) fe wmat(W5_st) type(ind) mod(sar)

effects nsim(499)

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis Basic panel data models
Basic spatial econometrics Static spatial panel models
Advanced spatial econometrics Dynamic spatial panel models
Summary

xsmle command

SLM 1 (eq. 13)

xsmle U NM, dlag(1) fe wmat(W5_st) type(ind) mod(sar)

effects nsim(499)

SLM 2 (eq. 14)

xsmle U NM, dlag(2) fe wmat(W5_st) type(ind) mod(sar)

effects nsim(499)

SLM 3 (eq. 15)

xsmle U NM, dlag(1) fe wmat(W5_st) type(ind) mod(sar)

effects nsim(499)

M. Herrera Spatial econometrics using Stata

Alternative models of dynamic SLM

Variable SLM 1 SLM 2 SLM 3

U(t 1) 0.46 0.75
NM 0.07 0.03 0.10
W U 0.75 0.76 0.99
W U(t 1) 0.30 0.51
Spatial effects (short run)
Directs 0.08 0.03 0.15
Indirects 0.19 0.08 12.52
Totals 0.27 0.11 12.37
Spatial effects (long run)
Directs 0.10 0.81 0.22
Indirects 0.44 1.25 0.66
Totals 0.34 0.45 0.44
AIC 1839.55 1972.80 1812.61

Nota: p < 0.05.

 Introduction
Exploratory Spatial Data Analysis
Basic spatial econometrics
Advanced spatial econometrics
Summary

Summing up
Stata has incorporated tools for spatial analysis.

ESDA can be carried out completely, as in others software.

Also, for cross-section data, the most common spatial

specications can be estimated by ML and/or IV/GMM.

For panel data, recent developments provide alternatives for

estimating static and dynamic models.

Main results of the impact of net migration:

Cross-section model: SLM shows a negative impact in
unemployment (long run eect).
Panel-data: Dynamic SLM shows a positive impact in short
run (orthodox theory) and negative impact in long run, total
eect (NEG theory).
M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis
Basic spatial econometrics
Advanced spatial econometrics
Summary

Summing up
Stata has incorporated tools for spatial analysis.

ESDA can be carried out completely, as in others software.

Also, for cross-section data, the most common spatial

specications can be estimated by ML and/or IV/GMM.

For panel data, recent developments provide alternatives for

estimating static and dynamic models.

Main results of the impact of net migration:

Cross-section model: SLM shows a negative impact in
unemployment (long run eect).
Panel-data: Dynamic SLM shows a positive impact in short
run (orthodox theory) and negative impact in long run, total
eect (NEG theory).
M. Herrera Spatial econometrics using Stata
 Introduction
Exploratory Spatial Data Analysis
Basic spatial econometrics
Advanced spatial econometrics
Summary

Some references
Theoretical references:

Anselin, L. and A. Bera (1998). Spatial dependence in linear regression models

with an Introduction to Spatial Econometrics, Handbook of Applied Economic
Statistics, pp. 237-289.

Brueckner, J. (2003). Strategic interaction among governments: An overview

of empirical studies, International Regional Science Review, 26(2).

Elhorst, J. P. (2013). Spatial Panel Models, Chapter 82, en Fischer y Nijkamp

(eds.) Handbook of Regional Science. Heidelberg: Springer.

Applied references:

Ertur, C. and W. Koch (2007). Growth, technological interdependence and

spatial externalities: Theory and evidence, Journal of Applied Econometrics, 22.

Le Gallo, J. and C. Ertur (2003). Exploratory spatial data analysis of the

distribution of regional per capita GDP in Europe, 1980 = 1995, Papers in
regional science, 82(2).

Fingleton, B. (2006). A cross-sectional analysis of residential property prices:

the eects of income, commuting, schooling, the housing stock and spatial
interaction in the English regions, Papers in Regional Science, 85(3).

M. Herrera Spatial econometrics using Stata

 Introduction
Exploratory Spatial Data Analysis
Basic spatial econometrics
Advanced spatial econometrics
Summary

Some references
Books:

Anselin, L. (1988). Spatial Econometrics: Methods and Models.

LeSage, J. and Pace (2009). An introduction to spatial econometrics.

Elhorst, J. P. (2014). Spatial econometrics: from cross-sectional data to spatial

panels.

Stata:

Belotti, F. et al (2016). XSMLE: Stata module for spatial panel data models
estimation. Statistical Software Components.

Drukker, D. M. et al. (2011). A command for estimating spatial-autoregressive

models with spatial-autoregressive disturbances and additional endogenous
variables. Econometric Reviews, 32, 686-733.

Drukker, D. M. et al. (2013). Creating and managing spatial-weighting matrices

with the spmat command. Stata Journal, 13(2), 242-286.

Pisati, M. (2008). SPMAP: Stata module to visualize spatial data. Statistical

Software Components.

M. Herrera Spatial econometrics using Stata