How Does Democratic Socialism Impact Economic Growth?

8/20/2019 How Does Democratic Socialism Impact Economic Growth?

1/21

December 2, 2015AbstractDemocratic socialist, Bernie Sanders, has become the center of attention for many potential left

wing United States voters, in the 201 presidential elections! "is candidacy raises the #$estion% issocialism an effective way to growth an economy& Using panel data controlling for co$ntries and

years, ' generated a m$ltit$de of regression models and a statistical analysis based on (D) per capita growth, in regards to a set of e*planatory variables in lie$ of democratic socialism! 'n an

effort to assess the practicality of democratic socialism, this st$dy was commenced! Some of themodels ' developed co$ld not +$stify any impact of certain socialistic val$es toward economic

growth! n the contrary, correlation and some statistical significance between the characteristicsand economic growth was discovered!

"- D.S D./A'

S'A3'S/ '/)A

.4/' (-"&


2/21

olantonio 1

'ntrod$ction

Democratic socialism is an ideology involving a combination of political

democracy and social ownership over the means of prod$ction, a socialist economic

system! he concept of democratic socialism has been raised into the p$blic eye recently

by ermont Senator, Bernie Sanders! Sanders plans to r$n for the 201 presidential

election and has gained a significant following in less than a year, since his

anno$ncement! )erhaps this is d$e to 526 of Democrats favoring socialism! .ven more

prevalent, 576 of Democratic voters wo$ld be o8ay with voting for a socialist 9Bla8e:!

Bernie;s r$n for presidency raises the #$estion if socialism is an effective way to operate

an economic system! By e*ploring different economic aspects comprising democratic

socialist policy, we can potentially learn how effective democratic socialism wo$ld be at

growing not only the United States economy, b$t even the economies of other nations!he e*planatory variables st$died, as well as (D) per capita growth were

meas$red as percent;s ta8en from the -orld Ban8 Development 'ndicators! )ercent

indicators were ta8en as a system of scaling the panel data and for a base of consistency

among all variables in st$dy! 't is important to $nderstand that this st$dy was completed

$sing $nbalanced panel data, with information dating bac8 to 170 from 177 co$ntries!

he socialist democrat characteristics foc$sed in this st$dy incl$de% government

e*pendit$re on ed$cation, p$blic health e*pendit$re, s$bsidies and other transfers, wage

and salaried wor8ers, and total ta* rate! .ach variable was ta8en as a piece of a greater

p$


3/21

olantonio 2

health e*pendit$re and wage and salaried wor8ers were all discovered to be significant at

the 16 level in negatively affecting economic growth per capita!

3iterat$re eview

A past st$dy on health e*pendit$re theory was completed investigating the impact

of (D) per capita and p$blic financing as a percentage of total financing! his article

stresses the need for more macroeconomic st$dies on health e*pendit$re impact, as well

as $pdated replications of data to $nify res$lts of previo$s analysis! )oliticians,

academics and administrators contin$o$sly $se health e*pendit$re as a disc$ssion point

as a share of (ross Domestic )rod$ct! "ealth e*pendit$re between co$ntries varies

vastly!'n this health e*pendit$re st$dy light is shed onto some of the not as apparent

effects behind health e*pendit$re between co$ntries! 't also compiles literat$re on health

e*pendit$re econometric theory, which has been of s$bstantial interest in and o$tside of

health economics, for decades! /ost efforts from different st$dies, to find possible

significance in variables to e*plain health e*pendit$re has generally raised more

#$estions! Also, many of the independent variables st$died alongside health e*pendit$re

is sparse and only br$shed $pon 9(erdtham:!'n another investigation, government e*pendit$re as a whole was of interest in

affecting economic growth! (overnment e*pendit$re;s impact on economic growth is of

partic$lar interest as it spar8s the #$estion of +$st how government impacts the economy!

(overnment e*pendit$re is $sed to represent government impact beca$se there is not

eno$gh positively correlated dimensions of economic welfare available to represent the

$ne*plained variation! .ven so tho$gh the partial relationship with government

e*pendit$re is still intrig$ing, beca$se government involvement is of s$ch interest!


4/21

olantonio =

)er capital real (D) was $sed as the meas$re for the dependent variable,

economic growth! he independent variable, government e*pendit$re, was meas$red by

government spending in this st$dy! /any wo$ld arg$e more spending wo$ld increase

(D), while others wo$ld spec$late that government spending wo$ld decrease (D),

generally! 'n this st$dy, these contrasting views are p$t to the test, not attempting to

e*plain the impact of different e*pendit$res individ$al impact!D$e to theoretical limitations behind government e*pendit$re, wea8ens the data!

-ith that said the negative beta val$es, fo$nd in this st$dy, for most co$ntry income

levels co$ld be insightf$l! he res$lts are consistent with a pro> free mar8et viewpoint, in

that growth of government h$rts economic growth 93anda$:!

Another st$dy investigates the implications of fiscal policy on economic growth!

/any economists hold an opinion on fiscal policy as a growth determinant for economic

well>being of a partic$lar co$ntry, in some respect! 't is a well reno$nced belief that

elements of fiscal policy s$ch as ta*ation, p$blic investment and more contrib$te to

growth! he empirical evidence giving light on the importance of fiscal policy in

infl$encing economic growth is rather sparse! -ith that said, the meas$rement error in

the data sets $sed is rather s$bstantial, b$t there is #$ite a bit of information at the same

time!

As a res$lt of this fiscal policy st$dy, any indication that ta* rates are a

determinate of economic growth, is delicate! 't is e*ceptionally diffic$lt to determine the

effects of ta* policy on growth, with this st$dy hinting ta*es may not even matter in

terms of economic growth! a$sality #$estions still linger to the effect of the fiscal

policy based variables $sed on economic growth 9.asterly:!


5/21

olantonio ?

3iterat$re has also been p$blished on the impact of immigration on host co$ntry

wages, employment and economic growth! An indisp$tably to$chy iss$e thro$gho$t

ind$striali


6/21

olantonio 5

an alternative hypothesis, the social economic variables $sed had an impact in e*plaining

economic growth!

Data Analysis

o test the impact of social economic variables on (D) per capita growth, the

following pop$lation model was $sed% y 0 C 1* C where y is the 6 ann$al (D) per

capita growth and * contains government e*pendit$re on ed$cation and p$blic health

e*pendit$re meas$red as 6 of total government e*pendit$re! Also contained within *%

s$bsidies and other transfers as a 6 of total e*pense, total ta* rate as a 6 of commercial

profit, and wage and salaried wor8ers as a 6 of total employed! .ach variable of interest

was meas$red as a percentage, in order to standardi


7/21

olantonio

e*pendit$re comprises of intermittent and capital, b$dget based government spending,

social health ins$rance f$nds and grants!

a*es withheld and remitted to ta* a$thorities are e*cl$ded from total ta* rate as a

6 of commercial profits! otal ta* rate does meas$re the ta*es b$sinesses m$st pay after

ded$ctions and any commercial profit e*emptions are made! 't is important to note that

some ta* rates will be range above 1006, beca$se b$sinesses can be ta*ed in vario$s ta*

brac8ets based on commercial profits, depending on different co$ntries fiscal policies

94ew Eor8 imes:! -age and salaried wor8ers are defined as wor8ers who have held a

+ob with contracted paid employment! -age and salaried wor8ers can be viewed b$t not

defined as a f$nction and representation of legal employment within a co$ntry!

-ith s$ch range of variables of interest, comes a wide range of observations

between variables! @or (D) per capita growth, F15? observations were gathered, while

for government e*pendit$re on ed$cation only 15G= observations were fo$nd! -ith that

said, the social economic variable data was $nbalanced as panel data with (D) per capita

growth! he difference in observations between variables ca$ses somewhat of a loss of

efficiency in the data gathered! Unfort$nately, the missing data is systematic and ca$sed

by a limitation and inconsistency in data provided by the -orld Ban8 Development

'ndicators which were $sed for this st$dy! .ven with this iss$e of total observational

variation, paneling the variables $sed with co$ntries and years is essential to controlling

s$ch large samples! Unbalanced data may be the res$lt, b$t is not a deliberate concern!

Using both fi*ed and random effects models, in 3S regression analysis, will acco$nt for

a range of possible concl$sions!


8/21

olantonio G

Descriptive statistics for each variable depict the shape of their respective

distrib$tions! (D) per capita growth holds a mean of abo$t 2!16 with a standard

deviation of abo$t !G6! -ith a minim$m val$e of 56 and a ma*im$m val$e of abo$t

1F26, ' relatively spread o$t distrib$tion co$ld be drawn! -ith a standard deviation of

?!56, p$blic health e*pendit$re wo$ld have a more condensed distrib$tion ranging from

abo$t !16 to =?!?6 9able 1:! 't can been recogni


9/21

olantonio F

independence of each variable of interest, a correlation matri* is displayed within able

2!

All variables of interest have a correlation between >0!5 and !5, indicating wea8 to

no correlation! ariables with the highest correlation incl$de s$bsidies and other

transfers in relation to government e*pendit$re on ed$cation and p$blic health

e*pendit$re! -age and salaried wor8s had a partic$larly wea8 b$t still prevalent

correlation to p$blic health e*pendit$re and s$bsidies and transfers! he most important

item to recogni


10/21

olantonio 7

variable so controlling for time is an especially necessary action, especially in the m$lti>

linear regression, where only =7F observations were $sed!

Both the fi*ed effects and the fi*ed time effects models fo$nd government

e*pendit$re on ed$cation, p$blic health e*pendit$re and wage and salaried wor8ers

significant at the 16 level! Additionally, the fi*ed effects model showed s$bsidies and

transfers significant at the 106 level! 'n a fi*ed effects m$ltilinear regression, p$blic

health e*pendit$re was significant at the 106 level! otal ta* rate was significant at the

56 level $nder a fi*ed effects model, b$t was fo$nd insignificant when controlling for

time! 'n a fi*ed time effects m$ltilinear regression, p$blic health e*pendit$re was

discovered significant at the 56 level!

-ith that said, the fi*ed effects model based tables showed significance between

specific social economic variables and (D) per capita growth! Using rob$st fi*ed

effects, for each percent increase in government e*pendit$re on ed$cation, (D) per

capita growth is estimated to decrease by !07? percent! @or every percent increase in

p$blic health e*pendit$re, (D) per capita growth decreases by !0G= percent! @or every

percent increase in e*pense of s$bsidies and other transfers, (D) per capita growth is

raised by !00F percent! @or every percent increase in wage and salaried wor8ers of total

employed, (D) per capita growth decreases by !02= percent! 'n m$ltilinear fi*ed effects

regression, for each percent increase in p$blic health e*pendit$re, (D) per capita growth

raises by !5G percent! Also for each percent of commercial profits increase in total ta*

rate, (D) per capita growth is increased by !117 percent! 't is also important to recognitest significance at the 106 level!


11/21

olantonio 10

-hen controlling for time, coefficient o$tp$ts for a rob$st fi*ed effects model

were #$ite different! -ith each percent increase in government e*pendit$re on

ed$cation, (D) per capita growth is estimated to decrease by !07G percent! @or each

percent increase in p$blic health e*pendit$re, (D) per capita growth is decreased by !0G2

percent! @or each percent increase in wage and salaried wor8ers of total employed, (D)

per capita growth decreases by !02 percent! 'n m$ltilinear fi*ed time effects regression,

for every percent increase in p$blic health e*pendit$re, (D) per capita growth increases

by !== percent! 'n this model, the +oint variable probability test is significant at the 16

level!

'n the rob$st random effects model, controlling for co$ntries, for each percent

increase in wage and salaried wor8s of total employed, (D) per capita growth decreases

by !0G percent! 'n a m$ltilinear regression model, with rob$st random effects and

controlling for co$ntries, for each percent increase in p$blic health e*pendit$re, (D) per

capita growth is estimated to increase by !5G percent! Under the same model, for every

percent increase in total ta* rate as a segment of commercial profits, (D) per capita

growth is estimated to increase by !117 percent!

/ost of the variables of interest which were fo$nd most significant, were only

significant when individ$ally regressed with (D) per capita growth! 'ndivid$al

regressions co$ld have demonstrated a greater significance, beca$se more so of each

regressions total variation was e*plained by their models, than that of the m$ltilinear

regressions! his is portrayed by the ad+$sted 2 val$es, given in the fi*ed time effects

tables! -hen time was not controlled for, this theory was shown to not be as li8ely as

relevant for variables, as less of the total variation for each model co$ld be e*plained!


12/21

olantonio 11

@or that reason, it co$ld be arg$ed that the coefficients for fi*ed effects are not as

acc$rate as the coefficients for fi*ed time effects, beca$se significantly less variation is

acco$nted for!

oncl$sion

he impact of p$blic health e*pendit$re, total ta* rate and wage and salaried

wor8ers on (D) per capita growth were most apparent! verall, s$ggesting that

individ$ally these significant variables negatively impact economic growth, while

together in m$ltilinear regression hold a positive impact! 'n a more real world scenario,

these social economic variables definitely come hand in hand! -ith that said, it is most

important to recogni


13/21

olantonio 12

he positive impact on economic growth that these variables have, can s$pport

ideologies s$ch as that a healthy wor8 force is a more prod$ctive one and that ta*ing big

b$siness is healthy! Ees, more information can and sho$ld be gathered on the impact of

increasing total ta* rate and p$blic health e*pendit$re! @or e*ample, the effect of total

ta* rate and p$blic health e*pendit$re on employment! 'f employment decreases, d$e to

these social economic variables, then nat$rally the economy wo$ld be s$ffer! Statistically

tho$gh, increasing these variables wo$ld be healthy to growing the United States

economy and sho$ld be raised accordingly!


14/21

olantonio 1=

eferences

Bla8e, Aaron! Io$ld a Socialist Act$ally Be .lected )resident&I Washington Post ! he

-ashington )ost, 1G A$g! 2015! -eb! =0 4ov! 2015!

.asterly, -illiam, and Sergio ebelo! I@iscal policy and economic growth!I Journal of monetary

economics =2!= 9177=:% ?1G>?5F!

@riedberg, achel /!, and Jennifer "$nt! Khe 'mpact of 'mmigrants on "ost o$ntry -ages,

.mployment and (rowthL! The Journal of Economic Perspectives7!2 91775:% 2=M??! -eb!!!

(erdtham, Ulf>(!, and Bengt JNnsson! I'nternational comparisons of health e*pendit$re% theory,

data and econometric analysis!I Handbook of health economics 1 92000:% 11>5=!

I"ow 't;s )ossible to )ay a 1006 a* ate!I The New York Times! he 4ew Eor8 imes, 02 @eb!

2012! -eb! 01 Dec! 2015!

3anda$, Daniel! K(overnment .*pendit$re and .conomic (rowth% A ross>co$ntry

St$dyL! outhern Economic Journal ?7!= 917F=:% GF=MG72! -eb!!!

-orld Development 'ndicators, he -orld Ban8


15/21

olantonio 1?

ables

Table 1 - S$mmary Statistics able% Economic !rowth and ocial Economic "ariables

Variable Metric Observations Mean Standard

Deviation

Minimum

Value

Maximum

Value!#P per

capita

!rowth

6 ann$al F15? 2!10? !5? >?!77G 1F2!=5

!ov$

E%penditure

on Edu$

6 of gov!

e*pendit$r

e

15G= 15!05= ?!77 =!22 ??!F

Health

E%penditure&

public

6 of gov!

e*pendit$r

e

=50? 11!201 ?!50? !077 =?!?12

ubsidies and

other

transfers

6 of

e*pense

17G =G!50F 21!051 !0F 70!F

Total Ta%

'ate

6 of

commercia

l profit

1G7= ?G!G= =G!=0? G!? ==7!1

Wage and

salaried

workers

6 of total

employed

2217 G0!G= 17!G=2 1!? 77!


16/21

olantonio 15

Table 2 - orrelation /atri*% Economic !rowth and ocial Economic "ariables

GDP per

capita

Growth

Gov.

Expenditure

on Edu.

Health

Expenditure,

public

Subsidies

and other

transfers

Total Tax

Rate

Wae and

salaried

wor!ers!#P per

capita

!rowth

1!0000

!ov$

E%penditure

on Edu$

0!00G5 1!0000

Health

E%penditure&

public

> 0!22?7 > 0!022 1!0000

ubsidies

and other transfers

> 0!125 > 0!?0=1 0!?11G 1!0000

Total Ta%

'ate

0!017F > 0!177? 0!0=5G 0!2G25 1!0000

Wage and

salaried

workers

> 0!2F? > 0!2=G1 0!?5G2 0!?G=2 > 0!01?1 1!0000


17/21

olantonio 1

Table 3 - /$ltilinear egression% @i*ed .ffects /odel

(1) (2) (3) (4) (5) ()

!ov$

E%penditure

on Edu$

> !07?OOO

9?!0F:

>!177

9!1F2:

Health

E%penditure&

public

> !0G=OOO

9!022:

!5GO

9!?21:

ubsidies and

other

transfers

!00FO

9!005:

>!00F

9!0?F:

Total Ta%

'ate

> !002

9!002:

!117OO

9!052:

Wage and salaried

workers

> !02=OOO9!005:

> !1?F9!15F:

onstant ?!0FOOO

9!?=F:

=!=75OOO

9!275:

2!110OOO

9!21=:

2!502OOO

9!1F0:

?!0=GOOO

9!=5?:

=!G11

910!0G:

bservations 1555 =?25 1720 1G2 21G =7F

)rob! P @ 0!000 0!001 0!0G0 0!=27 0!000 0!0GG

!d"# $ 2 0!012 0!00= 0!002 0!000 0!01? 0!0F7

4otes% ob$st standard errors reported! oefficient significance noted at levels 16, 56 and106 with OOO, OO, O!


18/21

olantonio 1G

Table 4 % /$ltilinear egression% @i*ed ime .ffects /odel

(1) (2) (3) (4) (5) ()

!ov$

E%penditure

on Edu$

> !07GOOO

9!025:

>!2FG

9!21:

Health

E%penditure&

public

> !0G2OOO

9!02=:

!==OO

9!=0G:

ubsidies and

other

transfers

!00

9!00?:

!00=

9!0=2:

Total Ta%

'ate

> !00?

9!002:

>!017

9!0=F:Wage and

salaried

workers

> !020OOO

9!005:

!000

9!117:

onstant ?!005OO

9!=57:

=!070OOO

9!511:

!=F0

91!01?:

=!F=5OOO

9!==1:

=!011OOO

9!15:

=!G11

910!0G:

bservations 1555 =?25 1720 1G2 21G =7F

)rob! P @ ! 0!000 0!000 0!000 0!000 0!000

!d"# $ 2 0!150 0!0?5 0!15 0!07G 0!170 0!0F7

4otes% ob$st standard errors reported! oefficient significance noted at levels 16, 56 and106 with OOO, OO, O!


19/21

olantonio 1F

Table 5 % /$ltilinear egression% andom .ffects /odel

(1) (2) (3) (4) (5) ()

!ov$

E%penditure

on Edu$

> !01G

9!0?1:

>!177

9!22:

Health

E%penditure&

public

> !02

9!05?:

!5GOO

9!2:

ubsidies and

other

transfers

!015

9!01?:

>!00F

9!0=F:

Total Ta%

'ate

> !002

9!00:

!117OO

9!01=:

Wage and

salaried

workers

> !0GOOO

9!021:

> !1?F

9!1?7:

onstant =!G25

92!=20:

5!?25OOO

91!F12:

G!2F5OOO

92!2=G:

5!71OOO

91!757:

!F=OOO

91!2==:

>5!22F

9F!2??:

4otes% ob$st standard errors reported! Data controlled by co$ntry! oefficient significancenoted at levels 16, 56 and 106 with OOO, OO, O!


20/21

olantonio 17

Appendi*

o$ntries in St$dy

Afghanistan ongo, ep! 'ndia /ongolia

Albania osta ica 'ndonesia /ontenegroAlgeria ote dQ'voire 'ran, 'slamic ep! /orocco

Andorra roatia 'ra# /o


21/21

olantonio 20

SomaliaSo$th AfricaSo$th S$dan

SpainSri 3an8aSt! itts and 4evisSt! 3$ciaSt! incent and (renadines

U

How Does Democratic Socialism Impact Economic Growth?

Documents