chapter 12. interaction between economic and social variables: the transformational growth matrix

7/28/2019 Chapter 12. Interaction between economic and social variables: the transformational growth matrix

1/93

12. Interaction between economic and social variables: the transformational growth matrix *

Edward Nell

12.1. INTRODUCTION

The object here is to explore the interactions between economic growth and change, on the one hand,and changes in social, environmental, political and population variables, on the other. Theseinteractions are likely to affect social and economic well-being, for better or worse, and very possiblyquite strongly. Economic Growth could be supported and enhanced or it could be undermined, evenreversed. The economy is embedded in society, after all; the object will be to show this in ameasurable way, and to show as precisely as possible why and how this matters.

Keeping the analysis simple, but still realistic, economies here will be assumed to be made up ofsectors industry, agriculture, services and social classes workers, landlords, business owners,bankers.1 There are economic interactions between sectors and social classes; analyzing these is part ofthe theory of growth and fluctuations. But we want to focus on a different kind of interaction, alsoinvolving both sectors and classes, but this time between economic advances and the resulting pressure

or development of social variables. When sectors and classes interact along the social dimensions, weshall argue, it is likely to change the patterns of development and it may lead to stagnation and failureto develop.2,3

12.2. THE TRANSFORMATIONAL GROWTH MATRIX

To examine these questions we must set up quite a general framework, in which we can study howdifferent kinds of economic change can interact in various ways with many aspects of the social order.But in what follows we

289


2/93

290 Institutional and social dynamics of growth anddistribution

will keep the discussion of the purely economic side as simpleas possible, allowing us to define and examine a wide range ofsocial, political and demographic variables. We will stick tostraightforward cases of economic growth resulting from capitalaccumulation and technological advances. At this point we will

present a full spectrum of non-economic variables, though laterwe will simplify this, too. Even so, we will see that theinteraction between the economic side of a country and the otheraspects of its social order turns out to be complicated and

important.To display the variables and their interactions in the most

general way, we will set them out as a matrix, in which thecoefficient will indicate the relationships. (Since we areexamining growth, we call this the Transformational GrowthMatrix.) First we will list and describe the variables to beconsidered. These will appear in the row and column vectors.Usually a variable will be a composite, made up of a number ofdifferent but related elements that tend to move together, so thatit can be represented by appropriate proxies.4 Measurementscales will have to be chosen or developed, and they willnecessarily be approximations. However, we will not deal with

the empirical issues here. Our aim is to present the approach,and outline the kinds of relationships that we can see betweeneconomic advances and social, political and demographicvariables.

12.2.1. Indicators for the Variables

First, the variables:

Economic Economic Growth (EconG): GNP levels and growth, level and

growth of employment, standard measures of sectoral

growth; human development index, for comparison withpurely economic.

Environmental issues Environment (EnvH): Pollution of soil, waste disposal air

pollution gases and particulate matter; acid rain; waste heatand energy discharge and the effects on climate; water quality


3/93

and availability.

Social issues Socialization of Adolescents (AdlSoc): preparing adolescents

for adult working life, contact between adolescents and non-parental adults, apprenticeship and other training programs,school drop-out rates, measures of substance abuse.

Education (Educ): literacy rates, school enrollments, educationspending as a percentage of government budget, of GNP,

primary, secondary, and college graduation rates, researchprograms, specialized training


4/93

Interaction between economic andsocial variables 291

programs,

news

pa

persand

books

pu

bl

ished,


5/93

library book loans, net schoolenrollment male and female(primary, secondary, tertiary),literacy, ratio of female tomale literacy.

Health: infant mortality,mothers condition inchildbirth, life expectancy,morbidity and prevalence ofdisease, access to hospitals,availability of medicines,hospitals, doctors per thousand population,

indicators of public health. Household Standard of Living

(HStndL): not only the levelof the average standard ofliving, but also its distribution

the degree of inequality.Under 5 infant mortality,shared household income ratio

between highest 20% andlowest 40%, population usingimproved drinking water andsewage facilities. Public

goods, crime rates by incomedistribution, radios andtelevisions per capita. Ginicoefficients will give anoverall picture; poverty ratesfor different poverty levels,and measures of concentrationof wealth and ownership ofland will portray the extremes.

Political Social Infrastructure (SocInf):

transportation,

communications, police andthe criminal justice system,

public infrastructure - roads,bridges, sewers and watersupply, harbours, airports,garbage and waste

management;

p

ublicadmini

stration,landmana


6/93

gement and zoning, militaryand defense spending.

Effective and DemocraticGovernment (EffGov):Representative and responsivegovernment, civil order, sizeof middle class, percentagevoting, effectiveness of thelegal system and the courts,training of the police, theworking of a free press,human rights and religiousfreedom.

Demographics Population (PopP): Size,

growth, age distribution,median age, birth and deathrates, natural reproductionrates, fertility, % under 15, %over 65, life expectancy at

birth. Sex ratios, if available.

PopP, AdlSoc and Educ togethergive us people along with theirskills, what they know and what

they can do. EconG and EnvHtell us what is being done, whatis being produced, who isworking, in what ways, and withwhat effects on the world inwhich the society is set. SocInfmeasures what the state and thesociety as a whole provides as aframework to undergird andregulate social life and economicactivity. Health and HStndL showhow what is being produced

benefits the people, in bothpublic and private dimensions. Itgives us a measure of the benefits

being delivered and theirdistribution. And finally EffGovtells us how order is maintainedand disputes resolved.


7/93


12.2.2. The Matrix

The object here is to show how EconG interacts with the rest;we argue that this will be the key to successful development.Accordingly the top row will show how EconG depends on eachof the other variables, and the first column will give the impactof EconG on the other variables. For example, EconG may bestimulated or supported by EconG itself or it may not. It willnormally benefit from improvement in the EnvH, as well as the

four social variables, AdlSoc, Educ, Health, and HouseholdStndL. Improvements in the Political variables, SocInf andEffGov can also be expected to benefit EconG.

The next line below shows EnvH as depending on EconG,EnvH itself, AdlSoc, Educ, Health, StndL, SocInf, EffGov, andPopP; below that we will have AdlSoc depending on each of theothers, and so on, for Educ, Health SocInf, HStndL, EffGov, andPopP, respectively.

Each depends on the others, and the interdependence can bepositive, negative, or zero. And, of course, the dependence canchange, either as the result of the pressures of globalization, thechanges brought about by development, or by policy, or as a mix

of all these.The bottom row of the following table shows the way PopPdepends on the other variables. Higher EconG, for example,could go either way. By bringing in improved technology, itwould improve medicine, and lower death rates, so increasingPopP. This is the likely short-term response. But by bringingincreased economic opportunities, especially for women, itcould lead to lower fertility, reducing PopP, though it may taketime for this to emerge. In the same way, a better EnvH will leadto fewer deaths from respiratory disease and foul water, but ahealthy environment might also lead to more chances forenjoyable leisure, reducing fertility. Good AdlSoc should lead tofewer unwanted pregnancies and smaller families, representingdeliberate choices, thus lowering fertility. A higher standard ofliving, at least initially, increases population pressure, because itreduces deaths. But in the long run, a higher HStndL will lead toreduced fertility as women take advantage of greateropportunities with confidence that their smaller families will


8/93

survive and live healthier lives. A positive relationship betweenthese economic variables and PopP can be seen in manydeveloping countries today, reflecting the fact that better dietand public health will increase longevity and lowers childhoodmortality, while the long run effects have yet to showthemselves.

The Matrix is not designed to make general statements; aseparate Matrix must be developed for each country, and it can

be expected to change over time. Indeed, these changes willmark the development of the country. Each


9/93


casemust

beinvestiga

tedem

pirically,andtheinvestigationmust

beupdate

dregularly.

Yet


10/93

when we assemble the data in theright way, some surprisinggeneral patterns can be seen. Theresults of empirical study can bewritten as the set of equations

below, arranged in a Table. Onthe Left in a column we have theamounts of each of the variables;they are each set equal to thecombined effects on them of theothers, shown on the right. The

plus or minus mark in each cellindicates whether the variable on

the left is related positively ornegatively to the variable at thetop of each column. Strongrelationships are indicated in

boldface. (The particularconfiguration shown here reflectsa stylized summary of some ofthe problems developingeconomies face now. Later wewill consider the way thesevariables are related in thedeveloped countries.) All the

variables are endogenous,although most will also have anexogenous component.

The matrix developed hereshould make it possible to see

patterns. An advantage of ourapproach is that these can beseen even when we are not ableto specify the relationshipsnumerically. Sometimes all wecan say is that variables arerelated positively, negatively, ornot at all. In such a case all wecan enter are +, , and 0 inthe appropriate cells of thematrix. But even with thislimited information, not only canwe identify significant causallinkages, we see how they

interactandfeed

backonone

another.Asweseehere,si

mplevisualinspection

nomathematics


11/93

needed can sometimes tell uswhether the interactions will endup in a vicious or a virtuouscycle.

As a first approximation, therelationships can be taken aslinear5. The justification is thatthe changes in the variables will

be small.

Table 12.1. Signs of the coefficients

EconG ENV AdlSoc

Educ Health StndLSocInf EffGov PopP

+ + + + +EnvH +or + + + ?+AdlSo +or 0 + + +Educ + + + + + +Health +or + + + + +StndL +? + + + +SocInf +or + ?+ ? + ?EffGo +? + + + + +Po P +or +or - - +or- +or

[There should also be aconstant term in each equation,to capture the influences notaccounted for by the statedvariables. 6 This will beimportant later.]


12/93


If the impacts shown are all positive, and if each impact isinterpreted to represent a fraction of the total change of thevariable, then basing the system on suitable initial conditions, itmight be reasonable to expect a unique solution. That is, solvingthe equations would tell us how large, relative to each other, thevariables must be. Since such a solution will determine relativeamounts, it would be possible to use this system to examine theimpact of a policy change which is, indeed, the main point. If,for example, the degree of change in EconG were imposed from

outside, the system would allow us to calculate the effects on allthe other variables.7 [Note that the case where the coefficients ina row or column sum to unity is especially useful; it gives rise toMarkov chains in which successive adjustments converge to aunique solution.]

With nine equations and nine variables we have eighty-onerelationships to consider or seventy-two, if we ignore thequestion of self-reinforcing relationships. It can all be presentedneatly, as a Matrix, with the nine variables arranged in rows andcolumns. The first row shows for example, how EconG dependson the other variables, while the first column shows how theother variables depend on EconG. The second row shows how

EnvH depends on the other variables, and the second columnshows how the others depend onEnvH. The third row andcolumn then show AdlSoc, the fourth Educ, the fifth, Health,and so on, finishing with PopP.

Table 12.2. Matrix of coefficient signs

EconG EnvH AdlSoc Educ Health StndL SocInf

EffGov PopP

Eco ----- + + + + + + +Env + ------ + + ?+ + + +

Adl +or + ----- + + + + +

Edu ?+ ?+ + + + + +

Hea -?+ + + + ------ + + +

Stn +or + + + + ------ + +

Soc +or + + ?+ ? + ------ +


13/93

Eff +? + + + + + + - - - -Pop +,- +,- ? - - - + - -

Lets look at the variables and how they might relate to oneanother in various economic conditions, taking them insuccession. Important cases include countries that depend onPrimary Exports, Dualistic economies, economies trying toindustrialize using out-dated technologies and second-


14/93


handequi

pment,and

economiesstillrelyingontraditionalandcrafttechnologies

.

EconomicsReadin


15/93

g across the top row shows howEconomic Growth depends onthe other variables. Theserelationships will tend to hold inall of the cases. An improvedEnvironment will tend to supportEconomic Growth; it should leadto more productive agricultureand to a healthier labor force.Better Socialization of Adolescents will mean a moreskilled workforce. Both improvedEducation and better public and

private Health measures willraise productivity and reduce sickdays. A better SocialInfrastructure will translate into amore secure workforce, while ahigher Household Standard ofLiving provides larger marketsand better informed consumers,and promotes better health. Italso supports but does notguarantee better socialization ofadolescents. A more Democratic

Government helps to ensure astable civil order and a reliableand uncorrupt system of justice.Finally, economic Growth will bemade more difficult byPopulation pressure,

And it works in reverse; just asimprovements in these variables

promote EconG, declines orreductions will hamper economicadvance. If adolescents are not

properly socialized, they will notfit into an urban work force, andthey will not function well ascitizens. This will be reflected inweaker productivity and adecline in the HouseholdStandard of Living, as relativeswill be under pressure to support

them.Pooreducationandhealt

hwillonlycom

poundthes

edifficulties.Allthiswill

put

pressureon


16/93

social infrastructure, and willcreate tensions within democraticinstitutions.

The EnvironmentThe Environment will very likelycome under pressure from EconGin two cases, Primary Exportingand Dualistic economies, butmay not be so pressured in othercases. Faster growth tends togenerate increased pollution andcontributes to global warming,

especially in Primary Exportingand Dualistic economies.Pollution is, famously, anexternality, that is, a cost whichneither buyer nor seller pay,unless forced to by regulation orlaw. This is unlikely in weaklyintegrated systems. But fastergrowth also provides theresources and the newtechnologies to clean up and

preserve the environment,although this may be difficult in

economies undergoing OutdatedIndustrialization (think of Peru orArgentina in the 1960s and 70s,or parts of India even today).Prosperity also elevates theregard placed on a sustainableand healthy environment. Forthese reasons some have arguedthat starting from a low level,Economic Growth will first tendto worsen the Environment, butat higher levels further economic

growth will improve theenvironment. This is sometimescalled an EnvironmentalKuznets Curve analogous toKuznets original hypothesis thathistorically economic growth


17/93


first worsened, then improved the distribution of income. Theevidence, however, is inconclusive, both in regard to theenvironment and for the original claim. As for the othervariables, better Education and Health probably have littleeffect; at best they may contribute marginally to a betterEnvironment. But better Social Infrastructure, a higher Standardof Living and greater Democracy will all tend to encourageimprovement in the Environment. On the other hand, in general,when Population increases the Environment will tend to suffer.

Social variables The Socialization of Adolescents will tend to altered by

EconG in virtually any developing economy. Higher EconGwill lead to an outflow of families from rural areas into thecities, where they will lose the traditional support ofneighborhood and kin. The older generation will lack theskills and knowledge to provide guidance to the young. InPrimary Exporting and Dualistic economies EconG may notcreate many new opportunities; youth unemployment may

become a major social issue. On the other hand in Small-Scale Craft and Industrializing economies, higher growth will

provide the both tax revenues and employment opportunities

in the cities, making it possible to offer education andtraining to the young. The effect of the other variables isstraightforward, except, perhaps, in one case. HigherPopulation growth will put a strain on Socialization, makingit more difficult, but a better Environment and improvedSocial Infrastructure, a higher Standard of Living and a moreDemocratic Government all tend to encourage bettersocialization of the young. Better Education will certainlyimprove AdlSoc, and better Health is likely to. A warning,however: in the short run improved Education could lead to ageneration gap between youth and parents. This is astandard theme in Bollywood movies, and is evident even in

todays Ireland. Education: Economic Growth can go either way; growth canpromote education by demanding greater skills and a higherquality of labor. But it could also draw children and theunskilled into low-level long-hour jobs, keeping them out ofschool. Which effect it has depends on the kind of growth

being promoted. In general the positive result is more likely


18/93

in the strongly integrated economies, the negative in theweakly integrated ones. Better AdlSoc will make it easier;improved social infrastructure and a higher standard of livingwill tend to improve education. A more DemocraticGovernment will strengthen it. But a cleaner and betterEnvironment, by itself, may not have much impact one wayor the other on education although people may learn a gooddeal by cleaning up! Health strongly and positively impactseducation better health means better learning. Note thechain of positive linkages: a better Env leads to


19/93


betterHea

lth,whichleadsto

bette

rlear


20/93

ning, which, in turn, is likelyto promote a better Env!These are the kinds of connections to look for. A risein Pop pressure, on the otherhand, will generally create

problems for education. Health: In the well-integrated

economies, EconG willprovide more resources; andemployers will want a healthylabor force, so Health shouldimprove; but even in these

economies, and certainly inPrimary Exporting andDualistic ones, economicexpansion might create toomuch disruption, leading tocrowding and congestion thatwill reduce the quality andavailability of health care.AdlSoc and Education willimprove it, the latter quitestrongly. Better SocInf and ahigher StandL will both

improve health, and greaterEffGov will open the way forpeople to demand better healthmeasures. Pop pressure canreduce health.

The Household Standard ofLiving should improve withadvances in EconG. However,Economic Growth can lead tosuch an intense concentrationof income and wealth in thehands of a small group thateveryone else is actually leftworse off. It will also tend to

be improved by increases inthe other variables, but it willcome under pressure fromincreases in Population.

Politicalvariables

TheSoci

alInfrastru

cturewill

be

putu


21/93

nder pressure by increasedEconG, but at the same time,growth will bring increasedresources. In the weaklyintegrated economies, theformer effect is likely to

predominate, in the stronglyintegrated the latter.Population will press on theInfrastructure, but it will tendto be strengthened as a resultof increases or improvementsin all the other variables.

Effective /DemocraticGovernment should beencouraged by EconG, butthings are not so simple; therecan be difficulties. Well-

balanced economicdevelopment, encouraging thegrowth of a middle class, as inthe integrated economies, willcertainly be supportive ofdemocracy, (think of India orChile today). But there are

other patterns of economicdevelopment. The kind ofEconomic Growth that resultsin a high concentration ofincome and wealth, or thatfocuses on an extractiveindustry, oil or minerals, or a

plantation crop, may not be socongenial. Indeed, famously,such an economic structurefavors an authoritarian ordictatorial government. As forthe other variables, EffGovwill be made more difficult byPopulation pressure, but willnormally be encouraged byimprovements in all the othervariables.

Demographics

Population

pressurede

pendson

bot

hbirthsanddeaths.Economic

andsocialim

provementswillgenerallyreduc


22/93

e deaths, thus increasing Popinitially, but


23/93

298 Institutional and social dynamics of growth and distribution

in the long run such improvements will lead to a decline infamily size. Reading across the first row, we see:

that the effects of Economic Growth and an improvedEnvironment will tend to go both ways, in the short run

probably increasing Pop, but in the long run and ultimatelyreducing it;

that Population pressure will tend to be reduced by improvedSocialization of Adolescents (fewer out-of-wedlock births,later marriages);

that Educ will tend to lower Pop, through knowledge of birthcontrol and increased awareness of opportunities for women that improved Health will reduce deaths, which will tend in

the short run to increase Pop, but as better Health decreasesinfant mortality and childhood deaths, fewer births will beneeded to ensure surviving children and fertility will tend tofall

that better Social Infrastructure will tend to reducePopulation pressure (by providing for old age and sickness,reducing the need to have large families to assure care in oldage;

that a higher Standard of Living will increase PopP initiallybecause it will reduce mortality and deaths during childbirth -but in the long run it will lead to lower fertility;

that Democratic Government will probably reduce PopP- forexample, by offering more opportunities for women, whowill then be inclined to have fewer chidlren.

12.2.3. Analysis

Just by glancing at the matrix we can see that the largest numberof +or- and ? cases are to be found under Population andEconomic Growth.

Consider PopP, reading along the bottom row: under present-day conditions it is likely that PopP will be increased by

improved economic growth, a better environment and a higherstandard of living. As a result, PopP itself (now read down thevertical column) will react back negatively not only on thosevariables, but also on the others. Greater PopP will make itharder to socialize adolescents, will create crowding andcongestion, hindering education and economic progress, causing


24/93

health problems and environmental stresses; moreover, it tendsto overtax the social infrastructure, bid wages down andotherwise lower the standard of living, and so make democraticgovernment more difficult. That is to say, an economic advancecould undermine itself, by generating population pressures thatwill undo it. This is the modern version of Malthus, and theIron Law of Wages, where wage gains lead to populationincreases which


25/93


eventually

pressurewa

gesbackdowntosu

bsistencelevels(Clark,2007).Ifther

eisto

bedeve


26/93

lopment, PopP must becontrolled; (think of incentives tohave fewer children in China.)

On the other hand, it works inreverse, too. If Population

pressure can be reduced, all theother variables, both social andeconomic, will be improved. Sowe need to develop policies toensure that increases in theeconomic variables will notincrease Population pressure.(But PopP control must not go

too far; France and Italy arefacing an aging population that isdeclining in numbers. That is onereason for permitting sociallytroublesome immigration.)

Now look at the column underEconG. This shows the effectsof a rise in Economic growth,with greater investment in newtechnologies raising productivity,developing new sectors and

bringing increased employment.

These effects increase theresources available to society.This should lead toimprovements in the othervariables. But economicinvestment does not necessarily

bring general economic andsocial improvement. On thecontrary, it will always destroy aswell as create, and, especially inweakly integrated economies, itcan cause terrible problems -destroying the old ways of life(as corporate farming does, forexample), driving people off theland in search for work, therebyundermining the way the youngare socialized. New resources arecreated, but in a weakly

integratedeconomy,thosew

hofeelthedestructiveim

pa

ctmayhavenoaccesstothenewresources.M


27/93

oreover, as we have seen,economic decisions respond tothe market; pollution costs areexternal (costs paid by neither

buyer nor seller), so unregulatedindustrialization normally leadsto widespread environmentaldamage, especially air and water

pollution. If the new industriesare unregulated they will oftenhave unsafe or unhealthyworking conditions, creating newhealth hazards. We have already

discussed how the migration oflabor that takes place in order tosupply the new industries withworkers may result in haphazardhousing developments, withunhealthy sewage and wastedisposal, and contaminated water.These shifts of population canoverburden school systemsalready under pressure.Economic development generallyleads to at least some expansion

of social infrastructure, but theexpansion is not necessarilyproportional, so the institutionsmay become overburdened.Finally, economic growth neednot necessarily lead to a rise inthe general standard of living; ifthe benefits are concentrated andthe costs spread widely, most

people could end up worse off.It might seem that these are the

relationships we would expect tofind in a developed economy. Butthat is not quite so. Certainly in adeveloped economy theEconomic and the Social,Environmental and othervariables do support each other.But in a fully developed

economythesu

pportofthe

socialandotherrelationshi

pswouldtendto

beinde

pendentofthecurre


28/93

nt


29/93


EffG + + + + + + + -------- -

PopP - - - - - - - - ------

changes in the economy, certainly in the short run. That supportwould not vary with the current level of EconG. The Soc, Envand other variables would certainly tend to support economicdevelopment both in the short and the long run, but the positivedevelopment of those variables would be independent of any butthe largest fluctuations in the economy. This of course is made

possible by the fact that a developed economy has a high levelof productivity, and so can support social, environmental, etc.,development without regard to current variations in theeconomy. More on this point later.

Table 12.3. Matrix of ideal coefficient signs

EconG EnvH AdlSoc Educ Health StndL SocInf

EffGov PopP

Eco ---- + + + + + + +Env + ------ + + + + + +Adl + + ----- + + + + +

Edu + + + + + + +

Heal + + + + ------ + + +

Stnd + + + + + ------- + +

SocI + + + + + + ------ +

Some variables will be fast-acting, others will exert theirinfluence only over longer stretches of time. Economic growthwill act quickly, as will the Standard of Living; in both cases thechanges will be felt through markets. But changes in SocialInfrastructure will be felt a little more slowly, while the impactof changes in the Socialization of Adolescents, Population

Pressure and Democratic Government will take a long time todevelop.The Population variable is critically important because it has a

strong long-run impact on the others. An increase in PopP willworsen all the rest, and they will in turn react on each other.


30/93

PopP can change rapidly, if the death rate falls or rises, but thechanges are slower when the change is due to changes infertility.

The Environment/Economic relationship need not be positive(+); indeed, in present conditions it appears to be negative thatis, higher Economic Growth leads to a worsening of theEnvironment. The worsening of the Environment then leads tolower performance in many of the other variables, ultimatelyreacting back to reduce or impair EconG.


31/93


Butif

policiescould

be

develo

pedtoensurethatEconomicGrowthledtoa

be

tterEnvironm


32/93

ent, this could be cruciallyimportant. For then the betterEnvironment would improve allthe rest, and, as we have seen, itis fairly reasonable to think thatthe other variables will be

positively related to each other(+). Then, assuming that

population growth will dependnegatively () on the others, a

positive Environment/Economicrelationship will tend to bringabout a pattern in which the

growth of each of the otherfactors reinforces the rest in a

positive (constructive) spiral. Wewill explore these relationships ina moment.

12.2.4. Sustainable Globalization

At this point we can derive aninteresting implication of ourapproach, one that provides animportant contribution to the

current debate over globalization.Opening up new areas of trade,building new factories,outsourcing and flows of hotmoney the processes of globalization, in short all tendto generate rural-to-urbanmigration; they call for workersto develop new skills; theyrequire additional infrastructure,they generate pollution, createneeds for new public healthmeasures, etc. In general theylead to social changes. But thesesocial changes react back on theglobally driven economic

processes. As we have seen thisraises the question whether thisinteraction is sustainable, that is,

whether,givenitsim

pactonde

velo

pingsocieties(and

perh

apsalsoontheadvancedworld)the

proces


33/93

s will support itself, and will becarried on, or whether it willgenerate counter forces that willtend to bring it to an end.

Using the matrix we candefine a precise condition thatshows what is necessary fordevelopment to be sustained,answering the question, whenand whether the development

process, driven by globalization,can be sustained. Of course, to dothis the equations have to be

written out with actual numbers,even if they are only estimatednumbers. For this purpose weshould treat PopP separately,since the issues there are on amuch longer time scale. Weshould also separate outEffGovov. (It might also beuseful to examine Env on itsown, but it could be included.)Then, given suitable conditions,the remaining equations can be

solved, and we haveeconomic impact on society =

social impact on the economy

This holds in equilibrium, andit says that when the equationsare in balance, the economyscontribution to the growth of thesocial variables just balances thesocietys contribution to thegrowth of the economy. Eachreinforces the other. This is the

condition for sustainabledevelopment. When there is aneconomic advance it will changethe social variables in the samedirection; but the social variableswill act reciprocally back on the


34/93


economic realm, also moving it forward, and the two effects willbe just the right magnitude to keep the balance between the tworealms. (Nell, Mayor and Errouaki, 2008) Economic advanceleads to better education, improved public health and moreexpansive housing (higher StndL). But, in turn, the improvededucation, public health and housing have arrived at just theright level (not too much, not too little) to support that degree ofeconomic advance.

If this condition does not hold, we do not have an

equilibrium; either society will be contributing more to theeconomy than it receives back, or vice versa. Of course, anequilibrium balance can only be defined if it is possible tomeasure the impacts, so that we can write equations. But evenwhen measurement is not possible we can see that if theelements of both rows and columns are positive, the economicimpact on society and the social impact on the economyreinforce one another. They may not be in exact balance without numbers we have no way of knowing - but we can saythat they work in the same direction.

Look back at the earlier version of the matrix, where manyelements in the EconG column were negative, but all the

elements in the row were positive. Under those conditions theeconomic impact on society not only could not possibly be inbalance with the social impact on the economy, the two couldnot possibly be mutually reinforcing. One will be likely toundermine the other, or worst of all, each will tend to underminethe other. The Econ variable will depend positively on goodeducation, on effective public health, on a clean environment;

but economic advances will put schools under strain, will createcrowding and disease in the cities, while polluting theenvironment. These deteriorating conditions will thenundermine the economic advance. Something very like thishappened throughout Africa following independence.

12.3. DYNAMICS AND DEVELOPMENT TRAPS

To carry the argument further, it will be useful, first to simplifyour approach, and then to explore the dynamics the patterns ofinteraction more fully.


35/93

So we will simplify by looking only at the interactionsbetween certain parts of the matrix, holding everything elseconstant, rather like the partial equilibrium analysis ineconomics. (Our approach will be open to some of the samecriticisms as partial equilibrium, but it allows us to isolate the

problems peculiar to each of the important sets of variables.) Wewill combine the social variables into one composite, and look atthe interaction of Soc and Econ.8 Then we can hold the Socrelationships constant, and


36/93


examinetheinteractio

nofEconwithEnv,withEffGov,andwithPo

pl.Ineachcase

wewillex

pl


37/93

ore how these interactions canturn into various kinds of viciouscircles and development traps,modeling the interactions interms of simple stabilitydynamics.9 Such modeling willhelp us to see exactly whichaspects of the interactions cause

problems (subject, of course, toour simplifying assumptions.)Having identified problematicaspects and hopefully havingclarified the dynamics, we might

be able to suggest policies.The various models below

basically partition the matrix intosimplified components Econand Soc, Econ and EffGov, Econand PopP. Soc, for example,will be an aggregate of thevariables of the matrix, such asEduc, Health, SocInf and AdlSoc.In each case we take Econ as itappears in the matrix, and theneither treat the other elements as

constant, and explore the relationbetween Econ and anothervariable, e.g. EffGov or PopP, orcombine several variablestogether into a compositevariable, e.g. Soc, made up bycombining the Social Variables.Then we show how thesecomponents of the matrix caninteract so as to lead todevelopment traps.

By a development trap wemean a set of relationships inwhich important variables arerelated in such a way that anyadvance forward sets upoffsetting movements in othervariables that undermine that

advance.Forexam

ple,inAf

rica,andinBangladeshsom

eyears

back,aidandeconomicgrowt


38/93

h improved health - which led toa population explosion, puttingadditional pressure on land andwater, whereupon healthdeteriorated again. Cf J. Sachs.

Development traps areespecially serious when theconfiguration of the matrix isunfavorable, but it is important tosee that traps can emerge evenwhen the overall configuration isfavorable. Of course, a favorableconfiguration makes it much

easier to devise policies to avoidthe traps.

Lets now explore thedynamics. This is a difficultsubject in economics, but it isincreasingly apparent that it isthe key to many of the mostdifficult problems in the field(financial instability,unemployment, inflation,

business cycles, etc.) and in ourcase here, it is the key to

understanding how globalizationimpacts on development. But atthis point we have to draw onsome mathematics. We arelooking at how incentives leadagents in the different sectors tointeract, and these interactionsare channeled by the structure ofthe system in ways that can leadto outcomes nobody expects and sometimes nobody wants!

We will start with Econ andSoc, where the latter is acomposite made up of Educ,Health, AdlSoc and SocInf. Themain question is, will theincentives created by theinteraction of Econ and Soc tend

toleadthesocietytothe

posit

ionwherethetwomutua

llysupportoneanother?Orwillthe


39/93


incentives drive the system away from that point, somethingnobody wants! Or perhaps the interaction will just leave thingsstagnant?

12.3.1. A Simple Two Equation Example

The coefficients of the matrix represent the slope of an impliedlinear function connecting the two variables; thus the coefficientshowing the impact of EconG on Health is the slope of a linear

function showing how Health varies as EconG increases ordecreases. The coefficient showing the effect of Health onEconG is similarly the slope of the function showing howEconG can be expected to vary as Health gets better or worse.Our first example will examine the interaction of the variablesgoverned by these linear functions.

To simplify, lets leave out the environment and democraticpolitics, and take the social variables as a group, represented bya simple index, Soc. Then we can illustrate the general

principles with a simple case in which there are just twoequations and two unknowns:

Econ = S(Soc) and

Soc =E(Econ),

where S( ) and E( ) are the two functions. The first says thatvarious levels of Soc support or, in a stronger interpretation,generate or help to generate corresponding levels of Econ,where higher levels of Soc lead to higher levels of Econ. Thesecond says that various levels of Econ generate correspondinglevels of Soc. But in this case, Econ could either damage orsupport Soc. If it damages it, then the effect will be negative,that is, the impact of Econ will lower the level of Soc; if itsupports it, the effect will be positive, so the impact of Econ willraise Soc.

Lets first consider the linear case with one equation negative,one positive. We can write our equations:

Soc =AB(Econ), and

Econ = C+D(Soc)


40/93

B and D are the coefficients drawn from the matrix, sayinghow Econ affects Soc, and how Soc affects Econ. Note that

besides the coefficients, we must consider constant terms (whichare not part of the matrix); these terms describe what Soc orEcon would be in the absence of effects from the other. (Theequations have intercepts, as well as slopes.) The first equationsays that increases in Econ will bring reductions in Soc. Thesecond says that increases (declines) in Soc will bring increases(declines) in Econ. If there


41/93


aretwoequations,one

witha

positiveandonewithanegativeslo

pe,thereha

sto

bean


42/93

intersection. Of course it mightbe negative! But it exists. Onerelation is positive upwardsloping and one negative,downward sloping; so there is anintersection somewhere. If it is in

positive space, it can beconsidered an equilibrium, thatis, the social and the economicwill be mutually supportive.There could equally well be anintersection where one or both ofthe variables was negative; in this

case the relationship would bedestructive rather thansupportive. (In ordinary languagethis says that these variablesaffect one another, in waysdescribed by the equations, andthat there is a point where theimpact of Econ on Soc just

balances the opposite impact ofSoc on Econ.)

Soc

A C E

c

o

n

Figure12.1.

Intersection

betweenrequiredand

generated

Socas

functionsof

Econ

S

olvingfor

Soc Re uired


43/93

the level of Econ, we substitutethe first equation in the secondand rearrange:

Econ = [C+DA]/{1 +DB}

Now we can see that if thenegative relationship were to turn

positive, the intersection wouldbe at a much higher level of bothEcon and Soc. In that case:

Econ = [C+DA]/{1

DB}, which is clearly

greater.


44/93


A Soc

C

Soc

Required

Soc

Generated

Econ

Figure 12.2. Another form of the intersection betweenrequired and generated Soc

This is illustrated on the diagram. But it should be intuitivelyobvious also. If Econ has a weaker negative effect on Soc, the

line will be flatter; if it has no effect the line will be horizontal.But as the downward sloping line swings up, the intersectionwill be further out, i.e. at a higher level of Econ. Note that it is

possible for the Soc generated by Econ to always lie above theSoc needed line (in the positive quadrant); in such a case itwould never be a constraint on economic advance but this isnot likely]. This would be the case ifDB = 1; ifDB > 1, thiswould also be true, but there would be an intersection innegative space.

Next suppose the society is not in that equilibrium lets saythat the level of Econ is a little less than the intersection orequilibrium level. Can we expect social or market forces to drivethe system back to the balancing level? Will there be appropriate

incentives?Lets recall what the equilibrium means. It is the level at

which the Soc generated by Econ would be just equal to the Socrequired to support Econ. This is an important relationship, inregard to being able to maintain a program of economicexpansion, whether market driven or planned. If the level of Soc


45/93

generated or supported by a particular level of Econ is equal toor greater than the level required to support that Econ, then thateconomic activity can continue or move ahead. But if the Socgenerated is less than the level needed to support that rate ofEcon advance, the Econ will have to be cut back. But we startedfrom the assumption that the level of Econ was just


46/93


alittlelowerthanthe

mutuallysu

pportinglevel.IfEconiscut

backthenthesy

stemismovin


47/93

g away from equilibrium therelationships are working

perversely.This can be seen most easily

looking at a diagram. There aretwo lines, one representing Socgenerated by Econ, the other Socneeded by Econ, with Econ onthe horizontal axis, Soc on thevertical. We assume that bothlines are positive, and that theyintersect in the positive quadrant.But where are the intercepts, and

which line is steeper? In Figure12.2 the Soc generated lies abovethe Soc required at all levels ofEcon below the intersection.Since this is so, Econ can easily

be increased, so the societywould be likely to move to theequilibrium. Above theequilibrium, Soc required will begreater than that generated, soEcon will tend to fall back. Butsuppose, as in Figure 12.3, that

Soc required is high even whenEcon is minimal, but rises slowly,while Soc Generated is very lowat low levels of Econ, but risesrapidly.

Soc

EconFigure12.3.

Unstab

le

interac

tion

betwee

n Soc

and

Econ

ThenatalllowlevelsofEcon

,Socrequiredwillexceed

Socgenerated,

Soc

Generated


48/93

resulting in a development trap!But at levels of Econ above the

balancing point, Soc generatedwill be higher than what isrequired, and the system canexpand freely the balancing

point is a takeoff position. Thesystem will find it difficult toreach or stay in its equilibrium,even though this equilibrium iswell defined.

Now we can set out a simpleexample of what we have been

saying all along. When Econ hasthe wrong kind of impact on Soc,the effects will react back andundermine Econ. This is avicious circle. To see exactly howthe


49/93


undermining (or, alternatively, the supportive process) works,as we have been discussing it, we need to express the impacts ofthe variables in a period analysis. That is, we divide up timeaccording to how the effects of the variables on each other workout. Mathematically, that means we set it out in Differenceequations. These are equations in which the variables aredated, which means that the values of the variables are thevalues that they hold in certain specific time periods.

Here we assume that Econ will play itself out in period 0; but

its impact will be felt on Soc in the next period, period 1.However, the effects of Soc on Econ will all take place in period1. So we have:

Soc1 =E(Econ0) and

Econ1 = S(Soc1)

The first says that impact of Econ on Soc takes time; theimpact of Econ now will be felt in the following periods Soc.(Econ growth will shift population from the countryside to thecities, and next period the schools will be crowded and healthfacilities overrun. Think of Mexico City, or Cairo, or Sao

Paulo.) The second equation says that the support of Soc forEcon is needed currently Soc now affects Econ now. (Todayshealth facilities keep todays workers healthy; todays roads and

bridges move todays goods and services.) Then substituting, wehave

Econ1 = S(E(Econ0)),

which will tell us how the system will evolve over time.Lets go back to our example, this time writing it as a first-

order linear difference system:

Econ1 = C+DADB(Econ0).

IfDB(Econ0 ) < C+ DA then Econ1 > 1; but ifDB(Econ0 ) >C+DA, then Econ1 < 1. As this suggests, the system alternatesaround the equilibrium; it will converge, however, only if theSoc line is steeper than the Econ line (Allen, 1968, pp. 813).On the other hand, if the relationship is positive, this becomes

Econ1 = C+DA +DB(Econ0)),


50/93

which clearly increases indefinitely. This would be self-sustaining growth.

A non-linear caseThe matrix gives us fixed coefficients, implying that theunderlying relationships are linear. But we do not have to staywith that assumption, if there are good reasons to think thevariables are related in a more


51/93


com

plicatedway.Theim

portantthingistoconsider

plausi

blerelationships,while

keepingthe


52/93

analysis simple enough that wecan visualize the dynamics.

We should be careful,however, when interpreting thesefunctions they are being used toexamine dynamics, and we areassuming that they are reliable,grounded in contracts,obligations or social mores. Yetthey may not be well grounded,or the grounds may be changing;it is entirely possible that some ofthe relationships we are

considering may shiftunexpectedly with social andcultural changes. We have arguedthat treating them asmathematical functions will beuseful; it will show us the various

possible patterns of interactionover time. But it should beremembered that, at times, thiswill be a stretch. In particular, itwould be unwise to assume thatthese relationships are always

reversible. That is, if the systemmoved along one of the functionsfrom A to B, it could reverseitself and return from B to A.Assuming reversibility istantamount to holding that timedoes not matter, an issue hotlydebated by economists. But iftime matters, when reversing,things may end up at a pointdifferent from A. Nevertheless,lets look further.

A plausible non-linear case:the relationships might both beones that increased slowly atfirst, then rapidly, then slowlyagain. That is, the dependentvariable rises slowly, but at anaccelerating rate, then rises

rapidly,

butdecelerates,until

itisincreasingonlyslowly

again,andthenflattensout.(Furtherincrea


53/93

se of the independent variablewill have no effect on thedependent.) These are known assigmoid curves. Suppose thisdescribed the effect of Econ onSoc as Econ increased, movingalong the horizontal axis in thediagram, the Soc which eachlevel of Econ could support,would first increase slowly, thenrise rapidly, and finally slowdown again. But the effect of Socon Econ would also show the

same form. As Soc increased,moving along the vertical axis,the Econ it could support wouldfirst rise slowly, then morerapidly, and then slowly again.

This may be hard to graspintuitively, but can easily bevisualized; the diagram plots thecurves with Soc on the verticalaxis and Econ on the horizontal.The two curves start from theorigin and intersect twice the

second intersection marking thepoint beyond which increases inthe independent variables haveno further effects on thedependent ones. Below the firstintersection, reading from thehorizontal axis, the curveshowing Econsupported/generated by Soc liesabove the curve showing the Socsupported/generated by Econ.After that point, it is just theother way around. The curveshowing Econ supported by Soclies below the curve showing Socsupported by Econ.

Lets interpret this. A level ofEcon is only viable can only bemaintained if the level of Soc

thatitsu

pportsisaslarge,or

larger,thanthe


54/93


level of Soc that it needs (i.e. the level that is needed to supportit). As Econ increases from a very low level, the Soc that itgenerates or supports will be low at first, and then rise rapidly,while low levels of Soc, will only support low levels of Econ.This means we have a development trap here: at low levels ofEcon, the Soc generated or supported will be less than the Socrequired to support those levels of Econ. Since the system startsout poor, it is caught; by itself it can never get going, eventhough if it did, it would reach a point where it would begin a

pattern of self-sustaining upward movement towards a high-level position of mutual support. . ]

Figure 12.4. Development trap and development support

Another version of the non-linear relationshipThis time lets consider not the levels of Econ and Soc, but theirrates of growth; lets also include Env along with Soc, andassume they move together.10 Both are negatively impacted byEcon growth. But in the absence of pressure from Econ, SocEnvwill grow on its own. When Econ growth speeds up, rural-to-urban migration increases, putting pressure on education and


55/93

health, and overstressing social infrastructure. In addition,pollution increases and environmental damage rises. HenceSocEnv will stop growing or even decline. But a stagnant ordepleted SocEnv will then, after a time lag, react back oneconomic growth, slowing it down. But slower economicgrowth, in turn will reduce rural-to-urban migration. But withslower growth


56/93


andlowermigration,

thesocialsystemwill

beabletoabsorbtheearliermigran

tsandcleanup


57/93

the environmental damage;SocEnv will resume its growth.Health and education willimprove, while the environmentrecovers. At this point growthcan resume, and the cycle willstart all over again. This, ofcourse, is analogous to thefamous prey-predator model.Growth is the predator, it feedson its prey, society andenvironment; but if it devoursthem too much, it cannot

continue, and must slow down.Once it does so, however, societyand environment can recover,and when they do, growth canresume (for the LotkaVolterraapproach, cf. Goodwin, 196611).

Four possible patterns ofinteraction between Econ andSoc have been examined: simplelinear ones, lagged linearinteractions, sigmoid non-linearrelationships, and a prey-predator

model. Many other possibilitiescould have been explored, butthese are more than just

plausible. They show how theinteraction can either undermineor augment economicdevelopment. The developmentof the economy cannot beconsidered in isolation.

12.3.2. Interaction Between

Government and the Economy

We can set Soc aside, continue tohold our other variables constant,and turn to a study of theinteraction between EconG andEffGov, that is, betweeneconomic advance and increases

ordecreasesinthedegreeof

effectivenessofgovernmen

t.(Thereare,ofcourse,manyotherkinds


58/93

of interaction betweengovernment and the economy.)Admittedly, measuring thedegree of governmenteffectiveness will be difficult,and there will be someunavoidable arbitrariness. Butmeasures have been proposedand we can draw on them,

bearing in mind that therelationships under examinationcannot be considered exact.

Note that we do not insist that

effective government isnecessarily democratic. China,for example, has been amazinglysuccessful, though it isnotoriously not democratic.

Neither is Singapore, anothersuccess story. Two issues are

paramount: providing voice tothose who are injured, so theycan demand to be compensated,and allowing pressures todevelop that will lead to renewal,

to the renovation of institutions,clearing out calcifiedadministrative structures andrestoring flexibility.

Lets consider an interestingproblem that could arise eventhough there might be positiverelations between EconG andEffGov. On the one hand, underappropriate circumstances a risein EconG can be expected togenerate an increase in theeffectiveness of government,EffGov. Economic growth willtend to bring an increase in themiddle class, and also give rise toa prosperous upper level of theworking class. Both will push forgreater representation, and will

trytoadvancetheircauses

politi

cally.Bothwill

pushtoeduc

atetheirchildren,and

bothwilldemand

bette


59/93

r public services. EconG


60/93


and EffGov are positively related, in that an increase in EconGtends to generate a rise in EffGov.

On the other hand, an improvement in the effectiveness ofgovernment (and very often in the degree of democracy) will(usually) tend to encourage an increase in EconG. Betteradministration, more adequate provision of public goods andinfrastructure, more reliable law and order, all will contribute tofurthering EconG. So here, too, EffGov and EconG are

positively related, but in this case, an increase in EconG depends

on a corresponding increase in EffGov.We have two positive relationships between EconG andEffGov; it might seem that no difficulties could arise. Not so.Consider a low level of EconG; if the degree of EffGovgenerated exceeds that required the level is sustainable. Supposeit rises, and it is still the case that the degree of EffGovgenerated is greater than required, but not by as much. Thenmove to quite a high level; here the degree generated is less thanrequired; in between there will a point at which they just

balance. We can see all this on a simple diagram.Very roughly, what it means is that at levels of EconG below

the point of intersection, the degree of EffGov generatedexceeds that required so the way is open for EconG to increase

further. But at levels above the intersection, the EffGov requiredexceeds that generated, so EconG cant be sustained, and willhave to decline.12

Eff. Gov.

Eff. Gov. Required


61/93

> EconFi ure 12.5. E ective Government and Econ


62/93


Nowsu

pposethatth

eslopesofthelinesarereversed,andthatatlevelsofEc

onG

belowth


63/93

e intersection the requiredEffGov exceeds the generatedEffGov. EconG will not besustainable; it will have to fall tozero. (For instance, key sectorsof the developing country may beeasily monopolized, leading tostagnation; democratic politicswould be needed to break up thetrusts.) By contrast, however, atlevels of EconG above theintersection, generated EffGovexceeds required at such high

levels of economic advancesocial dislocation will be high,

but so will opportunities; moneywill flow into politics, and classand sectoral conflicts will beintense. When the level ofEffGov generated exceeds thelevel required economicexpansion can move upindefinitely, no matter how high.

As noted Government policywill be called upon to direct and

orient economic development;but government action will surelyrequire support from Soc. Butgovernment action that favorseconomic development will verylikely bring improvements inSoc. That is, EffGov will tend togenerate Soc, while alsorequiring higher Soc. So EffGovand Soc will vary together; so itmight seem that would be no

problems. However, for reasonswe have already seen in the caseof Econ and Soc, this is not so.

Look at the diagram, replacingEffGov with Soc: the twolines Econ generates Soc, Econrequires Soc - are drawn so theyintersect in the positive quadrant.

Iftherequireslineisflatter,

thenatlevelsofEcon

below

theintersectionthesystemisinatra

p;atth


64/93

ese levels of Econ more Soc isrequired than is generated, andthe position is unsustainable.(The economy needs moreeducation, more health care,

better infrastructure than it isgetting. Absent these, the level ofeconomic activity cannot besustained.) Above theintersection, however, thesituation is the opposite; moreSoc is generated than is neededto support the economy. This

could just lead to cab driverswith PhDs. But, moreoptimistically, if matched withsurpluses in other variables, itcould be the start of a takeoff intoself-sustained growth.

12.3.3. Interaction Between

Population and Economic Growth

The discussion so far has dealtwith short or intermediate term

relationships. Lets now considersome possible longer terminteractions between economicsand population pressures. Theseare, of course, much morehypothetical; many externalfactors can change and introduceunexpected influences. Therelationships themselves maychange because of newtechnologies, or new socio-

political conditions.

Nevertheless, it may still beilluminating to hold thesematters constant in our minds,and consider the relationships.They are after all, therelationships underlying the prosand cons of the debate over

Malthus.Remem

ber,Malthu

sarguedthatreformand

po

licy-induced

povertyreductionwouldnotw


65/93

ork; higher wages and higher


66/93

314 Institutional and social dynamics of growth and distribution

living standards would simply lead to increases in population,driving living standards down again. Population increasedgeometrically, but food supplies and higher living standardscould only increase arithmetically; the former would inevitablyoverwhelm the latter. But in fact economic activity, and so livingstandards, also grow geometrically, and the patterns ofinteraction are more complex, and result in quite a different

picture.At a minimum we have two relationships, enough to illustrate

the issues. There are two variables, economic growth and

population growth; one relationship shows the effects of Pop onEcon, the other shows the effects of Econ on Pop.The two relationships are:

Econ depends inversely on Pop; as Pop declines, Econ rises.Reduced pressure on natural resources allows for moreinvestment; reduced pressure on family resources allows formore investment in children, producing more highly educatedand healthier workers. On a diagram with Pop on the verticalaxis and Econ on the horizontal, this curve slopes down fromleft to right.

Pop depends on Econ, first rising with higher Econ (betterdiet, better health), then peaking and falling (smallerfamilies), finally flattening out at a low or zero level. That is,as Econ rises from a low level, it makes better health and diet

possible, so Pop grows faster, but as Econ goes higher still,women become educated and reduce their fertility, so Popslows down and growth declines to a low level. On the samediagram, this curve rises from near the origin to a peak, thenfalls, and flattens out.

As is evident in the diagram, there could easily be threeintersections of these curves. At least one would be unstable,according to the usual analysis of economists. If the first curvestarted very high and did not fall steeply, while the second rose

only a little before starting to fall, there could be no intersectionat all.

This has implications for the Malthusian debates. Contrary toMalthus it is clear that both Pop and Econ grow geometrically.But it is also apparent that there is no reason to expect them togrow at the same rate. Instead the question is, will they support


67/93

each other? At intersection points, Pop supports Econ to thesame extent that Econ supports Pop - they are mutuallyconsistent. This raises the question, are there forces that pullthem together, so that they will tend over time to grow in asupportive balance? This does not mean that they must grow atthe same rate; in fact, there may be a number of equilibrium

points, and some may be unstable, (by economists definitions- which may not always be appropriate!) The equilibrium

positions will generally not lie on the 45 degree line (the linealong which


68/93


Popgrowth=Econgr

owth).MalthusfearedthatPo

pwouldnormallygrowfas

terthanEcon.In


69/93

general this will not be true, norneed they grow at the same rate;on the contrary, in advancedcountries, normally, Econ > Pop,which implies that averageincome per capita will be rising.

Figure 12.6. Population growth and

Econ

Cost disease for Educ and HealthThis is a different kind ofdynamic problem, one thatdoesnt involve adjustment.Instead it concerns the way the

relative sizes and costs ofdifferent sectors will change overtime. As development proceedsover the long run the costs ofEduc and Health and somegovernment services will appearto rise relative to other costs. It

may

begintoseemthatthes

eservicesare

becomingm

oreandmoredifficultfortheeconomy


70/93

to afford. This is an illusion, asmany economists have argued. Inreality, they are easier for thesociety to afford; what makesthem seem relatively moreexpensive is that productivity inother sectors has been growingmore rapidly, compared to theservices in question. Slower thanaverage productivity growth in agiven sector or industry implies arise in the costs of that sectorrelative to those in others.

Productivity in certain servicessuch as Education and Healthcannot


71/93


increase rapidly some jobs just cannot be performed faster -but those who work in these sectors must be highly trained sotheir wages and salaries must keep pace.

Consider an example: an opera company is putting on CosFan Tutte, the orchestra plays and the singers sing; they are goodand do it well. Nearby there is an assembly plant in whichskilled workers put together refrigerators from importedcomponents. It takes two hours for a batch of refrigerators to beassembled from start to finish; that is also the time required to

sing the opera. Lets suppose that the number of workers and thenumber of singers and musicians are the same, and that they are

paid the same; then the cost of an opera performance and a batchof refrigerators are also the same. Now the time and motionengineers reorganize the work at the assembly plant, and withsuitable incentives, it speeds up, and a refrigerator can beassembled in one hour. But it still takes two hours to sing theopera; it wont work trying to speed up the singing. So now therefrigerators cost only half as much; that is, the cost of the operahas doubled, relative to refrigerators.

Services in Education and Health (and some other areas)require people to spend time with other people; nursing, medical

care, and teaching all take time and require person to personcommunication. Like opera singing, these services cant easilybe speeded up or be done by machine. (Of course productivitycan be improved in other ways, for example, with betterequipment but such improvements are likely to be expensive,and may improve the product or the service, rather than reducethe cost.) Care providers, teachers and research workers arehighly skilled, and must be paid accordingly; moreover asaverage pay rises, pay in these skilled services must keep pace even though productivity is growing more slowly than average.It will therefore seem that costs are rising out of control; theyare not. They appear to be rising, in fact, because other sectorsof the economy are becoming more productive.

12.3.4. Economic Growth and the Growth of Government

Growth of the economy brings with it a rise in urbanization, aswe have seen13. The number of transactions between peoplenormally depends on the density of the population in a given


72/93

area given the transport and communications system. Urbanpopulation density is obviously higher than rural, and urbantransportation and communications are normally much superior.So a rise in urbanization typically leads not only to an increasein transactions, but to a more than proportional increase. Again,typically, this tends to reduce transaction costs, and also leads toeconomics of scale and scope. The presence of such economies,and also of externalities resulting from concentration, is well-known to promote agglomeration, which in turn


73/93


willtendtogeneratein

creasesin

productivity(Acs,2006).Theexternalitiesandsp

illovereffects


74/93

are particularly strong in regardto human capital, skilledworkers and especially the

professional classes. (Simon andNardinelli, 1996) So an increasein urbanization could be expectedto lead back to a rise in

productivity, which could, inturn, have a positive effect ongrowth.

These are commonobservations, and figurefrequently as assumptions in

discussions of the reasons formigration to urban areas, and forlocating industry in urban areas.

However, urbanization will notcontinue indefinitely; a shift fromlow or irregular growth to strongand regular growth will lead to arise in urbanization, setting off arate of migration from thecountryside that will continue fora long time. But the migrationwill slow down and come to a

halt when the gains fromadditional migration are offset atthe margin by the rising costs ofcongestion. At this point industrywill seek to locate its expansionelsewhere perhaps back in thecountryside, perhaps in a newarea of agglomeration, or

possibly overseas and migrantswill find it sufficiently difficult tofind accommodation, jobs, healthcare, etc, that they will look forways to survive in the informaland illegal economy.

As urbanization increases,transactions increase more than

proportionally. If we think of thisin terms of a network, the in-movement of migrants increases

thenum

berofnodes

perun

itofspace(thesamecan

besaidfortheformationofnew

businesses)


75/93

. This can be thought of as anarithmetical increase in the nodesin a given area of living space.But, as a first approximation, thenetwork consists of the directrelations between the nodes, andthis increases per unit space in amuch faster rising series:

Number per unit space: 1 2 3 4 5 6 7 8

9

Relations per unit space: 0 1 3 6 10 15

21 28 36

The expenses of governmentarguably will depend not on thenumber of people, but on therelationships and activities thatarise between them; it is toencourage and support theseconnections and activities thatgovernment build infrastructure(roads, sewers, light and power,water, etc.), provide police andcourts, undertake regulation, and

support education and publichealth. (Of course not all the

possible direct relations willbecome actual; also, besides thedirect relations, there can behigher order relationships.) So, asa very simple firstapproximation, governmentexpenditure will tend to be

proportional, not to the numberof people, but to the connections

between them. It will therefore

rise with urbanization, but willrise faster than the urbanpopulation; however, it will thenslow down and stop rising whenurbanization slows down andstops.


76/93


So as population shifts from rural and lightly populated areasto urban intensely populated areas, the number of transactionsand activities in the urban areas increases at a faster rate than the

population intensity; government expenditures will then risewith the transactions activity.

These models show not only the variety and complexity ofpossible interactions between the variables of theTransformational Growth Matrix, but they demonstrate that the

problems of the economy cannot be separated from those of the

society, its politics and its environment. (See also Keuning,1996, 1997) Economics as a subject cannot be isolated, excepttemporarily for the purposes of argument, from the other socialsciences.

The economy is supposed to be generally independent ofthe rest of society; it produces the goods and services thatsupport the rest of society, but only insofar as these needs aremanifested through the market; and it is not itself directly orimmediately dependent on the rest of society. Indirectly, and inthe long run, yes, of course, the economy depends on andinteracts with the social system. But not in the short run, and notimmediately or directly. So when the economy is well-

developed, as in the advanced countries, the matrix will exhibitthis independence; for the short run case, many of the cellsrepresenting the interaction between EconG and the socialvariables will show zeroes. But this will never be the case for adeveloping economy.

12.3.5. Advanced Socio-Economic Development

This suggests a way of defining economic development.Typically in traditional societies, there will be a balance betweenEcon, Env and Soc for small changes in established economic

practices. But substantial and innovative economic advance willcreate turmoil and have a negative impact on the social

variables. Successfully developing economies will exhibit apositive relationship, a virtuous cycle. Then economicdevelopment can be said to reach a high or advanced levelwhen social and environmental activities can be supported on anindependent basis, funded, so that they do not depend on thesuccess or failure of current economic activities. That is when


77/93

the economy emerges, as an aspect of the society (partially)independent of the rest.

We can rewrite the matrix again, now showing that in a fullydeveloped economy many short-run connections can be severed,so that the different areas are independent of one another. Thiswill be represented by 0s. In an advanced economy a rise inEconG means an increase in growth and in productivity. So insuch an economy we can expect the acceleration principle towork; EconG will therefore have a positive effect on itself. Animproved Env will surely support EconG, but better AdlSoc isunlikely to have much


78/93


immediateim

pact.Im

provedEduc,Health, ahigherHStndland

betterSocInf

mayallhave


79/93

an encouraging effect. Betterpolitical institutions are unlikelyto have much effect in the shortrun, but they might.

Now look at the effects ofhigher EconG on the othervariables, bearing in mind thatwe are thinking of the short tomedium term, the next year orso. Of course, higher EconG can

be expected to show up as higherHStndl. But otherwise it will nothave any great impact on any of

the other variables. The sectorsthat these variables represent arealready supported anddeveloping according to plan,independently of whether or notthe economy is running strongly,that is to say, independently ofthe business cycle. If they aresupported by taxes theexpenditures they require will besustained by deficit spendingduring downturns; to the extent

they are private, funding willcarry them through difficulttimes.

Table 12.4. The case of advanced

development

EconG EnvH AdlSoc

Educ Health HStndL SocInf

EffGov PopP

Econ + + 0 + + +

EnvH 0 ------- 0 0 0 0AdlS 0 0 ----- 0 0 0

Educ 0 0 + 0 0

Healt 0 + 0 + ------ 0

HStn + + 0 + + -------

SocIn 0 0 0 0EffG 0 0 0 +

PopP 0 0 0 0

Thista

blesugg

eststhatonceacountryisdevelo

ped,further

progress

islikelyto


80/93

come chiefly through thepolitical arena. Effective andresponsive government, EffGov,will have a positive effect on allthe social variables. A moresensitive, better functioninggovernment, especially oneresponsive to the public, will beable to offer improved servicesto the economy, to monitor theenvironment more carefully, to

provide programs to counseladolescents, to promote

education and health, improvethe distribution of income andsocial services, thereby raisingthe household standard of living,.And it will most likely work toimprove social infrastructure,and finally it is likely to react

back on itself, and move towardsimproved and more democratic

practices. All of these will befurthered or supported by strongand effective approach to

demand management whichcould be developed around a


81/93


public service employment program, operating as an automaticstabilizer as suggested elsewhere.14

12.4. Concluding Remarks

Now lets review what we have accomplished; initially we setup the Transformational Growth Matrix to show the way thevariables impact on each other; now we have indicated thedynamic implications, sketching how those impacts play out

filling in the initial sketches.

1. First, we have broken down the barrier between economicand social/environmental analysis. Our approach shows howeconomic variables interact with social and environmentalones. Contrary to what some mainstream economists

believe, not only can economics not be isolated from the restof society, but in principle, we can show just how it impactson the other aspects of society and how society reacts backon it. These relationships can be modeled precisely, and wecan see that there are many possible positions of mutualsupport (equilibrium, economists would say), both stable

and unstable.2. Second, this also lets us show how and to what extent theeconomy can become independent of the rest of society asdevelopment proceeds. The separation of the economy fromsociety is not inherent or necessary; it emerges as a result ofdevelopment and will normally remain partial andincomplete.

3. Third, weve provided a method not only for examining thistheoretically, but one which can be applied practically. Itsflexible and can be used with ordinal or cardinal measures,even without numbers at all; this could be important in

practice, since the statistics in developing areas are oftenpoor to non-existent.

4. Fourth, weve derived the condition for a balanced mutuallysupportive relationship, one where, for example, the effect ofEcon on Soc is just balanced by the reciprocal effect of Socon Econ.

5. Fifth, weve shown how the matrix can be partitioned, and


82/93

the coefficients used to defin

chapter 12. interaction between economic and social variables: the transformational growth matrix

Documents