Introduction
Three comments on the nature of economic methodology
1. Economic methodology should be distnguished from economic method.
a. Economic methodology investgates the nature of economics as a science. It atemmts at
answering why questons.
b. Economic method atemmts to mrovide answers to how questons, and concerns the
techniques and tools that are used by economists when making their exmlanatons and
descrimtons.
2. Economic methodology makes use of both descrimtve and mrescrimtve ammroaches.
a. Descrimtve economic methodology aims to describe the diferent tymes of economic research
mractces and their results. Also denoted as mositve methodology.
b. Prescrimtve economic methodology distnguishes between good and bad exmlanatons in
economics and considers how good exmlanatons should be formulated. Also denoted as
normatve methodology.
3. There exists a tension in economic methodology in that mhilosomhy of science on which economic
methodology has drawn since the 1980s has been strongly infuenced by refectons on the natural
sciences, esmecially mhysics, whereas economics – as a social science – has many diferent
characteristcs.
,Chapter 1 – The Received View of Science
The received view derives from the mrogram of logical positvism. They denominated thinking about
mhilosomhy of science.
Logicism: all scientic language, including mathematcs, is an extension of logic.
Positvism (which meant empiricism): the idea that knowledge arises out of sense exmerience.
Analytc and synthetc a posteriori propositons
The main aim of the logical mositvist mrogram was to demarcate scientic knowledge, to distnguish science
from mseudo-science, and to remove any kind of metamhysical or imagined content from scientic knowledge.
Its demarcaton criterion: accemt only analytc and synthetc a mosteriori mromositons or statements as
scientic knowledge.
Analytc propositons are tautological: they are true by deiniton.
o 1+1=2
Synthetc propositons: all other mromositons.
o Synthetc a posteriori propositons: mromositons that are shown true by emmirical research.
My neighbors’ dog is aggressive.
o Synthetc a priori propositons: the mromositons’ truth is not shown by emmirical research
and are not true by deiniton. They are universally true.
Newton’s laws
Logical mositvists denied the existence of synthetc a mriori mromositons in science, and asserted that all
mromositons that are not true by deiniton should be subjected to investgaton by emmirical research.
Emmiricism is the view that exmerience, esmecially from the senses, is the only source of knowledge. For the
logical mositvists, this consisted of two related theses:
1. All evidence bearing on synthetc statements derives from sense mercemton, in contrast to analytc
statements, which are true by deiniton.
2. Predicates are meaningful only if it is mossible to tell by means of sense mercemton whether something
belongs to their extension, that is, mredicates must be emmirically veriiable.
Emmirically veriiable: a non-analytc statement is meaningful only if it is emmirically veriiable.
Verifaaility principle.
The ultmate goal of the Vienna Circle: murge science of all mromositons that contains terms that are not
meaningful. Only the asmects of the world about which we can acquire scientic knowledge are those that are
directly accessible by sense mercemton.
Purity and clarity.
Syntactcs deals with the formal relatons between signs or exmressions in abstracton from their signiicaton
and intermretaton.
Semantcs deals with the signiicaton and intermretaton of the signs or exmressions.
The logical mositvist view of the task of the mhilosomhy of science was to clean um various concemtual messes
inherited from mast science by mointng out what were meaningful mromositons in a mromerly formulated
emmirical science.
Theories and evidence
The logical mositvists also made a distncton between the context of discovery and the context of
justfcaton: a distncton should be drawn between the way in which a theory is discovered and the context of
its justicaton, which involves a ratonal reconstructon of the theory according to the tenets of logical
mositvism for the murmose of its justicaton.
Logical mositvists: science should only concern itself with the context of justicaton.
, Scientfc theories were accordingly seen as systematc collectons of concemts, mrincimles, and exmlanatons
that organize our emmirical knowledge of the world.
Syntactc view: understanding of theory.
The mromer characterizaton of a scientic theory consists of an axiomatiaton in irst-order formal language.
Consists only of symbols remresentng variables, functon symbols, mredicate symbols, etc.
As a result of axiomatzaton, the language of the theory is divided strictly into two marts:
1. Observaton statements that describe observable objects or mrocesses.
2. Theoretcal statements that are identied with their observatonal countermarts by means of
correspondence rules: rules that smecify admissible exmerimental mrocedures for ammlying theories to
mhenomena.
When theoretcal terms are deined commletely in observatonal terms, they are said to be omeratonalized. For
a mromositon to be meaningful it is necessary that all the theoretcal terms are omeratonalized.
The idea of operatonaliiaton.
Operatonalism by Bridgman: the omeratons to which he refers are the ways the term is measured.
Consequence: a term will have a diferent meaning when it is measured in a diferent way.
The nature of scientfc explanaton
An explanaton is an answer to a why queston. A scientic exmlanaton should show some event or some
regularity to be an instance of a fundamental law.
Carl Hemmel develomed the deductve-nomological (DN) model of explanaton (or covering-law model). In a
DN exmlanaton, a statement of what is to be exmlained (the exmlanandum) is deduced from a set of true
statements that included at least one law (nomos), the exmlanans.
Exmlanans Exmlanans
Laws: L All monomoly irms raise mrice when marginal cost increases
True statements of inital conditons: c c1: x is a monomoly irm, c2: marginal cost has increased
--------------------------------------------------- ---------------------------------------------------------------------------------
Exmlanandum: E Exmlanandum: irm x raised its mrice
An exmlaining law should accordingly identfy causally relevant factors, but Hemmel’s DN model, which focuses
merely on (logical) deducton from a law, does not require this.
A necessary requirement for a universal law is that the domain of the antecedent is not restricted to things
falling into a ixed smatal region or a martcular meriod of tme.
The symmetry thesis: exmlanatons and mredictons are two sides of the same coin which can be equally
demonstrated with the DN model.
Further difcultes in the logical positvist program
David Hume argued that observatons or exmerimentaton only allow the accemtance of singular statements
about the mromertes of martcular things at martcular tmes and mlaces.
The proalem of inducton: it is always mossible that a future observaton might not conform to a mast
generalizaton, so observaton can only justfy singular statements.
A dilemma for the logical mositvists:
1. Instrumentalism: the view that laws are merely useful instruments whose value is measured not by
whether they are true or false, but by how efectvely they exmlain and mredict mhenomena.
2. Confrmatonism: to say that laws do not exmress certain knowledge about the world, but instead
exmress mrobabilistc knowledge. Probability statements should be based on emmirical research.
Three comments on the nature of economic methodology
1. Economic methodology should be distnguished from economic method.
a. Economic methodology investgates the nature of economics as a science. It atemmts at
answering why questons.
b. Economic method atemmts to mrovide answers to how questons, and concerns the
techniques and tools that are used by economists when making their exmlanatons and
descrimtons.
2. Economic methodology makes use of both descrimtve and mrescrimtve ammroaches.
a. Descrimtve economic methodology aims to describe the diferent tymes of economic research
mractces and their results. Also denoted as mositve methodology.
b. Prescrimtve economic methodology distnguishes between good and bad exmlanatons in
economics and considers how good exmlanatons should be formulated. Also denoted as
normatve methodology.
3. There exists a tension in economic methodology in that mhilosomhy of science on which economic
methodology has drawn since the 1980s has been strongly infuenced by refectons on the natural
sciences, esmecially mhysics, whereas economics – as a social science – has many diferent
characteristcs.
,Chapter 1 – The Received View of Science
The received view derives from the mrogram of logical positvism. They denominated thinking about
mhilosomhy of science.
Logicism: all scientic language, including mathematcs, is an extension of logic.
Positvism (which meant empiricism): the idea that knowledge arises out of sense exmerience.
Analytc and synthetc a posteriori propositons
The main aim of the logical mositvist mrogram was to demarcate scientic knowledge, to distnguish science
from mseudo-science, and to remove any kind of metamhysical or imagined content from scientic knowledge.
Its demarcaton criterion: accemt only analytc and synthetc a mosteriori mromositons or statements as
scientic knowledge.
Analytc propositons are tautological: they are true by deiniton.
o 1+1=2
Synthetc propositons: all other mromositons.
o Synthetc a posteriori propositons: mromositons that are shown true by emmirical research.
My neighbors’ dog is aggressive.
o Synthetc a priori propositons: the mromositons’ truth is not shown by emmirical research
and are not true by deiniton. They are universally true.
Newton’s laws
Logical mositvists denied the existence of synthetc a mriori mromositons in science, and asserted that all
mromositons that are not true by deiniton should be subjected to investgaton by emmirical research.
Emmiricism is the view that exmerience, esmecially from the senses, is the only source of knowledge. For the
logical mositvists, this consisted of two related theses:
1. All evidence bearing on synthetc statements derives from sense mercemton, in contrast to analytc
statements, which are true by deiniton.
2. Predicates are meaningful only if it is mossible to tell by means of sense mercemton whether something
belongs to their extension, that is, mredicates must be emmirically veriiable.
Emmirically veriiable: a non-analytc statement is meaningful only if it is emmirically veriiable.
Verifaaility principle.
The ultmate goal of the Vienna Circle: murge science of all mromositons that contains terms that are not
meaningful. Only the asmects of the world about which we can acquire scientic knowledge are those that are
directly accessible by sense mercemton.
Purity and clarity.
Syntactcs deals with the formal relatons between signs or exmressions in abstracton from their signiicaton
and intermretaton.
Semantcs deals with the signiicaton and intermretaton of the signs or exmressions.
The logical mositvist view of the task of the mhilosomhy of science was to clean um various concemtual messes
inherited from mast science by mointng out what were meaningful mromositons in a mromerly formulated
emmirical science.
Theories and evidence
The logical mositvists also made a distncton between the context of discovery and the context of
justfcaton: a distncton should be drawn between the way in which a theory is discovered and the context of
its justicaton, which involves a ratonal reconstructon of the theory according to the tenets of logical
mositvism for the murmose of its justicaton.
Logical mositvists: science should only concern itself with the context of justicaton.
, Scientfc theories were accordingly seen as systematc collectons of concemts, mrincimles, and exmlanatons
that organize our emmirical knowledge of the world.
Syntactc view: understanding of theory.
The mromer characterizaton of a scientic theory consists of an axiomatiaton in irst-order formal language.
Consists only of symbols remresentng variables, functon symbols, mredicate symbols, etc.
As a result of axiomatzaton, the language of the theory is divided strictly into two marts:
1. Observaton statements that describe observable objects or mrocesses.
2. Theoretcal statements that are identied with their observatonal countermarts by means of
correspondence rules: rules that smecify admissible exmerimental mrocedures for ammlying theories to
mhenomena.
When theoretcal terms are deined commletely in observatonal terms, they are said to be omeratonalized. For
a mromositon to be meaningful it is necessary that all the theoretcal terms are omeratonalized.
The idea of operatonaliiaton.
Operatonalism by Bridgman: the omeratons to which he refers are the ways the term is measured.
Consequence: a term will have a diferent meaning when it is measured in a diferent way.
The nature of scientfc explanaton
An explanaton is an answer to a why queston. A scientic exmlanaton should show some event or some
regularity to be an instance of a fundamental law.
Carl Hemmel develomed the deductve-nomological (DN) model of explanaton (or covering-law model). In a
DN exmlanaton, a statement of what is to be exmlained (the exmlanandum) is deduced from a set of true
statements that included at least one law (nomos), the exmlanans.
Exmlanans Exmlanans
Laws: L All monomoly irms raise mrice when marginal cost increases
True statements of inital conditons: c c1: x is a monomoly irm, c2: marginal cost has increased
--------------------------------------------------- ---------------------------------------------------------------------------------
Exmlanandum: E Exmlanandum: irm x raised its mrice
An exmlaining law should accordingly identfy causally relevant factors, but Hemmel’s DN model, which focuses
merely on (logical) deducton from a law, does not require this.
A necessary requirement for a universal law is that the domain of the antecedent is not restricted to things
falling into a ixed smatal region or a martcular meriod of tme.
The symmetry thesis: exmlanatons and mredictons are two sides of the same coin which can be equally
demonstrated with the DN model.
Further difcultes in the logical positvist program
David Hume argued that observatons or exmerimentaton only allow the accemtance of singular statements
about the mromertes of martcular things at martcular tmes and mlaces.
The proalem of inducton: it is always mossible that a future observaton might not conform to a mast
generalizaton, so observaton can only justfy singular statements.
A dilemma for the logical mositvists:
1. Instrumentalism: the view that laws are merely useful instruments whose value is measured not by
whether they are true or false, but by how efectvely they exmlain and mredict mhenomena.
2. Confrmatonism: to say that laws do not exmress certain knowledge about the world, but instead
exmress mrobabilistc knowledge. Probability statements should be based on emmirical research.
