Title: Recording the History of Economic Exchange Alters That History
1Recording the History of Economic Exchange Alters
That History
- Presentation by John Dickhaut
- Purdue University
- October 26, 2006
2Overview
- What do we seek to explain?
- Why are we conducting an experiment?
- Whats the experiment about?
- What does the evidence suggest?
- Where are we headed with this research agenda?
3What do we seek to explain?
- Our focus is on the creation of a record of
exchange outside the human brain. - i.e., Biology, anthropology economics of the
journal entry. - Studying recordkeeping requires us to go in two
directions - Back to the brain to study antecedent conditions.
- Forward into relatively primitive economies to
study the consequences of recordkeeping. - Why is this research important?
- We want to explain the genesis of accounting,
which is the creation of a record of exchange
external to the brain.
4The Evolutionary Function of Memory
- The ability to remember past interactions with
ones environment and other strategic actors
within that environment is crucial to survival. - Present in simple organisms e.g., the e coli
bacterium. - Memory sustains cooperation organisms have to
recognize other individuals and keep score.
(Ridley 1996, 83) - Game theorists note strategies like tit-for-tat
work fairly well but require some ability to
recall past moves. - Evolutionary psychologists conclude that memory
is an adaptation that primes humans for exchange
and social interaction.
5Accounting as Memory External to the BrainBasu
Waymire (2006) Story
- Records created outside the brain as exchange
complexity increases. The origins of accounting
derive from memory constraints within the human
brain. - Recordkeeping, law other exchange supporting
institutions evolve to replace unaided human
memory and informal sanctions for non-cooperative
behavior present in small groups. - Larger networked groups enabled by recordkeeping
other institutions enable complex exchange that
promotes division of labor that improves economic
welfare (Adam Smith). - Groups that evolve markets with effective
supporting institutions organizations will, all
else equal, more likely flourish (Douglass
North).
6Basu Waymire (2006) Predictions
- Formal recordkeeping will emerge as a mnemonic
aid as the complexity of exchange increases.
Recordkeeping sophistication will be a function
of - exchange complexity scale,
- the extent to which a given culture relies on
markets and a specialized division of labor - Evolved recordkeeping will feed back and promote
- more complex and advantageous exchange
- the co-evolution of exchange-supporting
institutions including language, standardized
weights measures and law that punishes
non-cooperators
7Three Ways to Investigate Predictions
- Archaeology Examine the characteristics of
ancient civilizations their recordkeeping
practices - Anthropology Examine ethnographic data drawn
from a spectrum of cultures - Experimental economics Examine use of
recordkeeping and its consequences in a lab
setting
8Archaeological EvidenceBroadly Consistent with
B-W Story
Sumerian marked envelope with matching tokens
(3300 BC)
9Archaeological EvidenceLimitations - In
Ascending Order
- Very difficult expensive to do
- Small samples involving more intensive research
of specific cases - Data are hard to time - i.e., dating artifacts to
within 100 years makes it hard to draw causal
inferences.
10Ethnographic SamplesData Sources
- Human Relations Area Files (HRAF) at Yale (some
available online). - Cumbersome hard to work with.
- Galtons Problem. Data lack independence.
- Standard Cross-Cultural Sample (SCCS)
- Designed to maximize data independence.
- Extensive data coded for 186 societies for over
1,000 variables.
11Examining Ethnographic SamplesSCCS
Recordkeeping Variable 149
- 1 None (Nama Hottentot 1,860 A.D.) n 73
- 2 Mnemonic Devices (Kung Bushmen 1,950 A.D.) n
49 - 3 Nonwritten records (Incans 1,500 A.D.) n 21
- 4 True writing, no records (Somali 1,900 A.D.)
n 12 - 5 True writing, with records (Babylon 1,700
B.C.) n 31
12Tests of AssociationCorrelations of
Recordkeeping with Measures of Economic
Complexity (SCCS)
13Ethnographic EvidenceStrengths Weaknesses
- Strengths
- Quantification allows for statistical analysis
- Large samples allow generalizability
- Weaknesses
- Still cant establish causality
14Why an Experiment?
- CAUSALITY
- Seek to re-create Baseline Mesopotamia
- That is, we use a simple experiment to
investigate whether the hypothesized economic
forces - Give rise to recordkeeping (Antecedents)
- Lead to fundamental changes in the nature of
economic exchange itself. (Consequences)
15Experimental Set-UpTrust Game
16Experimental Set-Up (contd.)
- Trust Game with two manipulations
- Exchange Complexity
- Simple One-on-one play for 10 periods
- Complex Each subject plays game simultaneously
with five others (5 on 5) for 10 periods - Recordkeeping Availability
- Yes (Computer screen available where doodling is
possible) - No (Computer screen not available)
17Exchange Complexity
Complex
Investors Trustees A1 B1 A2 B2 A3 B3 A
4 B4 A5 B5
Investors Trustees A1 B1 B2 B3 B4
B5
18Exchange Complexity (cont.)
Complex
Investors Trustees A1 B1 A2 B2 A3 B3 A
4 B4 A5 B5
Investors Trustees B1 A2 B2 B3 B4
B5
19Exchange Complexity (cont.)
Basic
Complex
Investors Trustees A1 B1 A2 B2 A3 B3 A
4 B4 A5 B5
Investors Trustees A1 B1 A2 B2 A3 B3 A
4 B4 A5 B5
20Four Primary Results
- Endogenous recordkeeping in complex exchange.
- Recordkeeping increases gains from exchange in
complex setting. - Recordkeeping promotes trust fairness in
economic exchange. - Recordkeeping alters fundamentally the evolution
of economic exchange.
21Finding 1 Recordkeeping Endogenous to Complex
Exchange
22Finding 2 Recordkeeping Produces Higher Gains
from Complex Exchange
23Finding 3 Recordkeeping Promotes Commitment
Reciprocity in Exchange
24Finding 3 (contd.) Recordkeeping Promotes Trust
Reciprocity in Exchange
25Finding 3 (contd.) Recordkeeping Promotes Trust
Reciprocity in Exchange
26Finding 4 Recordkeeping Alters Fundamentally the
Evolution of Exchange(Within Dyad Correlations)
27Finding 4 (contd.) Recordkeeping Alters
Fundamentally the Evolution of Exchange
(Correlations with Average ROI from Other
Trustees)
28Finding 4 (contd.) Recordkeeping Alters
Fundamentally the Evolution of Exchange
(Correlations with Average ROI Received by Other
Investors)
29ConclusionsWhat Have We Learned Here?
- Direct causal support for the hypothesis that
accountings recordkeeping function emerges from
complex exchange and enables beneficial exchange. - Powerful reminder that seemingly simple
frequently overlooked institutions exert
powerful effects on the organization of human
society. - Mind those debits credits!
30Experiments on the fundamental nature of
accounting institutions and conventions
- Contributors
- John Dickhaut, Greg Waymire, Sudipta Basu, Kristy
Towry, Ivo Tavlov, Radhika Lunawat, Baohua Xin
31Why these types of experiments
- Feynman-Critical role of experiments in physics.
- Emerging role of experiments in economics.
- Experiments in physics and economics are part of
teaching. - Maybe experiments could be used to do more in
accounting.
32An appropriate experimental platform
- A Platform should have some characteristics of
physics experiments - ExampleTwo Slit Experiments of Quantum
Mechanics. - Easy to add on to the structure of the experiment
to get deeper results. - A platform to study the fundamentals of
accounting should operationalize the notion of
investing capital and getting a return on
capital. - Without other accessories such as a legal system.
33A platform - The Investment Game
34The Universal Adoption of Income Reporting(Very
Early)
- John Dickhaut, Radhika Lunawat, Greg Waymire and
Baohua Xin
35Abstract
Why do firms report income numbers? Virtually
all established companies and known small
enterprises report income numbers. We consider
this question using four person economies with
two investors and two trustees each. Trustees
can precommit to reporting income before capital
is invested. Investors can then use the
precommit information as a proxy for trust. The
experimental design asks if competition for
reputation for being trustworthy among trustees
will lead them to tend to universally report
income.
36What is the issue?
- Why do firms report income numbers?
- Accounting a discipline that has worried about
its existence for the last 40 years. - Most serious researchers I know ask why does
accounting exist. - Actually never built a structure of why income
exists. - 1000s of studies interested in the properties of
income, dividends their relationship and how do
markets function. - In particular historical records of when firms
switched form reporting income to reporting
dividends. -
37What is augmented structure?
38Design of the experiment
- Adjust trust game so that the multiplier is
risky. - Setting has five or two person dyads per economy.
- Allow cheap talk.
- ManipulationGive trustees the ability to
precommit to having ?x, the amount of income
revealed at the end of period after dividends
revealed.
A1 B1 A2 B2 A3 B3 A4 B4 A5 B5
39Intuition
- Production Income.
- Investors will want to know which trustees are
skimming off the top and which are not. - A trustee not wanting to be perceived as
potential skimmer will choose to precommit to
reporting income. - Because no trustee will want to be left behind
eventually all trustees will commit to reporting
income. - An aside on precommitment
40Design and Methods
- Z tree
- Put on top of Z tree a mechanism for cheap talk.
- You could only talk with someone you were paired
with. - Put in option for trustee of having production
information revealed.
41Picture of design and data to date
Precommitment Possible
No Precommitment Possible
Risky
-One 10 person economy -Four 2 person economies
No data at all
Ambiguous
No data at all
No data at all
42Results (Data planned to be collected)
- No of precommitters through time.
- Efficiencies (with and without precommittment to
report income) - Investment conditional on precommittment vs.
Investment conditional on no precommittment - Return conditional on precommittment and return
conditional on no precommittment.
43Results for 5 X 5 Configuration
44Results for 2 X 2 Configurations
45Thought experiment on verifiable information
- VerifiableHard Information.
- Hard Information Signed receipt.
- Allow cheap talk.
- Give people the ability to write on each others
computer screens and send captured screens to
each other.
46Thought experiment on matching
- Complex setting.
- With recordkeeping.
- Different types of production rates for different
pairings.
47Thought experiment on creating accountancy
- Basic trust game complex setting repeated.
- Allow side payments.
- All information available to everyone. See if
someone is willing to differentially pay best
record keeper.
48The value of thought experiments for accounting
- Allows discussion across different accounting
domains - Analytical, empirical and experimental.
- Helps avoid possible wastage of resources on poor
experiments. - Allows discussion with non-accounting disciplines.
49Do you have some ideas for thought experiments?
- Some potential candidates for thought
experiments. - The emergence of accounting conservatism.
- The emergence of an enforcement mechanism.
- The emergence of auditing.
- What is the fundamental source of where the
accounting institutions and conventions came
initially and why will we see these institutions
emerge in our experiments? - THE BRAIN IS THE ORIGINAL ACCOUNTING INSTITUTION
AND INFLUENCES INDIRECTLY OUR INSTITUTIONAL
ACCOUNTING CHOICES.