Economics of Organisations - terminology

appropriable quasi-rents

• appears when some assets/investments are more valuable within the transaction than outside it

• they are a temporary economic rent obtainable by the owner of an asset with temporary supply constraints

• the difference between the value of the asset in its present use (ex-ante) and its next best alternative use (opportunity cost)

hold-up problem

• when opportunism leads parties to haggle and change contractual terms in order to capture appropriable quasi rents

• this can prevent efficient transactions from happening

• when appropriable quasi-rents are high, integrating the transaction within the firm is optimal

asset specificity

• relates to the inter-party relationships of a transaction

• it is the degree to which an asset of value can be readily adapted for other purposes

  • an asset with high specificity is useful only for certain tasks or in certain circumstances

  • an asset with low specificity is a more flexible resource, and therefore a more valuable asset

• applies to many settings

  • eg designing a software specific for a client (high)

  • worker learning a production method in a firm (high)

• results in hold-up problems

backhaul (Baker and Hubbard, 2004)

• the transportation of cargo on the return trip from point B to original point A

integration

• unification of control rights

double marginalization

• multiple firms in the supply chain applying their own markups, leading to higher prices

alienable capital

• able to transfer ownership

• eg physical capital

inalienable capital

• not able to transfer ownership

• eg human capital

vertical integration

• downstream producer acquires an upstream supplier

  • activities all carried out by the producer

spot markets

• the total amount of input the supplier producers and is price are determined in a competitive market based on the interaction of demand and supply

  • price is determined on a transaction by transaction basis

OLS method

• we use this when we have some data and we estimate the linear relation

• OLS finds the line (ie intercept and slope) that minimizes the (squared) distance between the line and dots

beta is not identified

• there are more regressors than equations

• we see this when we have OVB

confounder

• omitted variable

noise

• random (not correlated with any variable)

• zero mean (moves x up or down)

attenuation bias

• biases a variable towards zero

residuals

• the difference between an observation and the OLS line

balance tests

• hard to prove ‘random’ allocation of treatment T in natural experiments

• need evidence that T is ‘exogenous’ or ‘uncorrelated’ with any determinant of outcome

• don’t observe every possible variable but we can test if it is uncorrelated with some ‘observed’ determinants

• not perfect, but strongly suggestive test

• ie we are carrying this to prove that results are not driven by some pre-existing characteristics

LATE theorem

• it is possible to show that the coefficient from the IV regression (where Z is a dummy and X is also a dummy) captures the causal effect of X on Y among the compliers

  • compliers: X = 1 if Z=1 and X= 0 otherwise

  • other alternatives:

    • never takers: X = 0 always

    • always takers: X = 1 always

    • defiers: X = 1 if Z = 0 and X = 0 otherwise (assume they don’t exist because this is an odd relationship)

fixed effects

• set of dummies that ‘absorb’ any characteristic that changes at the defined level

• interpret in the same way as dummies

  • eg effect of CEO pay on profits when comparing firms within same State

• only works for observations that occur over time

• absorbs any heterogeneity at that level

productivity

• efficiency with which organisations transform inputs into outputs

  • not only firms but any organisation (schools, hospitals, government, etc)

r-squared

• measure of the explanatory power of the independent variable(s)

• when the regression line fits more of the observations in the dataset, it has a higher explanatory power

  • how much does the independent variable explain the variation in y?

incidental parameters problem

• when the presence of incidental parameters affects the maximum likelihood estimates of other parameters of interest

• problem usually arises in the context of panel data, where individual-specific parameters may relate to consumer, firm, or country fixed intercept effects

• it leads to inconsistent estimates of common panel regression coefficients, as the incidental parameters only figure in a finite dimensional probability law, involving only a finite number of observations

• relates to Hoffman and Tadelis (2021) where they did not include worker FE to avoid this issue

linear in means

• have averages for the variables and the averages enter the equation linearly

reflection problem in peer effects

• arises when it is difficult to distinguish between the influence of peers on an individual and the influence of an individual’s traits on their peers

• can lead to underestimation

peter principle

• in a hierarchy, employees tend to be promoted until they reach a level at which they are incompetent

managerial fiat

• the idea that workers only do things because their managers tell them to do it

hawthorne effect

• a policy may create a short-run improvement that fades out

  • ie workers notice there is a policy change so they increase their effort, but this eventually converges (due to behavioural responses)