

mestimators
Estimators that maximize a sample average. The 'm' means 'maximumlikelihoodlike'. (from NeweyMcFadden)
The term was introduced by Huber (1967). "The class of Mestimators included the maximum likelihood estimator, the quasimaximum likelihood estimator, multivariate nonlinear least squares" and others. (from Wooldridge, p 2649)
I think all mestimators have scores.

M1
A measure of total money supply. M1 includes only checkable demand deposits.

M2
A measure of total money supply. M2 includes everything in M1 and also savings and other time deposits.

MA
Stands for "moving average." Describes a stochastic process (here, e_{t}) that can be described by a weighted sum of a white noise error and the white noise error from previous periods. An MA(1) process is a firstorder one, meaning that only the immediately previous value has a direct effect on the current value: e_{t} = u_{t} + pu_{t1} where p is a constant (more often denoted q) that has absolute value less than one, and u_{t} is drawn from a distribution with mean zero and finite variance, often a normal distribution. An MA(2) would have the form: e_{t} = u_{t} + p_{1}u_{t1} + p_{2}u_{t2} and so on. In theory a process might be represented by an MA(infinity).

MA(1)
A firstorder moving average process. See MA for details.

macro
macro

MacroInstitutionalism
Generally, macroinstitutionalisation is seen as the tendency of organizations to arrange
their formal structure not in response to the technical needs of the organizations but in
accordance to certain widely accepted rules. This is done in order not to loose the
legitimacy towards important stakeholders like banks, clients etc.
(Scott, 1987,
Scott, 1995,
Zucker, 1991,
Meyer & Rowan, 1977).
Organizations are expected to conform to the institutionalized rules. So firms react to
these expectations of good practice rather than looking for the most rational solutions.
An example for this notion would be the implementation of computing facilities in
organizational settings just because competing firms use computing facilities in similar
settings too. "We arrive at the conclusion that formal organization, as it expands in a
domain or society, becomes less explicitly rational in its structure. ... Every aspect of
rationalized organizational structure comes under exogenous institutional control ..."
(Meyer, 1992: 268).
The consequence of this tendency is that within an organization the
institutionalized routines might be decoupled from the actual practice of the organization.
The formal rules signal to the environment that the organization complies with the
institutionalized norms of organizing. However, the strict appliance of the rules would
lead to inconsitencies. Therefore the organizational members (OM) have the freedom to
arrange the tasks in a way which they consider most efficient  thereby violating the
official rules (Meyer & Rowan, 1977: 357).

Magic cards
" the Gathering game scenario, players assume the roles of dueling wizards, each with their own libraries of magic spells(represented by decks of cards) that may potentially be used against the player's opponent. Cards are sold in random assortments, just like baseball cards, at retail stores. Launched in August 1993, this product has already grossed hundreds of millions of retail dollars, and now has over a million players worldwide."  description from Reily (1999), " Using Field Experiments to Test Equivalence Between Auction Formats: Magic on the Internet." AER, 89

main effect
As contrasted to interaction effect.
In the regression
y_{i} = aX_{i} + bXZ_{i} + cZ_{i} + errors
The bXZ_{i} term measures the interaction effect. The main effect is cZ_{i}.
This term is usually used in an ANOVA context, where its meaning is presumably analogous but this editor has not verified that.

maintained hypothesis
Synonym for 'alternative hypothesis'. "The hypothesis that the restriction or set of restrictions to be tested does NOT hold." Often denoted H_{1}.

Malmquist index
An index number enabling a productivity comparison between economy A and economy B. Imagine that we have an aggregate production function Q_{AA}=f_{A}(K_{A},L_{A}) that describes economy A and an aggregate production Q_{BB}=f_{B}(K_{B},L_{B}) that describes economy B. K and L stand for capital and labor inputs. We substitute the inputs of B into the production function of A to compute Q_{AB}=f_{A}(K_{B},L_{B}). We also compute Q_{BA=}f_{B}(K_{A},L_{A}) with the inputs from country A.
The Malmquist index of A with respect to B is the geometric mean of Q_{AA}/Q_{AB} and Q_{BA}/Q_{BB}. It will be greater than one if A's aggregate production technology is better than B's.

mantissa
Fractional part of a real number.

MAR
a rare abbreviation, for movingaverage representation

March CPS
Also known as the Annual Demographic File. Conducted in March of each year by the Census Bureau in the U.S. Gets the information from the regular monthly CPS survey, plus additional data on work experience, income, noncash benefits, and migration.

marginal significance level
a synonym for 'P value'

market
An organized exchange between buyers and sellers of a good or service.

market capitalization
Total number of shares times the market price of each. May be said of a firm's shares, or of all the shares on an equity market.

market failure
A situation, usually discussed in a model not in the real world, in which the behavior of optimizing agents in a market would not produce a Pareto optimal allocation. Sources of market failures:  monopoly. Monopoly or oligopoly producers have incentives to underproduce and to price above marginal cost, which then gives consumers incentives to buy less than the Pareto optimal allocation.  externalities

market for corporate control
Shares of public firms are traded, and in large enough blocks this means control over corporations is traded. That puts some pressure on managers to perform, otherwise their corporation can be taken over.

market power
Power held by a firm over price, and the power to subdue competitors.

market power theory of advertising
That established firms use advertising as a barrier to entry through product differentiation. Such a firm's use of advertising differentiates its brand from other brands to a degree that consumers see its brand is a slightly different product, not perfectly substituted by existing or potential competitors. This makes it hard for new competitors to gain consumer acceptance.

market price of risk
Synonym for Sharpe ratio.

Markets
In a very abstract sense, a market is where private goods are exchanged.
A good can be anything for which a welldefined property right exists.
This abstract definition is related to the concepts of an
allocation, the
competitive market equilibrium, and an
economic equilibrium in general.
More intuitively, the concept of a market describes the idea that the suppliers
of a product (or a service) meet the demand side, and both sides negotiate over the
price until an optimal combination of price and quantity is reached. Typically,
the supply side offers a higher quantity the higher the price, whereas the
demanded quantity falls the higher the price. In equilibrium, suppliers and consumers
trade at a price at which the supplied quantity equals the demanded quantity.

Markov chain
A stochastic process is a Markov chain if: (1) time is discrete, meaning that the time index t has a finite or countably infinite number of values; (2) the set of possible values of the process at each time is finite or countably infinite; and (3) it has the Markov property of memorylessness.

Markov perfect
A characteristic of some Nash equilibria. "A Markov perfect equilibrium (MPE) is a profile of Markov strategies that yields a Nash equilibrium in every proper subgame." A Markov strategy is one that does not depend at all on variables that are functions of the history of the game except those that affect payoffs. A tiny change to payoffs can discontinuously change the set of Markov perfect equilibria, because a state variable with a tiny effect on payoffs can be part of a Markov perfect strategy, but if its effect drops to zero, it cannot be included in a strategy; that is, such a change makes many strategies disappear from the set of Markov perfect strategies.

Markov process
A stochastic process where all the values are drawn from a discrete set. In a firstorder Markov process only the most recent draw affects the distribution of the next one; all such processes can be represented by a Markov transition density matrix. That is, Pr{x_{t+1} is in A  x_{t}, x_{t1},...} = Pr{x_{t+1} is in A  x_{t}} Example 1: x_{t+1} = a + bx_{t} + e_{t} is a Markov process For a=0, b=1 it is a martingale.
A Markov process can be periodic only if it is of higher than first order.

Markov property
A property that a set of stochastic processes may have. The system has the Markov property if the present state predicts future states as well as the whole history of past and present states does  that is, the process is memoryless.

Markov strategy
In a game, a Markov strategy is one that does not depend at all on state variables that are functions of the history of the game except those that affect payoffs. [Ed.: I believe random elements can be in a Markov strategy: e.g. a mixed strategy could be a Markov strategy.]

Markov transition matrix
A square matrix describing the probabilities of moving from one state to another in a dynamic system. In each ?row? are the probabilities of moving from the state represented by that row, to the other states. Thus the rows of a Markov transition matrix each add to one. Sometimes such a matrix is denoted something like Q(x'  x) which can be understood this way: that Q is a matrix, x is the existing state, x' is a possible future state, and for any x and x' in the model, the probability of going to x' given that the existing state is x, are in Q. (An example would be good here)

Markov's inequality
Quoting almost strictly from Goldberger, 1994, p 31:
If Y is a nonnegative random variable, that is, if Pr(Y<0)=0, and k is any positive constant, then E(Y) ≥ kPr(Y ≥ k).
The proof is amazingly quick. See Goldberger page 31 or Hogg and Craig page 68.

markup
In macro, the ratio of price to marginal cost. Can be used as a measure of market power across firms, industries, or economies.

Marshallian demand function
x(p,m)  the amount of a factor of production that is demanded by a producer given that it costs p per unit and the budget limit that can be spent on all factors is m. p and x can be vectors.

martingale
Same as martingale difference sequence.

martingale difference sequence
** This definition is not usable as is ** A stochastic process {X_{t}} is a martingale (or, equivalently, martingale difference sequence) with respect to information {Y_{t}} if and only if: (i) EX_{t} < infinity (ii) E[X_{n+1}  Y_{0}, Y_{1}, ... , Y_{n}] = X_{n} E(g_{t+1}) = g_{t}.
Martingale differences are uncorrelated but not necessarily independent.

mass production
'A production system characterized by mechanization, high wages, low prices, and largevolume output.' (Hounshell, p.305) Usually refers to factory processes on metalwork, not to textiles or agriculture. The term came into use in the 1920s and referred to production approaches analogous to those of the Ford Motor Company in the US.

Matching Pennies
 Player Two  C  D  Player One  C  1,1  1,1  D  1,1  1,1  A zerosum game with two players. Each shows either heads or tails from a coin. If both are heads or both are tails then player One wins, otherwise Two wins. The payoff matrix is at right.
There is no Nash equilibrium to this game.

Matching pennies
Extremely simplistic, symmetric, two player 2x2 game (which is said to be played by children),
in which each player chooses either Head or Tail. If the choices differ, player 1 pays a dollar
to player 2; if they are the same, player 2 pays player 1 a dollar.
This game does not have an equilibrium in pure strategies, but the unique equilibrium involves
each player selecting one of the two actions with equal probability. The game illustrates that
interactively optimizing behavior may create the need to take actions randomly, in order
not to be predictable by the opponent. For the exact determination of mixed equilibrium strategies,
the assumption of expected utility is important.
For a realworld situation closely resembling this game, think of penalty shooting in sports: both
the goalkeeper and the player who shoots the ball play randomized strategies. They randomize
their actions (left or right, upper corner or not) in a way such that the other player cannot improve
by either action he takes, given the own probabilities of selecting the actions.

Matlab
A matrix programming language and programming environment. Used more by engineers but increasingly by economists. There's a very brief tutorial at Tutorial: Matlab. The software is made by The Mathworks, Inc.

maximin principle
A justice criterion proposed by the philosopher Rawls. A principle about the just design of social systems  e.g., rights and duties. According to this principle the system should be designed to maximize the position of those who will be worst off in it.
"The basic structure is just throughout when the advantages of the more fortunate promote the wellbeing of the least fortunte, that is, when a decrease in their advantages would make the least fortunate even worse off than they are. The basic structure is perfectly just when the prospects ofthe least fortunate are as great as they can be."  Rawls, 1973, p 328

maximum score estimator
A nonparametric estimator of certain coefficients of a binary choice model. Avoids assumptions about the distribution of errors that would be made by a probit or logit model in the same circumstances.
In the econometric model: the dependent variable y_{i} is either zero or one; the regressors X_{i} are multiplied by a parameter vector b. y_{i} often represents which of two choices was selected by a respondent. b is estimated to maximize an objective function that is given by an expression: max_{b} sum_{i=1 to N} [(y_{i}.5)sign(X_{i}b)]
where i indexes observations, of which there are N, and the function sign() has value one if its argument is greater than or equal to zero, and has value zero otherwise.
b chosen this way has the property that it maximizes the correct prediction of y_{i} given the information in X. Notice that although the maximum value of the maximand may be well defined, b is not usually uniquely estimated in a finite data set, because values of b near betahat would make the same predictions. Often, however, b is estimated within a narrow range.

MBO
Stands for Management BuyOut, the purchase of a company by its management. Sometimes means Management By Objectives, a goaloriented personnel evaluation approach.

mean square error
A criterion for an estimator: the choice is the one that minimizes the sum of squared errors due to bias and due to variance.

mean squared error
The mean squared error of an estimator b of true parameter vector B is: MSE(b) = E[(b  B)^{2}] which is also MSE(b) = var(b) + (bias(b))(bias(b)')

measurable
If (S, A) is a measurable space, elements of A are Ameasurable.

measurable space
(S, A) is a measurable space if S is a set and A is a sigmaalgebra of S. Elements of A are said to be Ameasurable.

measure
A noun, in the mathematical language of measure theory: a measure is a function from sets to the real line. Probability is a common kind of measure in economic models. Other measures are the counting measure, which is the number of elements in the set, the length measure, the area measure, and the volume measure. Length, area, and volume are defined along lines, planes, and spaces just as one would expect, and they have the natural meanings. Formally: a measure is a mapping m from a sigma algebra A to the extended real line such that (i) m(null) = 0 (ii) m(B) >= 0 for all B in A (iii) m(any countable union of disjoint sets in A) = the sum of m(each of those sets) The third property is called the countable additivity property. An example: imagine probability mass distributed evenly on a unit square. Probability is then defined on any area within the square. The measure (probability, here) is the size (area) of the subset. The kinds of subsets on which measures such as probability are defined are called sigmaalgebras (which see).

measure theory
measure theory

measurement error
The data used in an econometric problem may have been measured with some error, and if so this violates a basic condition of the abstract environment in which OLS is validly derived. This turns out not to be seriously problematic if the dependent variable is affected by an iid meanzero measurement error, but if the regressors have been measured with a meanzero iid error the estimates can be biased. There are standard approaches to this problem, notably the use of instrumental variables. Paraphrasing from Schennach, 2000, p 1: In a linear econometric specification, a measurement error on the regressors can be viewed as a particular type of endogeneity problem causing the disturbance to be correlated with the regressors, which is precisely the problem addressed by standard IV techniques.

mechanism design
A certain class of principalagent problems are called mechanism design problems. In these, a principal would like to condition her own actions on the private information of agents. The principal must offer incentives for the agents to reveal information. Examples from the theoretical literature are auction design, monopolistic price discrimination, and optimal taxation. In an auction the seller would like to set a price just below the highest valuation of a potential buyer, but does not know that price, and an auction is a mechanism to at least partially reveal it. In a price discrimination, the seller would like to offer the product at different prices to groups with different valuations but may not be able to identify which group an agent is a member of in advance.

Mediator variable
In general, a given variable may be said to function as a mediator to the extend that it accounts for the relation between the predictor and the criterion. Mediators explain how external physical events take on internal psychological significance. Whereas moderator variables specify when certain effects will hold, mediators speak to how or why such effects occur
(Baron & Kenny, 1986, p. 1176).
Path diagram:
IV = independent variable
OV = dependent (outcome) variable
The authors clarify the meaning of mediation, with introducing a path diagram as a model for depicting a causal chain. The basic causal chain involved in mediation is diagramed in the figure above. "This model assumes a threevariable system such that there are two causal paths feeding into the outcome variable: the direct impact of the independent variable (Path c) and the impact of the mediator (Path b). There is also a path from the independent variable to the mediator (Path a).
A variable functions as a mediator when it meets the following conditions:
(a) variations in levels of the independent variable significantly account for the
variations in the presumed mediator (i.e., Path a),
(b) variations in the mediator significantly account for variations in the dependent
variable (i.e., Path b), and
(c) when Paths a and b are controlled, a previously significant relation between
the independent and dependent variable is no longer significant,
with the strongest demonstration of mediation occurring when Path c is zero.
In regard to the last condition we may envisage a continuum.
When Path c is reduced to zero, we have strong evidence for a single,
dominant mediator. If the residual Path c is not zero, this indicates the operation of
multiple mediating factors. Because most areas of psychology, including social, treat
phenomena that have multiple causes, a more realistic goal may be to seek mediators
that significantly decrease Path c rather than eliminating the relation between the
independent and the dependent variables altogether. From a theoretical perspective,
a significant reduction demonstrates that a given mediator is indeed potent, albeit
not both a necessary and a sufficient condition for an effect to occur
(Baron & Kenny, 1986, p. 1176).

medium of exchange
A distinguishing characteristic of money is that it is taken as a medium of exchange, that is, in the language of Wicksell (1935) p. 17, that it is "habitually, and without hesitation, taken by anybody in exchange for any commodity."

meet
Given a space of possible events, the meet is the finest common coarsening of the information sets of all the players. The meet is the finest partition of the space of possible events such that all players have beliefs about the probabilities of the elements of the partition.

mesokurtic
An adjective describing a distribution with kurtosis of 3, like the normal distribution. See by contrast leptokurtic and platykurtic.

metaproduction function
Means bestpractice production function  depending on context, either the most efficient feasible practice, or most efficient actual practice of the existing entities converting inputs X into output y. Often in practice y is an agricultural output, and data from a sample of farms, and the metaproduction function could be estimated by estimating production functions for the farms and choosing among the most efficient ones. In the (macro) context of the quote below, the entities are not farms but countries, producing GDP. 'The term 'metaproduction function' is due to Hayami and Ruttan (1970, 1985). For an exposition of the metaproduction function approach, see Lau and Yotopoulos (1989) and Boskin and Lau (1990).... The two most important maintained hypotheses [of this approach] are: (1) that the aggregate production functions of all countries are identical in terms of 'efficiencyequivalent' units of output and inputs; and (2) that technical progress in all countries can be represented in the commodityaugmentation form, with constant geometric augmentation factors....' The framework allows 'the researcher to consider and potentially to reject the maintained hypotheses of traditional growth accounting [such as] (1) constant returns to scale, (2) neutrality of technical progress; and (3) profit maximization.' (p66) An assumption related to the second maintained hypothesis above, which the theory depends on (p69) is that 'the measured outputs and inputs of the different countries may be converted into unobservable standardized, or 'efficiencyequivalent,' quantities of output and inputs by multiplicative country and output and inputspecific timevarying augmentation factors....' (where 'timevarying' seems to conflict with the requirement, above, that the augmentation factors be 'constant'.) (p69) In this approach 'countries may differ in the quantities of their factor inputs and intensities and possibly in the qualities and efficiencies of their inputs and outputs, but they do not differ with regard to the technological opportunities .... [T]hey are assumed to have equal access to technologies.' From p66, 69, 73 of Lau (1996).

metatheorem
An informal term for a proposition that can be proved in a class of economic model environments.

method of moments
A way of generating estimators: set the distribution moments equal to the sample moments, and solve the resulting equations for the parameters of the distribution.

MFP
Abbreviation for Multifactor productivty.

MGF
stands for 'moment generating function', which see.

Microeconomics
Microeconomics is the analysis of individual economic units and their interactions.
It includes the theories of the consumer (i.e., of households),
the producer (i.e., firms), and the markets in which they interact. The tools of
microeconomic analysis are also employed in other fields, such as the theory of optimal
taxation in public economics. Microeconomics is often contrasted with macroeconomics
which is concerned with economic aggregates, such as aggregate consumption and
aggregate production, unemployment, or with economic growth in general.

Minitab
Data analysis software, discussed at http://www.minitab.com.

Mixed strategy
In contrast to pure strategies, mixed strategies are strategies that
involve random draws. A mixed strategy is a probability distribution over
a player's (pure) strategies. For example in penalty shooting, the goalgetter
typically does not expect the goalkeeper to jump to the right corner for
sure, but he regards the goalkeeper's behaviour as a mix between the pure
strategies "jump to the right" and "jump to the left".

mixing
In the context of stochastic processes, events A and B (that is, subsets of possible outcomes of the process) "are mixing" if they are asymptotically independent in the following way.
Let L be a lag operator that moves all time subscripts back by one (e.g. replacing t by t1). Iff A and B are mixing, then taking the limit as h goes to infinity: lim Pr(A intersected with L^{h}B) = Pr(A)Pr(B).
The event L^{h} is the event B, but h periods ago; it's NOT some kind of stochastic ancestor of B.
If two events are independent, they are mixing. If two events are mixing, they are ergodic.
I *believe* that a stochastic process is mixing iff all pairs of possible values it can take, taken as events, are mixing.

MLE
maximum likelihood estimator

MLRP
Abbreviation for monotone likelihood ratio property of a statistical distribution.

models
Generally means theoretical or structural models. Can also mean econometric models which in this glossary are listed separately. models

Models of microeconomic decisions
Economics proceeds by building models of behavior. These models are supposed to
be simplified representations of reality which specify how variables in a system
relate to each other. Economists use many techniques in the construction and analysis
of economic models, but most of the techniques fall into the categories of optimization
analysis and equilibrium analysis.
Nearly all models of individual behavior in microeconomics are models of optimizing
behavior which can broadly be interpreted as rational behavior.
In building a model of behavior, economists are naturally led to identify agents that
make the choices, the kinds of choices that are feasible for them, how the choices of
other agents constrain them, and so on. Once the economist is able to write down an
optimizing problem describing an economic choice, he or she can apply the standard
mathematical methods of microeconomic analysis.
Once we have understood the nature of the optimal choice problem facing individual agents,
we can investigate how these choices fit together. In general, some of the variables
that influence a given agent´s behavior ? such as prices ? will be
determined, at least in part, by the behavior of other agents. An economic
equilibrium
is a situation of consistent optimal choices: No agent has an incentive to change any of
his choices, given his perceptions of the behavior of other agents.

Moderator variable
Moderator variables are important, because specific factors (e.g. context information)
are often assumed to reduce or enhance the influence that specific independent variables
have on specific responses in question (dependent variable).
Specifically within a correlational analysis framework, a moderator is a third variable that
affects the zeroorder correlation between two other variables. For example,
Stern, McCants & Pettine (1982)
found that the positivity of the relation between changing life events and
severity of illness was considerably stronger for uncontrollable events (e.g., death of a
spouse) than for controllable events (e.g., divorce). A moderator effect within a
correlational framework may also be said to occur where the direction of the correlation
changes. Such an effect would have occured in the Stern et al. study if controllable life
changes had reduced the likelihood of illness, thereby changing the direction of the relation
between lifeevent change and illness from positive to negative.
In the more familiar analysis of variance (ANOVA) terms, a basic moderator effect can be
represented as an interaction between a focal
independent variable
and a factor that specifies the appropriate conditions for its operation"
(Baron & Kenny, 1986, p. 1174).

modernization
Quoting from Landes: "Modernization comprises such developments as urbanization (the concentration of the population in cities that serve as nodes of industrial production, administration, and intellectual and artistic activity); a sharp reduction in both death rates and birth rates from traditional levels (the socalled demographic transition); the establishment of an effective, fairly centralized bureaucratic government; the creation of an educational system capable of training and socializing the children of the society to a level compatible with their capacities and best contemporary knowledge; and, of course, the acquisition of the ability and means to use an uptodate technology."

ModiglianiMiller theorem
that the total value of the bonds and equities issued by a firm in a model is independent of the number of bonds outstanding or their interest rate.
The theorem was shown by Modigliani and Miller, 1958 in a particular context with no fixed costs, transactions costs, asymmetric information, and so forth. Analogous theorems are shown in various contexts. The assumptions made by such theorems offer a way of organizing what it would be that makes corporations choose to offer various levels of bonds. The choice of numbers and types of bonds and stocks a corporation offers is the choice of capital structure. Among the factors affecting the capital structure of a firm are taxes, bankruptcy costs, agency costs, signalling, bargaining position in litigation, and differences between firms and investors in access to capital markets.

momentgenerating function
Denoted M(t) or M_{X}(t), and describes a probability distribution. A momentgenerating function is defined for any random variable X with a pdf f(x). M(t) is defined to be E[e^{tX}], which is the integral from minus infinity to infinity of e^{tX}f(x). A use for these is that the t^{th} moment of X is M^{(t)}(0), that is the t^{th} derivative of M() at zero.

monetarism
The view that monetary policy is a prime source of the business cycle, and that the time path of the money stock is a good index of monetary policy. As presented by Milton Friedman and Anna Schwartz, monetarism emphasizes the relation between the level of the money stock and the level of output without a detailed theory of why changes in the money stock are not neutral in the short run. Later versions posed an explicit basis for noneutrality in the form of barriers to information flow about prices.
In policy terms monetarists, notably Friedman, advocated a monetary rule, that is, a steady growth in the money supply to match economic growth, without allowing central banks room for discretion. If the rule is credible, public expectations of inflation be low, and thus inflation itself, if high, would fall almost immediately.

monetarist view
In extreme form: that only the quantity of money matters by way of aggregate demand policy. Relevant only in an overheated economy (Branson p 391).

monetary base
In a modern industrialized monetary economy, the monetary base is made up of (1) the currency held by individuals and firms and (2) bank reserves kept within a bank or on deposit at the central bank.

monetary regime
"A monetary regime can be thought of as a set of rules governing the objectives and the actions of the monetary authority."
Examples: (1) "A gold standard is one example of a monetary regime  the monetary authority is obligated to maintain instant convertibility between its liabilities and the gold. Th monetary authority may have considerable room to maneuver in that monetary regime, but it can do nothing that would cause it to violate its commitment." (2) "The same remarks would apply to a monetary regime obligating the monetary authority to maintain a fixed exchange rate between its own and another currency." (3) "A monetary regime of a very different sort could be based on a Monetarist rule specifying the rate of growth of some monetary aggregate. The basic distinction is between regimes based on a convertibility or redemption principle and those based on a quantity principle."

monetary rule
See the policy discussion in monetarism.

monetized economy
A model economy that has a medium of exchange: money

money
A good that acts as a medium of exchange in transactions. Classically it is said that money acts as a unit of account, a store of value, and a medium of exchange. Most authors find that the first two are nonessential properties that follow from the third. In fact, other goods are often better than money at being intertemporal stores of value, since most monies degrade in value over time through inflation or the overthrow of governments.
Theory: Ostroy and Starr, 1990, p. 25, define money in certain models "as a commodity of positive price and zero transaction cost that does not directly enter in production or consumption."
History: See this Web site on the History of Money.
Related terms: money

money illusion
'the belief that money [that is, a particular currency] represents a constant value'

moneyintheutilityfunction models
A modeling idea. In a basic ArrowDebreu general equilibrium there is no need for money because exchanges are automatic, through a 'Walrasian auctioneer'. To study monetary phenomena, a class of models was made in which money was a good that brought direct utility to the agent holding it; e.g., a utility function took the form u(x,m) where x is a vector of other commodities, and m is a scalar quantity of real money held by the agent. Using this mechanism money can have a positive price in equilibrium and monetary effects can be seen in such models. Contrast 'cashinadvance constraint' for an alternative approach.

monopoly
If a certain firm is the only one that can produce a certain good, it has a monopoly in the market for that good.

monopoly power
The degree of power held by the seller to set the price for a good. In U.S. antitrust law monopoly power is not measured by market share. (Salon magazine, 1998/11/11)

monopsony
A state in which demand comes from one source. If there is only one customer for a certain good, that customer has a monopsony in the market for that good. Analogous to monopoly, but on the demand side not the supply side. A common theoretical implication is that the price of the good is pushed down near the cost of production. The price is not predicted to go to zero because if it went below where the suppliers are willing to produce, they won't produce. Market power is a continuum from perfectly competitive to monopsony and there's an extensive practice/industry/science of measuring the degree of market power.
Examples: For workers in an isolated company town, created by and dominated by one employer, that employer is a monopsonist for some kinds of employment. For some kinds of U.S. medical care, the government program Medicare is a monopsony.

monotone likelihood ratio property
A property of a set of pdfs which is assumed in theoretical models to characterize risk and uncertainty because it makes more conclusions feasible and is often plausible.
Example: Let e ('effort') be an input variable into a stochastic production function, and y be the random variable that represent output. Let f(y  e) be the pdf of y for each e. Then the statement that f() has the monotone likelihood ratio property (MLRP) is the same as the statement that: for e2>e1, f(ye2)/f(ye1) is increasing in y. This says that output is positively related to effort, and something stronger, something like: of two outcomes or ranges of outcomes, the worse one will not become relatively more likely than the better one if effort were to rise. By relatively more likely is meant that the likelihood ratio, above, rises.
The set of pdfs for which the MLRP is assumed above is the set of f()'s indexed by values of e. Each holds that specified relationship to the others. In practice the MLRP assumption tends to rule out multimodal classes of distributions, and this is its main effect. (By multimodal we mean those with multiplepeaked pdfs.)
Normally e is scalar, taking on either discrete or continuous sets of values. An analogous definition, for a multidimensional (vector) e, is feasible. Whether it is used in existing models is not known to this author.

monotone operator
An operator that preserves inequalities of its arguments. That is, if T is a monotone operator, then: (i) iff x>y, then Tx>Ty, and iff x<y, then Tx<Ty.
Same basic meaning as monotone transformation.
The most common monotone operator is the natural log function. For example in maximum likelihood estimation, one usually maximizes the log of the likelihood function, not the likelihood function itself, because this is more tractable and the log is a monotone operator so it doesn't change the answer.

monotone transformation
A transformation that preserves inequalities of its arguments. That is, if T is a monotone transformation, then: (i) iff x>y, then Tx>Ty, and iff x<y, then Tx<Ty.
Same basic meaning as monotone operator.

Monte Carlo simulations
These are data obtained by simulating a statistical model in which all parameters are numerically specified.
One might use Monte Carlo simulations to test how an estimation procedure would behave, for example under conditions when exact analytic descriptions of the performance of the estimation are not algebraically feasible, or when one wants to verify that one's analytic calculation for a confidence interval is correct.

MoorePenrose inverse
Same as pseudoinverse.

morbidity
Incidence of ill health. It is measured in various ways, often by the probability that a randomly selected individual in a population at some date and location would become seriously ill in some period of time. Contrast to mortality.

mortality
Incidence of death in a population. It is measured in various ways, often by the probability that a randomly selected individual in a population at some date and location would die in some period of time. Contrast to morbidity.

MSA
Same as SMSA.

MSE
mean squared error (which see)

multifactor productivity
Same as total factor productivity, a certain type of Solow residual.
MFP = d(ln f)/dt = d(ln Y)/dt  s_{L}d(ln L)/dt  s_{K}d(ln K)/dt where f is the global production function; Y is output; t is time; s_{L} is the share of input costs attributable to labor expenses; s_{K} is the share of input costs attributable to capital expenses; L is a dollar quantity of labor; K is a dollar quantity of capital.

multinomial
In the context of discrete choice models, multinomial means there are more than two possible values of the dependent variable, the choice, which is a scalar.
For specific constructions see multinomial logit and multinomial probit.

multinomial logit
Relatively easy to compute but has the problematic IIA property by construction. Multinomial probit with correlation between structural residuals does not suffer from the IIA problem but is computationally expensive. (Ed.: I don't know why the IIA problem gets sucked into this when the actual different between logit and probit is the functional form.) Multinomial logit is available in more software packages than is multinomial probit.

multinomial probit
Multinomial probit with correlation between structural residuals does not suffer from the IIA problem but is computationally expensive. Multinomial logit which solves a similar problem is relatively easy to compute but has the problematic IIA property by construction. (Ed.: I don't know why the IIA problem gets sucked into this when the actual different between logit and probit is the functional form.) Multinomial logit is available in more software packages than is multinomial probit.

multivariate
A discrete choice model in which the choice is made from a set with more than one dimension is said to be a multivariate discrete choice model.

MundellTobin effect
That nominal interest rates would rise less than oneforone with inflation because in response to inflation the public would hold less in money balances and more in other assets, which would drive interest rates down.

mutatis mutandis
"The necessary changes having been made; substituting new terms."

MVN
An abbrevation for 'multivariate normal' distribution.
