Margaret Slade
Research Area
About
I am a professor emeritus in the Vancouver School of Economics at The University of British Columbia. I taught at UBC for 20 years before going to the UK to become the Leverhulme Professor of Industrial Economics at the University of Warwick. After six years in the UK, I retired and returned to Canada and the VSE.
My research is concerned with applied problems in Industrial Organization, especially those requiring the analysis of data. I am interested in both horizontal (same product market) and vertical (buying and selling) relationships and policies. I have advised government agencies in numerous countries as well as private parties.
Prior to becoming an economist, I worked as a mathematician for Shell Development, the US Geological Survey, and the Department of Scientific and Industrial Research of New Zealand.
I obtained my Ph.D. from The George Washington University in Washington, DC, USA.
Teaching
Research
No-poaching clauses (NPCs) have recently come under scrutiny due to their potentially anti-competitive impact on wages. However, they can also enhance efficiency. We use data from the US chain restaurant industry to assess the effect of such clauses on wages and we find robust evidence of a negative effect. Specifically, we find that the legal cases, proposed legislation, and negative attention surrounding NPCs, which led many chains to remove such clauses from their contracts, caused wages in those chains to rise by 5-6% relative to chains that never imposed NPCs. We attribute this effect to the removal of frictions and barriers to mobility in labor markets.[go to paper]
COVID–19 has affected all industrial sectors and regions of the world and, in many cases, the impact has been devastating. Nevertheless, there are large differences across sectors and regions, with a few winners as well as many losers. This paper assesses the impact of the virus on Canadian industries and provinces in the short and medium run. In particular, time–series forecasting models for 44 three–digit industries in each of the ten provinces are estimated. Those models, which answer the question: What would the economy have looked like absent the virus? are used to predict post–lockdown losses — percentage differences between forecast and observed output. Econometric analysis of the estimates is then used to quantify and rank industries and provinces in terms of short and medium term losses and
gains.[go to paper]
Vertical mergers have attracted much attention in recent years. In this paper, we assess the role of presumptions and likelihoods in vertical merger analysis and guidelines. We focus in particular on the role that we believe statistical evidence in general, and retrospective analyses more specifically, should play in determining presumptions. We also discuss how horizontal merger guidelines provide frameworks to analyze the horizontal issues that can be associated with vertical mergers. We conclude that while some vertical mergers may raise concerns, the evidence at this point does not provide sufficient guidance to develop presumptions related to strictly vertical issues.
[go to paper]
This article surveys recent empirical evidence on efficiencies and competitive harm that are associated with vertical mergers. It discusses both ex post or retrospective empirical studies that rely on post-merger data and ex ante or forecasting techniques that use pre-merger data. It develops the idea that, although there is a need for vertical merger screening tools, there are a number of problems that are associated with attempts to adapt horizontal screens to the vertical context. Mergers in the technology, media, and telecom sectors are emphasized because they tend to dominate contested vertical mergers.
[go to paper]
A Longstanding economic proposition predicts that vertical integration between companies with market power can benefit both consumers and firms by eliminating the “double marginalization” that characterizes pricing by the firms separately. Too often, however, this proposition is cited without sufficient attention to its underlying assumptions, leading to the possibility of its misuse in antitrust and regulatory proceedings involving vertical integration. This note outlines the strong assumptions of the basic model, explains how the conclusions differ when those assumptions do not hold, and analyzes certain unique settings that require special analysis. It cautions that policy should pay closer attention to traditional supply and organization efficiencies of vertical integration rather than the double marginalization scenario.
[go to paper]
The Sécretan Syndicate controlled 80% of world copper production in the late 1880s. The downstream firms set the price and contracted for all of the output of the leading producers. Despite those restraints, the Syndicate was short lived. This paper describes the cartel and the role of the restraints in disciplining the upstream firms. It also quantifi es the effect of the vertical restraints on supply, demand, and inventory holdings. Finally, it takes a broader look at commodity market corners | why they are attempted and why they usually fail | and their implications for antitrust policy.
The standard real–options model predicts that increased uncertainty discourages investment. When projects are large and take time to build, however, that prediction can be reversed. We investigate the investment/uncertainty relationship empirically using historical data on opening dates of new U.S. copper mines — large, irreversible projects with substantial construction lags. Both the timing of the decision to go forward and the price thresholds that trigger that decision are assessed. In particular, we build upon a reduced form analysis to construct a structural model of entry. We find that, in this market, greater uncertainty encourages investment and lowers the price thresholds for many mines.
We show how production and pricing data can be used to estimate merger-related efficiencies using pre merger data. We assess the changes in static and dynamic firm efficiency and marginal costs that are expected to occur post merger. To do this, we jointly estimate firm level returns to scale, technical change, TFP growth, and price-cost markups. We implement our empirical model using data for the North American (US and Canadian) brewing industry, and we use the estimated model of firm technology to evaluate the changes that are associated with the merger between Molson and Coors. We forecast nontrivial increases in returns to scale and declines in marginal costs and we verify those results by analyzing the impact of the merger retrospectively using post merger data. [go to paper]
We review important developments in Empirical Industrial Organization (IO) over the last three decades. The paper is organized around six topics: collusion, demand, productivity, industry dynamics, interfirm contracts, and auctions. We present models that are workhorses in empirical IO, and describe applications. For each topic, we discuss at least one empirical application using Canadian data.
Using US firm level panel data we simultaneously assess the contributions to productivity of three potential sources of research and development spillovers: geographic, technological, and product market (“horizontal”). To do so, we construct new measures of geographic proximity based on the distribution of a firms inventor locations as well as its headquarters. We find that geographic location matters more for productivity than the other spillover mechanisms with both intra- and inter-regional spillovers being important. The geographic location of a firms researchers is much more important than its headquarters. These benefits may be the reason why local policy–makers compete so hard for the location of local R&D labs and high tech workers.
See policy column at VoxEU, here
Using a newly constructed data set, this paper explores the forces that led to the rise and fall of the U.S. copper industry. A number of factors are evaluated including depletion of the resource base, cost–lowering technical change, formation of beliefs concerning profitability concentration of ownership into a few hands, and changes in productivity and ore grades. I find that entry patterns were principally determined by the introduction of the steam shovel, which enabled open–pit mining. Although other technological factors profoundly changed the type of mine that entered, their effect on entry was weaker. Revisions of expectations of success (mining rushes) were statistically significant but economically less important. Contrary to conventional wisdom, I find that concentration of ownership encouraged entry. Finally, resource depletion was primarily responsible for the decline in entry in later years. However, I argue that the same forces that contributed to success also led to decline.
The purpose of this paper is threefold. First, we give an overview of the general direction the spatial econometrics literature has taken without attempting to provide a representative survey of all interesting work that has appeared. Second, we identify a number of problems in spatial econometrics that are as yet unresolved. Finally, we provide advocacy for the notion that new spatial econometric theory should be inspired by actual empirical applications as opposed to being directed by what appears to be the most obvious extension of what is currently available."]
We review the empirical literature that extends and tests the Hotelling (1931) model of the optimal depletion of an exhaustible resource. The theory is briefly described to set the stage for the review of empirical tests and applications. Those tests can be roughly divided into two broad categories—descriptive and structural—and we discuss the strengths and weaknesses of each before presenting the empirical studies of optimal extraction under conditions of exhaustibility. We also discuss some econometric pitfalls that applied researchers face when attempting to test the model."]
I discuss the use of simulation techniques to evaluate unilateral effects of horizontal mergers and the pitfalls that one can encounter when using them. Simple econometric models are desirable because they can be implemented in a short period of time and can be understood by non experts. Unfortunately, their predictions are often misleading. Complex models are more reliable but they require more time to implement and are less transparent. The use of merger simulations and the sensitivity of predictions to modeling choices is illustrated with an application to mergers in the UK brewing industry. There have been a number of brewing mergers that have changed the structure of the UK market, as well as proposed but unconsummated mergers that would have had even more profound effects. I assess two of them: the successful merger between Scottish&Newcastle and Courage and the proposed merger between Bass and Carlsberg–Tetley."]
We provide a new central limit theorem (CLT) for spatial processes under weak conditions which, unlike existing results, are plausible for most economic applications. In particular, our CLT is designed for problems that have some, but not necessarily all, of the following features: i) Agents choose the locations of observations to maximize profits, welfare, or some other objective. ii) The objects that are chosen (e.g., stores or brands) interact with one another. For example, they can be substitutes or complements. iii) Interaction can be complex. In particular, interaction between i and j need not depend only on the distance between the locations of i and j, but can also depend on distance to or location of other observations k, or possibly on the number of other such observations.
We propose a semi–structural discrete–choice model that can be used to estimate static or dynamic decision rules. By semi–structural we mean that our estimator is motivated by a structural model, and we estimate the choice rules that are associated with that model. However, we do not uncover the underlying structural parameters. Our estimator can be used to advantage when games involve a rich set of choices that might be correlated across decision makers and when the object is to uncover the effect of historic behavior (on, for example, prices, sales, and profits). We apply our discrete–choice estimator to study decision rules for a dynamic game of price and advertising competition.
We show that the one–step (‘continuous updating’) GMM estimator is consistent and asymptotically normal under weak conditions that allow for generic spatial and time series dependence. We use the new procedure to estimate a dynamic spatial discrete–choice model with fixed effects that enables us to study operating decisions for mines in a real–options context. We find that the data are more supportive of a mean/variance–utility model than of a real–options model.
In this paper, we assess how the characteristics of product and forward markets affect commodity–price distributions. In particular, we assess the levels and volatilities (means and standard deviations) of the spot prices of the six commodities that were traded on the London Metal Exchange in the 1990s. The theories that we examine can be grouped into four classes. The first considers how product–market structure and forward–market trading jointly affect the spot–market game, the second explores the links between product–market structure and spot–price stability, the third assesses whether forward trading destabilizes spot prices, and the last relates the arrival of new information to price volatility and the volume of trade. We find support for traditional market–structure models of the price level but not of price stability. In addition, increased forward trading is associated with lower prices. Finally, although we find a positive relationship between increased trading and price instability, the link appears to be indirect via a common causal factor.
This paper analyzes the role that economic space plays in private and public decision making. Both geographic and characteristic space are considered. The choice of spatial location, whether it be a physical location or a product position, can have significant consequences for other economic decisions, and those secondary decisions are the focus of the paper. In particular, it is useful to have methods of measuring economic proximity in geographic and characteristic space and to have ways of determining the implications of that proximity for market outcomes. I summarize some of the methods that my coauthors and I have developed for estimating economic distance and for assessing its implications.
In this paper, I look at four models of firm profitability: two taken from Industrial Organization, one from Finance, and one from the Economics of Exhaustible Resources. Only one predicts that there will be a positive relationship between firm profitability and the structure of the market in which the firm operates, and only that one views high profits as an indication of monopoly power. Nevertheless, most antitrust authorities base their policies on a belief in those relationships. Using panel data from 14 nonferrous–metal mining and refining markets, I find strong empirical support only for the market–structure model.
Mergers in the UK brewing industry have reduced the number of national brewers from six to four. The number of brands, in contrast, has remained relatively constant. We analyze the effects of mergers on brand competition and pricing. Brand-level demand equations are estimated from a panel of draft beers. To model brand-substitution possibilities, we estimate the matrix of cross-price elasticities semiparametrically. Our structural model is used to assess the strength of brand competition along various dimensions and to evaluate the mergers. In particular, we compute equilibria of pricing games with different numbers of players.
[go to paper] [go to Appendix - Mergers, Brand Competition, and the Price of a Pint]
Market power and joint dominance are examined in U.K. brewing. I assess unilateral andcoordinated effects, where the latter is equated with joint dominance, and show how one can distinguish between the two econometrically. The application makes use of two demand equations: the nested logit of McFadden [1978a] and the distance-metric of Pinkse, Slade, and Brett [2002]. The two equations yield very different predictions concerning elasticities and markups. Nevertheless, although there is evidence of market power using either demand model, that power is due entirely to unilateral effects. In other words, neither model uncovers evidence of coordinated effects (tacit collusion).
We examine theoretical predictions and econometric evidence concerning franchise contracting and sales-force compensation and suggest a number of factors that ought to influence the contracts that are written between principals and agents. For each factor, we construct the simplest theoretical model that is capable of capturing what we feel to be its essence. The comparative statics from the theoretical exercise are then used to organize our discussion of the empirical evidence, where the evidence is taken from published studies that have attempted to assess each factor's effect on the power of agent incentives. We also discuss theoretical issues and empirical results pertaining to a few topics that have been addressed in the literature but that do not fit easily into our simple modeling framework.
The value of managerial flexibility is assessed empirically using data on prices, unit costs, ore extraction, grade, reserves, and metal output for a panel of twenty one Canadian copper mines. A real-option model is estimated and solved to yield the value of the project as well as the option value that is associated with flexible operation. Most previous empirical researchers consider the initial-investment decision but neglect the possibility of flexible operation thereafter. Moreover, although they assume that price is stochastic, they ignore cost and reserve uncertainty. Finally, they usually model price as a geometric Brownian motion or other nonstationary process. Transition equations for three state variables, copper price, unit cost, and remaining reserves, are estimated here. Differences in assumptions (e.g., flexible vs. inflexible operating policies, stochastic vs. deterministic state variables, and mean reverting vs. nonstationary stochastic processes) are found to lead to large differences in estimated project and option values.