1.Explain the difference between the results presented in Tables 1 and 2. How does the difference in difference estimator establish the main result?
2. Explain each of the robustness results presented in Table 5. Explain the result of the quantile regressions in Table 6. How do the quantile results relate to Figures 1 and 2?
3. Describe in detail the basis for the 3rd and 4th arguments supporting collusion as an explanation for the main results.
4. Describe in detail the basis for the 5th and 6th arguments supporting collusion as an explanation for the main results.
Instructions ECON7012/8012 – Topic 6b Retail grocery markets A/Prof Jordi McKenzie, Department of Economics Paper Details • Genakos, C., Koutroumpis, P., & Pagliero, M. (2018), ‘The impact of maximum markup regulation on prices’, The Journal of Industrial Economics, 66(2), 239-300. Abstract Markup regulation is a common, yet understudied type of regulation. We analyze the repeal of maximum wholesale and retail markup regulation in an oligopolistic and vertically nonintegrated market. By comparing the prices of products affected by regulation before and after the policy change and using unregulated products as a control group, we find that abolishing regulation led to a significant decrease in both retail and wholesale prices. Our analysis provides indirect but consistent evidence that markup ceilings provided a focal point for collusion among wholesalers. Introduction • State imposed markups are common across markets ― 60 % of low and middle-income countries regulate wholesale or retail markups in the pharmaceutical industry • Governments justify markups on grounds of protecting consumers ― If binding, markup ceilings force some firms to reduce prices ― If not binding, prices will not be affected ― Hence, the average price is expected to fall • No evidence on actual impact of markup regulation on prices ― Existing studies on the impact of price regulation challenge this seemingly uncontroversial prediction ― Maximum markup regulation may also have this perverse effect ― However, does not imply markup regulation will have similar effect Introduction • This paper examines the impact of maximum markup regulation on retail and wholesale prices in an oligopolistic and vertically nonintegrated market • Take advantage of repeal of maximum markup regulation in the market for fresh fruits and vegetables in Greece June, 2011 ― Regulation consisted of maximum wholesale and retail markups on virtually all fruits and vegetables (imported or locally produced) ― Nonetheless, five products – apples, lemons, mandarins, oranges and pears – were exempt from regulation • To identify impact of deregulation, compare prices affected by regulation before and after policy change, using unregulated products as control group Preview of Results • Using difference in difference (DiD) methodology, find that abolishing markup regulation led to a 6% drop in average retail prices ― Find deregulation had direct effect on wholesalers and only indirectly affected retailers, who adjusted prices in reaction to wholesale prices • The retail and wholesale prices of goods in the control group were unaffected • Also find price dispersion increased (particularly at bottom of distribution) as a consequence of deregulation in retail and wholesale market Preview of Results • Given maximum markup regulation is typically instituted to protect consumers, surprising to observe a decrease in prices after deregulation ― Not consistent with markup regulation’s having the sole effect of constraining firms with higher markups ― Results point to large and unexpected effects of regulation that significantly impact consumers • Second part of the paper investigate hypothesis that maximum markups served as a focal point for collusion and explore alternative hypotheses • Number of factors facilitating collusion are present in the wholesale market 1. Operates as licensed market, with small stable group of operators 2. Wholesalers operate in close proximity and trade (homogeneous products) on daily basis, making it possible to share information Markup Regulation • Markup regulation constrains distribution of markups and only indirectly affects the distribution of prices • Consider a market in which firms with heterogeneous marginal costs c are selling differentiated products at prices p ― Where a is the maximum markup • If regulation is binding for some firms, then expected price will fall after the introduction of regulation • Hence, first prediction is after deregulation see an increase in average prices Markup Regulation • Markup regulation is expected to reduce price variability if regulation is more binding for firms with higher marginal costs ― In other words, if there is a positive correlation between marginal costs, markups and prices • Assumption is realistic for fruits and vegetables ― higher marginal costs typically correspond to a higher quality product (e.g., taste and freshness, quality of service), which makes it possible to charge higher markups (more rigid demand). • Second prediction is an increase in price variability after deregulation, driven mainly by a movement in the right tail of price distribution • Markup regulation differs from price regulation: 1. Markup regulation is used in markets (for example, fruits and vegetables) in which price regulation would be impossible to implement due to high seasonality and uncertainty in production costs (e.g. caused by weather conditions) 2. Markup regulation limits benefits from cost reductions, while price regulation provides strong incentives to increase efficiency 3. Enforcement of markup regulations is generally more difficult, as firms may misreport or distort cost structure Markup Regulation vs. Price Regulation • The market for fruits and vegetables in Greece consists of three tiers ― At production level, market is very fragmented. ― The wholesale market is significantly more concentrated, with the Central Market operating as a closed (licensed) market ―Wholesalers mainly sell to retailers (supermarkets), but also to street sellers, grocery stores, and restaurants. ― At retail level, consumers buy either from street markets (58% in 2011 but declining), supermarkets (32% but increasing), and grocery stores or other corner shops (10%) • Until 2011, law provided for product-specific maximum markups ranging 8- 12% for the wholesale market, 20-35% for supermarkets, and 17-32% for street markets and grocery stores The Greek Market for Fruits and Vegetables Data • Dataset consists of three types of data: 1. First, weekly store level retail prices for each fruit and vegetable product category both from supermarkets and street markets in Athens, Greece • Sample covers one and a half years before and after the policy change, from 4 January, 2010, to 28 December, 2012 2. Median monthly wholesale fruit and vegetable prices from the Athens Wholesale Central Market 3. Weekly store-specific retail prices for 19 non-fruit and vegetable products sold in supermarkets in Athens during the same period Retail Prices Wholesale Prices Distribution of Retail Prices Distribution of Wholesale Prices • Identification is obtained within a difference in difference framework • Denote by Pijt the retail price of product variety i, in store j, in week t • The basic empirical specification is of the form: ― Postt is an indicator variable equal to one after deregulation ― Treati is an indicator variable for products affected by the regulation (treatment group) ― Postt × Treati denotes their interaction ― Xijt is a matrix of control variables ― eijt is a random shock with E(eijt |Postt,Treati,Xijt)=0 • b3 is crucial parameter capturing impact of policy change Identification and Empirical Methodology • Key identifying assumption is that price trends would be same (conditional on covariates Xijt) in treatment and control groups in absence of treatment • Assumption becomes increasingly credible as add appropriate controls in Xijt 1. Control for changes in the VAT rates 2. Include eleven month indicator variables, 53 store indicator variables, and 109 product variety-specific indicators in Xijt 3. Add interaction of month and product fixed effects, capturing the yearly price cycle of each product (we assume that varieties of the same product follow the same cycle) 4. Include a quadratic trend (measured in months) • Analysis of wholesale prices from Central Market uses same specification, with caveat that only median wholesale prices at a monthly frequency Identification and Empirical Methodology Results: Treatment Only Results: DiD Prices Results: DiD Wholesale Prices • Control group may be contaminated as contain substitutes/complements • Consider different control group of 19 non-fruit and vegetable goods Results: DiD Alternative Control Group Robustness Checks Results: Quantile Regressions • Dependent variable is residuals of regression of log of retail/wholesale price of product variety i, in store j, and week t on store, product, month × product fixed effects and a linear and quadratic trend measured in months Results: Wholesale Price Range • Monthly relative wholesale price range for each product, (maxit−minit)∕minit Results: Wholesale Price Minimum Results: Wholesale Price Maximum • Deregulation not specific to one or a small set of products • 86% of products have negative coefficient Results: Product Specific (Postt xTreati) • Negative impact of deregulation on retail and wholesale prices not consistent with the view that regulation is constraining firms with high markups • Main candidate explanation is that regulation facilitated collusive behavior ― Economic intuition underlying this idea is that (unconstrained) firms used the maximum markups as focal points for coordination, leading to increases in average prices ― Repeal of the law might have destroyed these focal points and led to significant price decreases Interpretation of the Estimated Impact • Collusion on focal markups makes it necessary for firms to infer their competitors’ markups from available data on costs and prices • Three main reasons to believe monitoring was possible: 1. Identity of (large) customers supplied by each wholesaler could be observed because of physical arrangement of Central Market 2. Wholesale transactions were far from confidential, although they were subject to negotiation between wholesalers and (large) buyers 3. Information on retail prices in supermarkets was widely available to competitors • Empirically test six different implications of the collusion hypothesis Interpretation of the Estimated Impact Arguments to Support Collusion 1. Collusion and Price Dispersion ― To the extent that prices and markups are positively correlated, the strong movement in the tail of price distribution is consistent with breakdown of collusion 2. Pass-Through Regressions ― Expect larger impact in markets where collusion easier to maintain ― Central Market has many characteristics favorable to collusion ― Pass-through regressions disentangle direct impact of the policy on the distribution of retail prices from the indirect impact through the effect on the wholesale price distribution (Table V, cols 3&4) ― Deregulation affected retail prices indirectly through wholesale prices, but no evidence of direct effect of deregulation on retail prices Arguments to Support Collusion 3. The Heterogeneous Impact of the Reform in Supermarkets and Street Markets ― That the effect of reform originated in the wholesale market also supported by differential effect in supermarketsand street markets ― Supermarkets typically buy all their grocery products from the wholesale market. ― Street vendors, on other hand, have access to a variety of small producers, or are producers themselves (see Table V, col 5) 4. The Impact of the Reform on Specific Products at Street Markets ― Street vendors rely on wholesalers for specific products, e.g. almost never buy lettuce from wholesalers, but rely on them for peaches • Column 1, reports results of benchmark model, but including only lettuce (‘low’) and peaches (‘high’) in treatment group • Column 2 confirms differential effect in supermarkets and street markets • Column 3 shows at street markets, deregulation had no impact on price