WPS8294 Policy Research Working Paper 8294 Micro-Level Analysis of Mexican Retail Markets and Their Response to Changes in Market Structure and Competition Policies Luis Fernando Sanchez-Bayardo Alvaro Gonzalez Leonardo Iacovone Jobs Cross Cutting Solution Area January 2018 Policy Research Working Paper 8294 Abstract This paper develops the following price indicators to mea- (increased efficiency) have the expected correlation with sure the relative efficiency (functioning) of markets: (a) measures affecting the functioning of markets. It considered price dispersion, (b) price volatility, and (c) price trans- changes in competition and entry of large retail stores in the mission (speed, completeness, and symmetry). The paper local retail market. These changes affect market efficiency in uses these indicators to study trends and conditions of the way theory would predict. The results suggest that these the outlet level in retail prices for common commodities indicators are good measures of the relative efficiency (func- sold throughout Mexico. The analysis examines price pat- tioning) of markets. The findings also suggest that efforts terns for each indicator across commodities, regions, and to monitor markets using these indicators may be useful. time. The descriptive results indicate that although there For example, for policy makers who are concerned about is (expected) heterogeneity in the behavior of these indica- the distributional effects of liberalizing trade, the indicators tors across commodities, location variables explain the most may predict where price impacts will be felt the most and variation in the indicators. There are clear and persistent by whom. In addition, the indicators provide preliminary regional- and commodity-specific effects. Thus, the study information about relative competition levels, which may concludes that Mexico is not one, well-integrated national be helpful in saving the time and effort of the competi- market. The study tested whether changes in these indicators tion authorities and possibly making them more effective. This paper is a product of the Jobs Cross Cutting Solution Area. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at agonzalez4@ worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Micro-Level Analysis of Mexican Retail Markets and Their Response to Changes in Market Structure and Competition Policies Luis Fernando Sanchez-Bayardo, Alvaro Gonzalez and Leonardo Iacovone Keywords: prices, competition, entry, integration JEL Classification Codes: L10, L11, L81, F15, F61 TABLE OF CONTENTS Introduction .......................................................................................................................... 1 Literature review ................................................................................................................. 3 Data....................................................................................................................................... 6 Empirical strategy .............................................................................................................. 13 Descriptive statistics .......................................................................................................... 18 Regression results .............................................................................................................. 27 Conclusions ......................................................................................................................... 34 Appendix ............................................................................................................................. 36 References........................................................................................................................... 37 INTRODUCTION The choice to transact in a market—as a buyer, seller or intermediary—depends on price signals. Supposedly, when markets function well they send the ‘right’ price signals. Price levels, movements and trends represent the relative levels, changes and trends in the surplus and scarcity of goods and changes in the inherent value of these goods to both seller and buyer. Under these conditions, when parties transact, exchange benefits all. But markets may not function well and send the ‘right’ prices. Some markets may exhibit persistent and relatively higher price volatility, may be less well integrated to international and national markets and indicate relatively attenuated capacity to arbitrage across time and space. Most importantly, the relative poor performance of some markets may not be random nor transitory. In the price data from Mexico examined for this study, there are distinct spatial and temporal patterns that one would expect market forces to address. Ceteris paribus, markets function to attenuate heterogeneity in price levels, changes and trends. There was a lot of persistent heterogeneity in these price data. The motivation for this research was to identify and understand what prices, when and in which markets do not behave as theory would predict. To identify and comprehend the function of markets and price trends and conditions, this research focused on retail markets for perishable and non-perishable consumables of relatively homogeneous quality across Mexico. This study looks at three dimensions of market efficiency; namely, price dispersion, pass- through of costs to prices and price volatility. With respect to price dispersion, price variations across physical space are expected not to be sizeable or persistent when markets work efficiently and market agents face stiff price competition. Regarding pass-through, in efficient markets a change in costs of inputs will be fully reflected (passed through completely) in the prices of commodities that use those inputs with minimal delay. In addition, when input costs go up, prices will go up as quickly and as completely as when input costs go down —this is a symmetry condition of pass-through of competitive and efficient markets. Out of three proposed dimensions of market efficiency, volatility may be the least obvious dimension. However, volatility is indirectly related to efficiency. A benchmark on how much volatility is expected in an efficient market can be inferred. In well-functioning markets, volatility would solely be a function of the fundamentals of demand and supply. Because higher volatility introduces uncertainty, it increases the chances of buyer and seller error and it is something that both seller and buyer may be interested in reducing.1 In addition, higher volatility may introduce temporally missing markets, especially when information is perfect—buyers delay purchases when they know prices will go down and sellers withdraw 1 The error is based on buying when prices may go down or selling when prices may go up in the next period. Some may object to categorizing this as an error. Admittedly, this sounds a lot like timing the market and this can only be done, without error, in hindsight (ex post). However, the error is not that unless a buyer or sellers can buy or sell at a better price, they make an error. The point is rather that in more volatile markets, using an average price as a representation of the likely market price, is a poorer heuristic at any period in time than in a less volatile market. When the average is a poorer heuristic, this means that the buyer or seller may pay too high or sell too low, respectively, than would otherwise be the case. This is what is meant here as an error. 1 goods when they know prices will go up.2 In sum, higher price volatility indicates that a market will have to assume costs to overcome it. For that reason, higher volatility is assumed to be less efficient than lower volatility. As already explained, the functioning of markets in this study refers to the comparison of the behavior of prices based on the following three dimensions: dispersion, volatility and pass- through. The study follows on and is informed by a pilot study in Nigeria. The results of that pilot study showed the following:  Substantial heterogeneity in the dispersion, volatility and pass-through of markets across time and location for well-specified, homogenous commodities;3 and  This heterogeneity was robust across time and displayed distinct geographical and commodity-linked patterns.4 The results of the Nigeria study were preliminary and not published because the price data were of substantially poorer quality than price data available from Mexico. That said, we find similar results for Mexico. In addition to identifying and understanding how well markets function across Mexico, we use these same indicators of market efficiency to determine how markets respond to changes in market structure and competition policies. Following Iacovone et al. (2011), this study expands on identifying the economic effects of a large retailer’s entry into Mexico. More specifically, instead of looking at the entry’s effect on manufacturing, this study examines the effect of entry on the performance of markets in locations near and far from the physical location of new stores. For the US, Fishman (2006) provides insight on the effects of Wal- Mart on local markets. Driven mostly by constant reductions in transportation and production costs, Wal-Mart usually offers goods at lower prices than their competitors, in turn forcing the rest of retailers to drop their prices. Atkin et al. (2015) find similar results in prices for Mexico. Indeed, this study finds that a large retailer’s entry improves the functioning of markets as we have defined them. All things being equal, markets more immediately and directly affected from the entry of this large retailer function more efficiently after entry than those less affected. Also, this research exploits the existing information on competition in Mexico to analyze the effects of anticompetitive practices on the functioning of markets. Using the documents published by the Mexican competition agency, COFECE, on investigations, opinions and resolutions regarding the existence of anticompetitive activities in industries and regions across time, a database is assembled to identify markets with distortions in competition and use these data as a control when evaluating the functioning of markets. It is shown that markets where anticompetitive practices are present behave less efficiently when no regulator acts to correct such inefficiencies. The next section contains a literature review on the indicators of market efficiency used. The empirical strategy proposed to estimate the dimensions of efficiency of markets is discussed 2The assumption is that the price falls below cost since price was set competitively—at zero economic profit. 3 White and yellow maize, local and imported rice, sorghum, cement and bottled water. 4 After controlling for physical isolation, proximity to international borders, economic density, competition, proximity to production of the commodity in question and broad measures of poverty. 2 as well as the definition of market efficiency. Following this, the results are provided. The final section presents conclusions and policy recommendations. LITERATURE REVIEW This review of the literature on price dispersion, price volatility and price transmission is not meant to include the vast existing literature. Price dispersion Price dispersion refers to the existence of more than one price for the exact same commodity at the exact same time, across different, nearby locations (Eckard, 2004). The existence of price dispersion implies that markets may not be completely efficient. In markets with high levels of competition, when prices differ across locations, unexploited arbitrage opportunities exist. Arbitrage would force long-run price differences to cover transaction costs, and the ‘Law of One Price’ (LOP) would generally hold. However, studies indicate that there are persistent deviations from LOP across cities within the same country. For example, even for narrowly- defined commodities, Ardeni (1989) finds that LOP does not hold. Trade costs are detected in domestic markets as well. Trade costs are the most common explanation for violations to the Law of One Price (Crucini et al., 2005; Goyal, 2010; Lee, 2010; Blomberg and Engel, 2012). These costs include handling, marketing, breakage, stealing, spoilage, etc. (Crucini et al., 2005; Lee, 2010). Portugal-Pérez and Wilson (2008) and Atkin and Donaldson (2014) find that other market frictions limit the transmission of price signals and prohibit market arbitrage in African economies and lead to variation in prices for the same commodity. In addition to trade, arbitrage and other transaction costs, Stigler (1961) argued that information is costly for buyers and this leads to price dispersion as well. He provided the first model of buyer’s imperfect information and price dispersion that others developed further (Salop and Stiglitz, 1977; Pratt, et al., 1979; Salop, 1979; Varian, 1980; Burdett and Judd, 1983; Stahl, 1989). These models show that many prices can exist and persist in equilibrium because it may be too costly for buyers to identify exactly which sellers, in a market with many sellers, are offering the lowest price. Some sellers benefit from this limited ignorance and can charge a higher price than others. Theory on the importance of information for price convergence is supported by empirical evidence. Parker et al. (2013) look at grain price information in India acquired via mass text messages and find that price dispersion during an unanticipated short-term ban on mass text messages increased. Goyal (2010) estimates the impact of soy bean farmers having access to wholesale price information (via Internet kiosks and postings at warehouses) in India and finds that average prices rise and price dispersion decreases. A long-standing empirical literature documents considerable commodity price variability across space, especially in developing countries, with various empirical tests of market integration suggesting significant and puzzling forgone arbitrage opportunities (Fackler and Goodwin, 2001). Using variations in geographic coverage and rollout of mobile phone services, two studies found that price dispersion reduces dramatically for staple commodities, with bigger effects for fish in India (Jensen, 2007) than for grains in Niger (Aker, 2010). 3 Institutions seem to matter as well in the existence of price dispersions. A study by Gluschenko (2011) examines convergence to one price across the Russian Federation’s heterogeneous regions. He finds that markets with better institutions are more integrated, converging to a narrower band of prices for the same commodity, than regions where reforms have yet to take hold and market function is hampered by government interference, regulations and corruption. Blomberg and Engel (2012) find that price dispersion across Iraqi cities decreased during ‘the surge’, an increase in military presence, for a wide range of products. McCann (2012) shows that incidence of violence in Somalia augmented the bilateral price dispersion between Somali cities. But this might simply reflect increased transportation or trade costs as a reduction in military presence has an impact on illegal bribes (Olken and Barron 2009). Cross-sectional variation in market power can also lead to different prices in what may seem to be the same market. Markets might, for example, not be large enough to support enough trade for perfect competition and exhibit a monopoly or oligopoly structure (Abbott, 1994; Bresnahan and Reiss, 1991; Shepard, 1991; and Stewart and Davis, 2005). But even in large markets, there may be enough market power through segmentation that some price dispersion may exist. For example, Çağlayan et al. (2012) explain the persistent deviation across shops in Istanbul by the presence of monopoly power for shops located within residential areas, where customers shop when they arrive home from work. Price volatility From the perspective of the consumer, higher price volatility is associated to uncertainty and risk. The theoretical frameworks of Pratt (1964), Turnovsky et al. (1980) and Schmitz et al. (1980) reveal why price stability is preferred. Turnovsky et al. showed how a consumer’s preference for price stability depended on only a handful of parameters, and the authors then derived a similar measure for multiple commodities. Additionally, the empirical framework developed by Finkelshtain and Chalfant (1991, 1997) confirms predictions of the preference for price stability. On the supply side, under certain circumstances—even under high levels of competition—sellers and producers may benefit from price volatility (Oi, 1961). With respect to the issue of welfare and price volatility, in development economics questions are centered on how commodity price volatility affects the welfare of households in contexts where production and consumption smoothing is often difficult. Output price uncertainty generally causes firms to employ fewer inputs, forgoing expected profits in order to hedge against price volatility (Baron, 1970; Sandmo, 1971; Barrett, 1996 for farms). When the analysis of commodity price risk is extended to individual consumers (Deschamps, 1973; Hanoch, 1977; Turnovsky et al., 1980; Newbery and Stiglitz, 1981), these are generally thought to be price risk-loving for a specific commodity when the budget share of that commodity is not too large. In sum, there is no specific link between price volatility and market efficiency as defined and used in this study. However, given that volatility might either increase transaction costs— since risk has to be mitigated or lead to missing marketsless volatility will lead to more efficient market transactions than more volatility. For this reason, in this study, above average levels of price volatility are an anomaly that requires study and explanation. 4 Pass-through Price pass-through is the most common measure of how transmission between markets happens. Following Fackler and Goodwin (2001), pass-through happens when cost or price shocks arising in one location are transmitted to the other. Specifically, the market for good in region is said to be spatially integrated with that of region if a shock that shifts, for example, demand in but not in affects the price in both and . Even if large price differences exist between a pair of locations, if shocks are transmitted, then markets are considered to be integrated and efficient. The magnitude as well as the speed of the pass-through indicates how well markets transmit information about changes in costs or prices. For this reason, pass-through is directly related to the efficiency and integration of markets. In addition, the nature of pass-through may also indicate characteristics about the market structure (competitive or otherwise). The empirical literature finds that pass-through is often incomplete (Campa and Goldberg, 2005; Goldberg and Hellerstein, 2008). For example, a one dollar rise in the per unit cost of an input is not always transmitted as a dollar increase in retail prices for that unit. Even when accounting for lags several studies confirm this (Campa and Goldberg, 2006; Nakamura, 2008; Nakamura and Steinsson, 2008; Hellerstein, 2008; Goldberg and Campa, 2010; Nakamura and Zerom, 2010; Eichenbaum et al. 2011). Incomplete price pass-through is attributed to remoteness of markets, trade barriers, or government interventions. The magnitude of pass-through can also vary with market structures (for example, see Jackson, 1997). The level of competition in a market matters for the intensity of transmission of cost changes into final prices. Pass-through is complete (100 percent) in environments with perfect competition.5 For all other market structures, the expected pass-through is a bit ambiguous and depends on other conditions. With sufficiently high transport costs, the pass- through could even be zero with perfect substitutes.6 Under monopolistic competition, the pass-through decreases with the transportation costs (Engel and Rogers, 1996). Incomplete pass-through can therefore be the result of anti-competitive behavior (Domberger, 1979). In addition to its magnitude, pass-through from costs to prices may have varying delays or ‘speed of adjustment’ across markets (Dixon, 1983). The existence of pass-through delays is explained by the fact that business may incur costs (menu costs) to change prices (Rotemberg, 1983; Nakamura and Zerom, 2010; Klenow and Kryvtsov, 2008). Eichenbaum et al. (2011) argue that retail prices generally do not change in the absence of a change in costs but conversely changes in costs may not be sufficient to trigger a change in retail prices. Nakamura and Steinsson (2008) and Kehoe and Midrigan (2007) find that when sales (discounts) are excluded, prices are rather inertial: prices change on average every 11 months 5 In an environment with constant marginal costs and linear demand, pass-through for a monopolist is predicted to be 50 percent. In general, pass-through for the monopolist is defined by , while pass- through for a competitive markets is defined by: . 6 1 where 0 is the transportation cost. This equation illustrates the arbitrage between two points A and B when the good sold in A is a perfect substitute to the good sold in B. If the price in A is within the band expressed in the inequality, there is no exchange. On the edges of the band, pass-through is complete. 5 for the Nakamura and Steinsson study and every 4.5 months for the study carried out by Kehoe and Midrigan. Prices in some markets may be slow to change compared to costs because the physical good may only be a relatively small part of the unit costs. When this is the case, the buyer also pays for additional services (marketing or/and advertisement) and local costs (wage, labor, local taxes, local transportation services, retail services). These additional costs may be substantial compared to the unit cost of the physical good (Burstein et al. 2003). This implies that even a significant increase in the unit cost of an imported factor of production or in wholesale prices may increase retail prices by only a small fraction of total costs (Burstein et al., 2003; Burstein et al., 2007; Goldberg and Campa, 2010; and Nakamura and Zerom, 2010). For example, Goldberg and Hellerstein (2008) indicate that response of prices to exchange rates in the beer industry is small mainly because of local costs that are responsible for approximately half of the observed inertia of retail prices. So even though changes in exchange rates should affect the price of beer bought and sold internationally, changes in retail prices do not reflect changes in exchange rates. Behavior of consumer prices in Mexico A few studies for Mexico analyze prices from a microdata perspective. One particular finding in some of these studies is the asymmetry between the frequency of upward and downward price changes, with upward movements being more common, suggesting some inefficiencies in price adjustments (Gagnon, 2009 and Ysusi, 2010). Vaughan (2013) finds some evidence of market integration in the form spatial correlation of cities’ price indexes for the same good, but this only holds in the presence of structural breaks (i.e., a change of indexes from non- stationary to stationary processes); once a structural change takes place, cities’ prices seem to be driven by internal (idiosyncratic) factors rather than spatially or regionally. Finally, Gagnon (2006) identifies that in periods of low inflation (below 10-15%), such as during our sample period,7 the adjustment in inflation is mostly driven by the average magnitude of price changes rather than by the frequency of changes. The frequency of changes plays almost no role in such adjustment. Effects of the entry of a large retailer on prices There is documented evidence on the effect the entry of a large retailer has on retail prices. Basker (2005) and Fishman (2006) indicate that Wal-Mart charges lower prices in the markets it enters, inducing other retailers to reduce their prices too. Thus, a possible consequence of higher competition is a lower price dispersion. Fishman (2006) also hints at the possibility of a reduction in price volatility when Wal-Mart enters a market, because it smooths fluctuations in the seasonality of some goods’ prices. Relatedly, Atkin et al. (2015) find that the arrival of global retail chains has led to welfare gains in average Mexican households, mainly driven by reductions in the costs of living. DATA To identify and describe how prices for certain well-specified commodities function across time and space for markets across Mexico, we look at price dispersion, price pass-through and price volatility temporally and geographically. To explain how these dimensions were 7 The highest annual CPI inflation level in our sample is 4.97%. 6 formulated and estimated, we first provide the reader with a description of the data use and then the empirical strategy. The INEGI INPC microdata The data are a time series of urban retail prices collected by the Mexican Institute of Statistics, INEGI (Spanish acronym), originally collected bi-weekly to estimate consumer price indexes. Geographic data are mapped to the price data to have information on the location where the data were collected and to exploit heterogeneity in location characteristics in the analysis. There is also information on the type of outlets from which these retail prices were collected and this information is used in the analysis as well. In addition to understanding the patterns of how markets work, this novel retail price data set is also used to identify and understand how entry of large retailer stores affected the performance of the retail markets these new stores serve compared to those they do not. In addition, using the same micro-level price data, this research identifies and explains the effect of competition policy changes. The database used to study the functioning of markets consists of time series of retail prices collected by INEGI used for the computation of Mexico’s consumer price index (INPC). The INPC is published every two weeks and takes price quotes from 46 cities. The prices of cement are supplemented with INEGI’s database on residential construction prices, which has the same characteristics as the INPC data, but with fewer observations within cities. The period covered in this study goes from January 2010 to December 2015. Table 1. Cities and Regions in the INEGI INPC database Region City Abbrev. State Region City Abbrev. State Mexicali MXL Acapulco ACA Baja California Guerrero Tijuana TIJ Iguala IGU Northern La Paz LAP Baja California Sur Tulancingo TUL Hidalgo border Acuña ACU Coahuila Toluca TOL State of Mexico Juarez JUA Chihuahua Center‐ Cuernavaca CUE Morelos Matamoros MAT Tamaulipas south Puebla PUE Puebla Tepic TPC Nayarit Tlaxcala TLAX Tlaxcala Culiacan CUL Sinaloa Veracruz VER Northwest Hermosillo HMO Cordoba COR Veracruz Sonora Huatabampo HUA San Andres Tuxtla SAT Torreon TOR Campeche CAMP Campeche Coahuila Monclova MVA Tapachula TAP Chiapas Chihuahua CHIH Oaxaca OAX Chihuahua Oaxaca Jimenez JIM South Tehuantepec TEU Norteast Durango DGO Durango Chetumal CHE Quintana Roo Monterrey MTY Nuevo Leon Villahermosa VSA Tabasco Tampico TAM Tamaulipas Merida MID Yucatan Fresnillo FRE Zacatecas Mexico Mexico City Federal District and MX Aguascalientes AGS Aguascalientes City Metro Area State of Mexico Colima COL Colima Leon LEON Guanajuato Cortazar CTZ Center‐ Guadalajara GDL Jalisco north Tepatitlan TEP Morelia MOR Michoacan Jacona JAC Queretaro QRO Queretaro San Luis Potosi SLP San Luis Potosi All Mexican states and the Federal District (which roughly corresponds to the Mexico City Metro Area) are represented in the database with at least one city. INEGI groups these cities 7 in 7 regions (Table 1 (above) and Figure 1 (below)). As described by INEGI,8 the sample contains cities with a population of at least 20,000 inhabitants. The sample also includes the 10 most populated metropolitan areas. Cities are catalogued as being “small” (20,000 to 120,000 inhabitants), “medium” (over 120,000 to 600,000 inhabitants) and “large” (over 600,000 inhabitants). Given the nature of this database, the results of this study can only be attributed to urban markets. Prices are collected at seven types of outlets of different sizes and characteristics. These are classified as: supermarkets, department stores, specialized stores, public markets, convenience stores, stands and informal markets, and warehouse clubs. Additionally, cement prices are collected from retail construction materials stores. Figure 1. Cities and Regions in the INEGI INPC database INEGI’s complete INPC database includes both goods and services and follows the prices of 283 generic categories. That is, goods or services classified by their broad description. Within each generic, INEGI tracks several specifics, for which more detailed information exists, such as type/variety, brand, presentation/packaging, size, etc. Examples of generics are “tomato”, “detergent”, “milk”, “beef”, etc. Similarly, some specifics can be “tomato: roma type – unbranded – unpackaged – by the bulk”, “detergent: liquid – Tide brand – 1lt bottle”, “milk: 8http://www.inegi.org.mx/est/contenidos/Proyectos/INP/Default.aspx?_file=documento_metodologico_inpc.pdf 8 whole – Alpura brand – 1lt carton”, “beef: ground –unbranded – unpackaged – by the bulk”, among many others. INEGI follows a wide range of specifics within a certain generic category. The aim of casting a wide net is to gather as much information from markets as possible and to avoid capturing price behaviors that may be idiosyncratic to certain brands or varieties. In some cases, this wide net approach is also due to the fact that consumption habits and tastes vary regionally and the aim is to capture all habits across Mexico. A clear example of this is bean consumption: people from southern and central states tend to prefer (and consume more) black beans, whereas people from the north and the west usually consume pinto beans (and almost no black beans at all). Accordingly, INEGI’s price gatherings of the generic category “beans” in the south and the center of the country contain more information of the specific good “black beans” and much fewer price quotes of the specific “pinto beans”. The opposite is the case for cities in northern and western states. In sum, INEGI’s database contains detailed information of the characteristics of wide range of goods. We were able to track specifics precisely and to distinguish them from other goods that may be similar. As a result, it was possible to analyze goods that are exactly the same across cities and time; we compared apples with apples. Also, INEGI’s wide coverage of generics means that price data for these may be heterogeneous within some generic categories. In principle, higher levels of heterogeneity should not pose a problem for our analysis. As long as a specific is widely covered across and within cities and is consistently followed through time, heterogeneity in the price data is, if anything, a potential advantage to exploit in the empirical strategy. Challenges when using the INPC data While INEGI does follow all generics in all cities, there were cases where specifics surveyed in certain cities or regions did not match (strictly speaking) with the ones followed in other geographical areas. This seemed to be apparent independently of consumption habits and tastes or availability of certain specifics within a city. This sometimes happened because INEGI’s database is intended to be representative only at the national level, a requirement that allows degrees of freedom on goods selection across cities. An illustrative example for this is soda. Soda is widely available basically everywhere and it comes in different presentations, sizes, flavors and brands. There is no doubt that many brands and sizes (for example, Coca Cola – 600 ml bottle) are sold in all cities and in practically all types of outlets. Thus, it would be expected to count with several price quotes in all cities of such common specifics. However, what we see is that INEGI collected price data on certain varieties and brands in some cities and other varieties and brands in other cities. The price of specific “Coca Cola – 600ml” is widely tracked in a set of cities, while in others the specific “Pepsi – 600ml” is tracked. We know that both specifics are widely available and sold in all cities, but the data collection does not reflect this. This fact limited the number of exact matches of specific goods that could be compared across cities and represented a challenge to the empirical strategy. For some goods, we were able to obviate brand and analyze specifics of exactly same characteristics (variety, size, package, etc.) as if they were the same good. Unfortunately, we also had to drop entire generic categories simply because the set of generics was too heterogeneous and prices across these 9 difficult to compare. Among some of the dropped generics are detergents, dishwashing liquid, body soap and toilet paper. Table 2. Description of goods included in the analysis Unit of Group Good Variety Size Brand measure Beans Pinto and black 1 kg Most common brands per kg Grains Rice Super extra 1 kg Most common brands per kg Bolillo bread ‐ per piece Unbranded per piece Tortilla & Concha bread ‐ per piece Unbranded per piece bread Corn tortilla ‐ by the bulk Unbranded per kg Eggs Brown and white per piece Most common brands per piece Eggs Brown and white 12 pack Most common brands per piece Eggs & milk Eggs Brown and white 18 pack Most common brands per piece Eggs Brown and white 30 pack Most common brands per piece Milk Whole 1 lt Most common brands per lt Water ‐ 1 lt Most common brands per lt Water ‐ 1.5 lt Most common brands per lt Beverages Soda ‐ 2 lt Most common brands per lt Soda ‐ 600 ml Most common brands per lt Beef Chop by the bulk Unbranded per kg Beef Ground by the bulk Unbranded per kg Bee Rib by the bulk Unbranded per kg Beef Steak by the bulk Unbranded per kg Meat Chicken Roasted by the bulk Unbranded per kg Chicken Whole by the bulk Unbranded per kg Chicken Breast by the bulk Unbranded per kg Pork Rib by the bulk Unbranded per kg Pork Steak by the bulk Unbranded per kg Apple Golden by the bulk Unbranded per kg Banana Cavendish by the bulk Unbranded per kg Guava ‐ by the bulk Unbranded per kg Lime Key (Mexican) by the bulk Unbranded per kg Muskmelon ‐ by the bulk Unbranded per kg Fruits Orange Valencia by the bulk Unbranded per kg Papaya Maradol by the bulk Unbranded per kg Pear D'Anjou by the bulk Unbranded per kg Pineapple ‐ by the bulk Unbranded per kg Plantain ‐ by the bulk Unbranded per kg Watermelon ‐ by the bulk Unbranded per kg Avocado Hass by the bulk Unbranded per kg Cabbage ‐ by the bulk Unbranded per kg Zucchini ‐ by the bulk Unbranded per kg Carrot ‐ by the bulk Unbranded per kg Chayote ‐ by the bulk Unbranded per kg Jalapeño pepper ‐ by the bulk Unbranded per kg Poblano pepper ‐ by the bulk Unbranded per kg Vegetables Serrano pepper ‐ by the bulk Unbranded per kg Cucumber ‐ by the bulk Unbranded per kg Lettuce Romaine by the bulk Unbranded per piece Onion Common by the bulk Unbranded per kg Potato White by the bulk Unbranded por kg Tomatillo ‐ by the bulk Unbranded por kg Tomato Beefsteak by the bulk Unbranded por kg Tomato Roma by the bulk Unbranded por kg Salt Table salt 1 kg Most common brands por kg Sugar Refined 1 kg Most common brands por kg Other Sugar Refined 2 kg Most common brands por kg Cooking oil Mixed 1 lt Most common brands per lt Cement Portland 50 kg Most common brands per 50 kg Sample selection strategy We focused on analyzing consumable commodities—perishables and non-perishables. Moreover, we chose commodities that are bought and sold in relatively thick markets on both the demand and supply sides, have relatively low information costs to determine differences 10 in quality and can be relatively easily arbitraged across time and location (stored). The idea is to analyze prices of goods whose characteristics make the possibility of the existence of efficient markets a more plausible one. The set of goods selected for the analysis intends to comply with these requirements (see Table 2, above). A total of 54 specific goods were selected and are studied from January 2010 to December 2015, meaning that there are 144 periods of data gatherings. The database has over 1,770,000 observations and comprises prices from 3,386 retail stores located in each of the 46 cities. The selected specifics are homogenous: varieties. There few, negligible; differences in quality or other characteristics (such as brand) that are low and relatively easy to observe in the data; and presentation/packaging is fairly standard. A good number of these are perishable goods—fruits, vegetables, meat, and some varieties of bread and corn tortilla—with no brand, or packaging and sold in bulk. The non-perishables such as bottled water, soda, sugar, salt and cement are classified according to size, variety and package, while brand is removed. We clarify that, in order for two specifics of different brands to be considered the same, the brands had to be “similar” in terms of market coverage (i.e. they had to be quite common nationally, regionally or within a city) and in price levels. INEGI has deep coverage of the goods selected at the city level. In most cases, there are several price quotes within a city at a certain time period. This is an important requirement for the data when calculating market efficiency indicators, especially dispersion. More importantly, the sample of goods is highly representative in the basket of consumption of Mexican households belonging to the bottom 40 percent of the income distribution. They represent 29 percent of consumption expenditure of such households. This selection criterion left us be confident that the goods we follow share two important properties: 1) The markets of these goods are, with all probability, thick, given their high level of expenditure; and 2) These goods are indeed “meaningful” to households, especially to the poorest ones who are the most likely to be affected by changes in prices. The price data required cleaning. In some cases, this was due to missing prices for certain specifics, for which sometimes INEGI attributed either the price of a “similar” specific within the same store, the price of the missing specific from a nearby store, or an average price within a city or region (what INEGI calls precios imputados—imputed prices). Other cases related to misspellings and errors in data collection such as the misplacing of decimal points, prices not standardized and other (undecipherable) errors. When errors were clearly identifiable, they were corrected. Price quotes were dropped when these were precios imputados or the price levels were considered as outliers. We defined outliers as prices that suddenly changed and were below/above three standard deviations according to a 3-month moving window or those whose level was suddenly below/above three standard deviations of its price distribution nationally. International prices In addition to INEGI INPC data, other sources were used to gather data required to both calculate indicators and to use as controls for regressions. International prices of goods were required to estimate pass-through. International prices were not available for all commodities, especially for non-perishables, such as cement, soda or water that can be 11 considered non-tradeables. We relied on the database sources used to construct the monthly commodity prices of the World Bank’s Pink Sheet and the International Monetary Fund (IMF) Commodity Prices databases. Given that our database contains bimonthly prices, the sources mentioned above were used only as reference in finding more frequent price collection. For goods that did not appear on any of the World Bank and IMF databases, we collected international prices from the US Department of Agriculture (USDA), Market News. We used shipping point prices for the most common varieties and presentation in which goods are sold. These shipping point prices represent first on board (FOB) prices of open market (spot) sales by first handlers at point of production or port of entry on products of generally good quality and condition. The USDA data may not always represent a world price but the specificity and high frequency of the price data collection makes it a valuable source. Also, this database contains at least an international price of a country that happens to be Mexico’s most important trade partner. Table 3 contains a description of the data gathered. Table 3. International prices database description Intl' price Good Description Country of origin Unit of measure Source available? Beans  US average bid price US 100 lb bag US Department of Agriculture Rice  Bolillo bread Concha bread Corn tortilla*  US (Gulf), Maize (US No. 2, Yellow) US per tonne FAO Eggs  Shell Eggs: Weekly Combined Regional Milk Water Soda Beef  CIF US import price Australia per lb Meat and Livestock Australia Chicken  FOB broiler/fryer prices US per lb US Department of Agriculture Pork  Lean Hogs Futures (HEZ6) US Per tonne investing.com Apple  Shipping point average price US 36 & 40 lb carton US Department of Agriculture Banana  US East Coast ‐ Main Brands, f.o.r. Central America per tonne FAO, The World Bank Guava Lime  Shipping point average price Mexico 40 lb carton US Department of Agriculture Muskmelon  Shipping point average price Brazil, Central America, Mexico, US 30 & 40 lb carton US Department of Agriculture Orange  Shipping point average price US 38 lb carton US Department of Agriculture Papaya  Shipping point average price Brazil, Central America, Mexico 25, 40 lb carton; 3.5 kg container US Department of Agriculture Pear  Shipping point average price Argentina, Chile, US 45 lb carton; 18 kg container US Department of Agriculture Pineapple  Shipping point average price Central America, Mexico 20 lb carton US Department of Agriculture Plantain  US East Coast ‐ Main Brands, f.o.r. Central America per tonne FAO, The World Bank Watermelon  Shipping point average price Central America, Mexico, US Several packages, price per lb US Department of Agriculture Avocado  Shipping point average price Mexico, Chile, Peru 23 lb carton/flat US Department of Agriculture Cabbage  Shipping point average price US 50 lb carton US Department of Agriculture Carrot  Shipping point average price Mexico, US 25 & 50 lb sack US Department of Agriculture Chayote  Shipping point average price Central America, Mexico 20 & 40 lb carton US Department of Agriculture Cucumber  Shipping point average price Central America, Mexico, US 55 lb carton US Department of Agriculture Jalapeño pepper  Shipping point average price Mexico 28 lb carton US Department of Agriculture Lettuce  Shipping point average price US 40 & 50 lb carton US Department of Agriculture Onion  Shipping point average price US 25 & 50 lb sack US Department of Agriculture Poblano pepper  Shipping point average price Mexico 28 lb carton US Department of Agriculture Potato  Shipping point average price US 50 lb carton/sack; 2000lb sack US Department of Agriculture Serrano pepper  Shipping point average price Mexico 28 lb carton US Department of Agriculture Tomatillo  Shipping point average price Mexico 40 lb carton/container US Department of Agriculture Tomato  Shipping point average price Mexico, US 20 & 25 lb carton/flat US Department of Agriculture Zucchini  Shipping point average price Mexico, US 21 & 28 lb carton US Department of Agriculture Salt Sugar  US Sugar # 11 futures (SBH7) US per lb investing.com Cooking oil  Soybean oil futures (ZLZ6) US per lb investing.com Cement Market structure and competition data We tracked sanctions imposed by COFECE, Mexico’s competition agency to study prices in markets with possible competition issues. The voluminous set of documents on investigations, resolutions and opinions on competition matters was used to identify markets 12 with potential competition-related problems. COFECE’s research information covers analyses for different types of goods (or categories of goods), for several types of economic sectors (including retailing), in different regional levels. We found that during the period of analysis, COFECE fined industries in poultry, soda, tortilla, avocado and cement. The businesses fined by COFECE were large companies with presence at the national level, which allowed us to analyze the effect of the sanctions in all city markets. We also use data on the entry of a large retailer store in Mexican cities through time. This large retailer operates several types of retail stores aimed to sell to different types of customers. There are four store brands relevant to this study and we distinguished each for the analysis: 1) the first brand is basic and austere mini- and super-markets that mostly sell groceries and personal care products at the lowest price possible; 2) the second brand sells a wide range of goods, from groceries to electronics and small appliances; 3) the third brand is premium supermarkets focused on higher-end groceries; and 4) the fourth brand stores are a warehouse club. We are able to follow the openings of these four kinds of stores in all cities in our database. The large retailer provides information on store openings in their monthly sales reports, which are available in their webpage. Thus, we can estimate the effect of the entry of each of the store brands, either as the group as a whole or by type of store. The first brand stores are by far the most common of all this large retailer’s stores, and expanded aggressively during the time period (722 total openings in the sample, including both super- and mini- markets). Then, the second and fourth brand stores opened less frequently (63 and 40 openings, respectively), while third brand store openings were the least common (23 openings). EMPIRICAL STRATEGY The empirical strategy is divided into two activities. First, we estimate indicators to identify and understand how markets are working. This way we can distinguish those markets that seem to be working relatively well from those that work less efficiently. The second step consists in analyzing the effect that the described trade and market structure variables play in the functioning of such markets. It is important to understand what we mean by market efficiency through the use of the indicators discussed above, in the literature review. Defining efficient markets is the first task of the empirical strategy. Efficient market and its characteristics To identify and understand which markets are functioning well, we first define the performance criteria. Our criteria equate a well-functioning market with an efficient market. More precisely, since all of the indicators of efficiency used are relative, we define a market that is functioning relatively well as a market that is working relatively more efficiently than other markets. 13 Market efficiency refers to the situation in which prices reflect, fully and immediately, all information available, and where no opportunities of arbitrage exist.9 This definition sets a high standard. It is unlikely that many markets function with perfect efficiency. Many markets do not adjust prices at all times for all locations, for every buyer and seller in ways that fully and immediately reflect all relevant information that materializes. For a variety of reasons, inefficiency in markets is a fact of life. As a general rule, however, markets tend to adjust prices in response to new information. Some markets are better at doing this faster and more completely than others. The point is to understand why some markets can do this relatively well, while others seemingly do not. Since there is not a perfectly efficient market, this research examines gradations of efficiency by comparisons. The analysis focuses on questions such as: why arbitrage may be relatively more apparent in, say, Market 1 than in Market 2, why there may be relatively more complete and immediate cost pass-through in one market compared to another, why prices for good 1 are more volatile than for good 2, etc. This research is comparative in nature in the sense that the performance of markets is stacked against each other. For that reason, it is important to set benchmarks to which the proposed measures of market efficiency are compared. The canonical, hyper-efficient market is used as a benchmark in this study, despite the fact that perfectly efficient and fully competitive markets are, at best, rare. For the same reason that physicists use frictionless models as the starting point to describe important features of the physical world, economists use the functioning of a perfectly competitive and efficient market to provide insights into important features of how markets have the potential to work. When the functioning of the ideal and observed markets does not match, this mismatch affords the opportunity to explore why this is the case. Also, the variation in performance of markets among themselves—that is, heterogeneity in market efficiency—affords an opportunity for further research as well. Thus, comparisons between the ideal benchmarks and the data are used throughout this study. It is important to begin here by explaining the efficient market benchmark. The basic feature of an efficient market is that prices reflect all information available. If a market is efficient, price signals will be available to all market agents, in the blink of an eye and at zero cost. With every market participant having the same information, prices should behave in the same way. Stiff price competition is the engine that incentivizes all market agents to remain informed and to use all available information. The ability of buyers to switch, instantaneously, from higher priced sellers to lower priced ones, is based on two other features that characterize competitive markets: zero transaction costs and zero information costs. Zero transaction costs means that a buyer will assume no additional cost in switching from one seller to another. With no transportation costs, for example, there is no additional cost to buying from one seller as opposed to another. In addition to zero transaction costs, perfect information translates into instantaneous knowledge about all prices, and all dimensions that are required to complete the buying transaction with any seller. All sellers have to respond immediately to changes in prices or lose all sales or gain all sales but at prices that represent a loss for each unit sold. These 9Concerns about the efficiency of markets are most often relegated to discussion about financial markets. Indeed, the concept of market efficiency as described here was first introduced by Fama (1970) in the context of foreign exchange markets, but it can be easily extended to any market. 14 assumptions about instantaneous price changes and symmetry between upward and downward price movements will also be used to indicate if a market is working relatively efficiently and exhibits signs of the use of market power. In sum, the prices of efficient markets behave like one large, agglomerate market—very much like a school of sardines swimming in the ocean. Even with external shocks, prices in efficient markets would quickly converge back to equilibrium. The question for this study is to ascertain just how closely this characterization resembles the behavior of prices in actual markets. Definition of markets The definition of markets is the foundation to the rest of the empirical strategy; the market is the unit of analysis for all regressions, not descriptive statistics, however. Markets are defined as the combination of location, commodity and outlet type. This definition is far from perfect. A better definition would be one used by antitrust authorities. In antitrust economics, the relevant market is defined by the cross elasticity of demand or cross-price elasticity of demand. Elasticities measure the responsiveness of the demand for a good to a change in the price of another good.10 When consumers substitute one good in response to small but significant and non-transitory price increases with another product, antitrust economists conclude that these products compete against each other and are therefore in the same market. By using a location-commodity-outlet type triplet we likely define markets more broadly than the antitrust definition. This is most likely to affect the price dispersion indicator where dispersion may be the result of aggregating markets that do not belong together, and therefore have different prices. At the same time, the definition includes economically relevant factors that shape how products compete and prices are derived and respond. This makes our definition of markets congruent with the definition of ‘relevant markets’ from antitrust economics. Step 1: Estimation of Indicators In order to assess how markets are working, we estimate indicators to study markets in the following three dimensions:  Price dispersion;  Price volatility; and  Integration with international markets; o Pass-through speed o Pass-through magnitude o Pass-through symmetry As described above, each of the indicators plays an important role in understanding the functioning of markets. 10 It is measured as the percentage change in demand for the first good that occurs in response to a percentage change in price of the second good. 15 Price dispersion Price dispersion measures the variation of prices of stores within a market. It is calculated as the coefficient of variation using the prices of all stores of a certain market at a certain time period. The average price of the product in the defined market is used as the base price. This indicator is equal to 0 when prices are all the same across stores. The existence of high price dispersion can suggest the presence of barriers to arbitrage across stores. The coefficient of variation (CV) used as the measure for price dispersion is formally defined as: 1 ∑ (1) where is the price at outlet i at the two-week period t, and is the mean price. For the sake of simplicity, the descriptive statistics present price dispersion measures at a city level, that is, the variation of prices of all types of stores within a city. However, when analyzing the effect of market structure and trade policies, we use price dispersions at city and outlet level, in order to exploit the granularity of data. Price volatility Price volatility is a measure of the magnitude of price fluctuations, and is one of the fundamental features of most markets. A frequently used measure of volatility is the degree to which price deviates from its central tendency. The proposed indicator of price volatility is calculated as the standard deviation of price returns (i.e., price inflations). Equations (2) and (3) show the two steps required to calculate this indicator: ln / (2) (3) where is the price at period and is the price at period . Ideally, returns should be evaluated each period, or 1. When this is the case, returns are evaluated each two- week period. When a price series has a gap, the returns are normalized to be equivalent to two-week periods. Volatility is then calculated as the moving standard deviation of six months (twelve periods) of such returns. Similarly to the case of price dispersion, we use two definitions of volatility in this paper. The descriptive statistics present the average volatility within cities. The volatility indicators used in regressions with control variables are average volatilities at city and outlet level, in order to distinguish possible volatility patterns across store types. International market integration To assess the degree of integration of local markets to international markets, we estimate pass-through indicators for those goods where international prices are available. Pass- 16 through estimations give information about the magnitude and speed to which local prices respond to changes in international prices: Price pass-through magnitude refers to the long-term pass-through coefficient, i.e. the extent to which shocks in prices in international markets are transferred to domestic market locations in the long run.11 Price pass-through speed measures the time interval that it takes domestic prices to adjust to a shock in the long-run equilibrium between domestic and international prices. The cycle begins when price changes in the international market are reflected and ends when this change (of any magnitude) is reflected in retail prices in each location. The dynamics of transmission of shocks from international prices to domestic prices is estimated using the error correction framework of (Engle & Granger, 1987). This implies estimating a long-run relationship in the first place, between domestic and foreign prices as shown in Equation (4), where are world prices at time and are the residuals of the proposed regression. Under the hypothesis of integration of foreign and domestic markets, these residuals fluctuate around a stationary mean, their variance is also stationary, and represent, conceptually, temporary deviations from the long run equilibrium relationship between foreign and domestic prices. Once this equation is estimated, the residuals are then the focus of the empirical strategy to understand pass-through. (4) The second stage requires testing the dynamics of adjustment. Equation (5) is estimated by regressing ∆ , against ∆ , as well as the lagged errors from the long run relationship represented by , where ∆ , is the change in local prices from 1 to and ∆ , represents the change in world prices for the same time period. ∆ , ∆ , (5) , Indeed, this equation represents an error correction model (ECM), where estimates the short-term effect from world price changes to domestic price changes in the domestic market and captures the speed at which domestic prices adjust to a shock in the long-run equilibrium between domestic and international prices. Initially, the price adjustment is assumed to be symmetric with regard to positive and negative deviations. , , the lagged error form the long-term relationship, is the error correction variable. (6) ∆ , ∆ , Finally, to determine whether price transmission from international market to domestic market is asymmetric, the lagged residuals (i.e. the error correction term, ECT) from Equation (5) are separated into positive and negative components. This proposed estimation 11 This indicator is obtained from the coefficients that are statistically significant at a 10 percent confidence interval. 17 is shown in Equation (6). Positive components (errors) imply that domestic prices are temporarily above equilibrium, and that the processed adjustment implies a price decrease. Negative components (errors) imply the converse. This strategy was followed for all goods. We focus on series that appear to be cointegrated with the international prices or show to be stationary, otherwise any relation between local and international prices could be spurious. In order to evaluate this, we perform the Augmented Dickey-Fuller GLS test, which is better suited for short samples (in our case, at most series can be 144 observations long). In the descriptive statistics, we present pass- through indicators at city level, while in the regressions with the control variables we use the city-outlet level indicators. DESCRIPTIVE STATISTICS We describe and analyze how markets function in two ways. First, we use heat maps that characterize how price indicators behave for each good and city across the time series. Heat maps provide a visual idea of which goods and in which cities prices have a tendency toward higher efficiency. In the following section, the results of fixed effects regressions provide information on which characteristics play an important role in the way markets work. However, it is important first to understand the macroeconomic environment in which Mexico experienced during the period of study. Figure 2 indicates the evolution of prices, GDP, exchange rates and unemployment rates; low and falling CPI and exchange rates, a modestly falling GDP rate of growth and a very modestly falling unemployment rate. Figure 2. Macroeconomic indicators of Mexico during the period of study a) Inflation and Exchange rate b) GDP growth and Unemployment 6.0 18.0 7.0 GDP - annual growth rate Unemployment rate - % 17.0 6.0 5.0 16.0 5.0 4.0 15.0 4.0 3.0 14.0 3.0 13.0 2.0 2.0 12.0 1.0 CPI - annual growth rate 11.0 1.0 Exchange rate - MXN/USD (right) 0.0 10.0 0.0 Jan-2010 Jan-2011 Jan-2012 Jan-2013 Jan-2014 Jan-2015 Jul-2010 Jul-2011 Jul-2012 Jul-2013 Jul-2014 Jul-2015 Q1 2010 Q3 2010 Q1 2011 Q3 2011 Q1 2012 Q3 2012 Q1 2013 Q3 2013 Q1 2014 Q3 2014 Q1 2015 Q3 2015 Source: INEGI, Bank of Mexico. Source: INEGI. 18 Tables 4 to 9 provide visual information on the performance of each price indicator in the form of heat maps. In all cases except for the case of price levels, heat maps compare values of indicators across all goods and cities. A cell in green means that the value is in the lowest 10th percentile across the whole sample of indicators. A cell in red means that the value is in the 90th percentile while yellow is a value statistical median. In the case of price levels, colors compare prices by row, that is, by good, since it does not make sense to compare price levels across goods and cities but within goods across cities. Table 4 shows average price levels across cities for each commodity. Prices in the North tend to be higher, especially along the US-Mexico border and northwest regions. Cities in central and southern parts of the country tend to have lower prices. Noticeably, prices of certain goods are more or less consistently lower in the South vis-a-vis central states and vice versa. The South has, roughly speaking, lower prices for some, non-perishable (salt, sugar, cooking oil) goods as well as some meat products, but higher prices for vegetables and some fruits. Price dispersion patterns are clear by good, but a less distinct pattern across cities (Table 5). In general, dispersion seems consistent across goods more than location (or store type). Most fruits and vegetables indicate higher dispersion than non-perishables. Also prices for beans and bread indicate relatively higher price dispersion. On average, the standard deviation of prices is on average 0.15 the size of the average price within a city. Price volatility is largely influenced by the type of good. Regional factors, however, play a significant role (see Table 6) as well. Prices of fruits and vegetables are by far the most volatile, showing an average volatility of 16 percentage points (pp), while the average volatility of the rest of goods is four time less, around 4 pp. Recall that volatility is defined as the standard deviation within a 6-month period (12 observations), thus volatility of 16 pp implies that inflation rates of prices tend to move around 16 pp around its average inflation within such period. Geographical patterns are less obvious, but present. Roughly speaking, lower volatility is apparent in cities located in central states, including Mexico City. The extent to which shocks in international prices transmit to local prices is shown in Table 7. Pass-through measures shown in this table include coefficients of long-term relationship regressions whose significance is below 0.20 and whose series are either cointegrated or stationary. Values of zero in Table 7 refer either to correlated or stationary series whose long- term regression coefficients were equal to zero or statistically insignificant (p-value 0.20 or higher in this case). The magnitude of the pass-through in prices is particularly high for meat products, especially in the varieties of beef and chicken. In many cases, local prices overshoot when adjusting to shocks in international markets (shown in pass-through values above 100, implying that local prices absorb more than 100% the adjustment in international prices). Also, the magnitude of price pass-through is comparatively high for bananas and avocados. In the case of avocado, such pass-through magnitudes correspond with the fact that Mexico is a major exporter of this good. The rest of goods, on average, indicate a magnitude of price pass-through of around 36, which is a relatively low value. This implies that local prices in general tend to absorb 36% of the changes in international prices in the long-term. This is an indication that some Mexican markets, especially those for fruits and vegetables are relatively immune to the influence of changes in prices of international markets. This is not uncommon; perishable goods are more likely to be sold in local markets as compared to cement, soda or bottled water. 19 The speed of adjustment is the companion price transmission measure to the magnitude of pass-through. Table 8 contains results that indicate that the speed of adjustment in prices is perhaps more uniform across goods. In similar fashion to the magnitude of pass-through, Table 8 shows coefficients of Error Correction Model (ECM) regressions with significance below 0.20. Also, zero values are due to zero-valued coefficients or statistically not significant coefficients. On average, prices adjust at a rate of 22% each time period, implying that it takes approximately two and a half months (5 two-week periods) for local prices to fully incorporate shocks from international markets. Local prices of chicken are relatively more sensitive to changes in international prices in terms of both magnitude and speed to shocks. In comparison, many coefficients for meat products are not significant, suggesting that speeds of adjustment are much slower. Prices for many fruits and vegetables adjust relatively quickly, implying that, while they might not absorb shocks fully in terms of magnitude, these are absorbed faster than other goods, in some cases within a month. Price transmission asymmetries between positive and negative shocks are less frequent (see Table 9). In general, positive and negative shocks in international prices tend to appear at the same frequency in local markets. In some cases, prices for goods appear to have somewhat consistent positive asymmetries (i.e. local prices adjusting faster to positive shocks than to negative ones), such as papaya, carrot and chayote, and others with negative asymmetries like the case of watermelon and some meat products. As pointed out before, this asymmetry in price adjustments could indicate market power. However, to conclude that it is market power as the explanation for this finding would be premature. 20 Table 4. Average prices in cities Northern border Northwest Northeast Center north Center south South MX HMO CHIH LEON MOR MXL ACU MAT HUA MVA DGO MTY TAM QRO TLAX COR CAM OAX LAP JUA TPC CUL TOR JIM FRE AGS COL CTZ GDL TEP ACA IGU TUL TOL CUE PUE VER SAT TAP TEU CHE VSA MID JAC SLP TIJ Beans 24 24 23 21 18 23 24 25 18 24 20 20 19 18 20 22 19 21 21 21 24 23 26 23 21 22 20 18 24 22 24 20 23 23 21 22 19 21 19 21 20 18 20 20 18 24 Rice 18 16 15 12 13 12 14 19 12 12 16 16 13 14 16 13 14 11 13 14 16 13 15 15 12 12 15 14 13 13 15 17 15 15 16 12 12 14 16 12 12 13 17 15 17 16 Eggs ‐ per piece 2 2 2 2 2 2 2 2 2 2 2 2 1 2 2 2 2 2 2 2 1 1 2 2 1 2 2 2 1 2 2 2 2 2 2 Eggs ‐ 12 pack 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Eggs ‐ 18 pack 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Eggs ‐ 30 pack 2 2 2 1 1 2 2 1 1 2 2 2 2 1 2 2 2 2 2 2 2 2 1 2 2 2 1 1 2 1 1 1 1 1 2 2 Milk 15 14 15 13 13 12 13 14 14 13 12 12 13 14 13 13 12 12 11 11 11 13 12 12 12 11 14 12 13 13 14 14 13 14 14 13 14 14 15 13 13 13 13 13 13 14 Bolillo bread 3 3 3 2 2 1 2 2 3 2 3 1 3 2 2 2 2 2 3 3 2 2 3 3 2 2 2 2 2 2 1 1 1 2 2 1 1 1 1 1 1 1 1 1 1 Concha bread 5 5 5 5 5 5 4 6 5 4 5 4 5 4 5 5 5 4 4 4 4 5 4 4 4 4 4 4 5 3 4 5 5 4 3 6 3 4 5 4 4 4 5 4 5 5 Corn tortillas 14 12 13 11 11 12 12 12 14 12 11 13 13 12 10 12 12 10 11 12 11 10 11 10 11 11 11 10 12 11 9 11 12 10 10 11 10 10 12 10 10 11 13 12 12 10 Water ‐ 1lt bottle 8 8 7 11 10 9 7 8 7 9 6 8 8 6 8 8 6 9 9 8 9 7 8 5 7 7 8 6 7 6 8 6 8 6 7 6 6 7 6 7 6 11 5 10 6 8 Water ‐ 1.5lt bottle 5 7 6 5 5 6 7 6 7 5 4 5 5 6 5 6 6 6 7 5 6 5 5 5 5 5 5 7 6 5 8 6 5 6 5 5 6 5 5 6 5 6 Soda ‐ 2lt bottle 9 9 11 8 6 9 9 9 9 7 10 9 8 8 8 7 8 9 7 8 7 6 8 7 10 6 7 8 8 7 9 7 8 9 6 7 8 8 Soda ‐ 600ml bottle 15 15 16 15 15 15 14 15 15 15 14 15 15 13 14 14 14 15 14 13 12 14 13 14 13 15 13 12 13 13 14 14 14 13 13 16 14 15 14 14 14 14 13 15 13 14 Beef ‐ chop 110 111 86 82 86 112 94 107 61 86 64 78 103 72 106 89 128 86 94 101 88 103 75 104 82 93 92 87 105 111 109 97 109 119 83 100 91 79 95 109 117 85 86 90 103 Beef ‐ ground 81 83 93 86 73 87 82 102 92 81 70 81 78 84 71 67 80 76 85 89 81 86 78 77 82 79 91 80 81 78 85 93 89 85 80 81 71 71 78 78 88 84 75 86 74 97 Beef ‐ rib 62 68 69 89 64 78 69 68 86 60 68 65 78 68 62 59 70 119 95 79 75 91 72 73 88 63 82 70 72 101 84 88 104 111 92 66 66 57 62 51 62 52 62 60 91 Beef ‐ steak 99 107 100 84 98 84 96 102 105 109 92 89 103 104 90 90 101 102 98 96 98 99 96 92 98 98 99 101 97 106 90 113 105 102 93 103 102 91 105 97 107 93 95 97 95 103 Chicken ‐ whole 36 36 42 41 31 33 35 36 31 32 34 41 38 37 40 37 39 43 33 40 36 37 38 39 32 37 37 35 33 31 33 35 34 34 31 33 28 30 32 30 36 34 32 36 32 36 Chicken ‐ breast 62 85 91 70 68 60 68 55 59 54 51 52 59 56 59 55 61 57 59 66 59 55 63 62 57 52 56 62 59 68 66 63 47 57 58 62 57 46 61 60 63 Chicken ‐ roasted 89 85 97 75 85 98 84 108 83 85 100 79 93 90 81 80 98 94 92 72 78 89 86 87 70 76 76 89 90 78 79 72 88 68 82 91 81 68 69 65 67 63 98 104 71 68 Pork ‐ rib 80 75 83 71 62 69 68 75 70 88 82 67 70 74 70 73 79 77 72 60 72 89 74 75 74 74 82 83 77 84 78 79 71 78 81 84 78 67 72 80 80 83 80 75 70 76 Pork ‐ steak 67 76 68 72 64 67 64 65 69 72 60 71 64 70 69 67 74 73 68 65 70 81 72 76 69 72 76 68 69 69 72 70 67 72 75 75 74 69 67 66 79 70 61 71 68 75 Apple 26 27 27 22 25 36 31 33 28 24 31 25 27 33 30 25 31 29 31 29 28 27 31 30 32 29 31 30 24 32 29 28 33 32 28 35 31 30 30 34 31 35 30 30 Banana 11 11 14 9 10 10 10 11 11 11 9 10 10 9 11 10 9 9 8 9 9 10 11 8 10 10 11 10 11 11 10 10 8 10 9 10 10 8 9 6 10 9 9 8 10 11 Guava 24 23 23 23 20 24 18 25 25 25 19 21 19 17 18 23 18 19 15 18 16 17 18 16 16 21 17 22 17 15 15 18 14 15 18 16 16 18 19 20 16 21 16 24 20 18 Lime 16 16 16 14 14 15 14 15 14 14 14 13 14 14 12 15 14 13 12 13 12 13 12 12 11 12 13 12 13 13 13 13 13 12 12 12 12 13 11 12 13 12 9 16 12 14 Muskmelon 14 14 15 14 12 13 13 14 13 13 12 13 11 11 13 14 13 12 12 14 12 13 13 11 12 13 14 12 15 14 12 16 15 13 15 15 12 13 14 15 13 14 15 14 14 15 Orange 10 10 9 8 9 8 8 10 8 8 8 7 9 7 8 9 8 6 7 9 7 7 7 6 6 8 7 7 8 10 7 8 7 7 7 8 6 8 8 8 7 9 9 9 8 7 Papaya 20 20 20 19 18 18 15 18 18 19 17 19 18 17 17 17 17 17 15 14 15 16 15 13 14 16 16 14 13 14 13 15 14 14 14 13 11 13 14 14 13 12 11 14 15 15 Pear 29 28 32 34 31 42 32 31 27 27 31 32 29 30 33 34 36 34 32 33 30 32 28 33 31 32 35 31 33 31 30 33 33 33 31 36 32 33 35 33 33 31 34 34 36 31 Pineapple 16 14 13 13 13 12 11 13 13 13 12 12 11 11 11 12 11 13 10 13 11 14 13 11 12 12 13 12 12 14 13 12 14 12 13 12 12 12 11 12 11 14 12 13 13 13 Plantain 19 18 17 21 22 18 13 19 19 22 18 21 21 20 14 17 18 14 17 16 18 14 15 18 16 16 13 14 9 17 14 13 12 12 12 10 14 12 13 11 13 12 13 14 Watermelon 6 6 7 7 6 7 7 6 5 7 7 7 5 5 7 7 8 7 7 8 7 8 8 7 8 7 9 7 7 7 7 8 7 7 7 7 7 7 8 7 8 8 6 7 8 9 Avocado 36 35 36 30 30 33 35 36 31 29 32 34 31 28 32 34 32 31 30 35 31 33 32 33 31 30 33 27 35 36 32 36 35 33 31 34 30 34 36 34 32 39 37 35 30 35 Cabbage 9 9 7 5 7 6 7 7 7 6 5 4 6 4 6 5 7 6 8 7 6 6 6 7 8 6 6 6 6 6 6 7 6 6 13 6 7 6 6 7 6 6 5 7 7 Carrot 12 10 11 9 9 10 9 10 10 9 9 10 11 8 8 9 9 8 7 10 8 9 8 9 7 9 9 7 9 9 9 9 9 9 8 8 8 9 10 9 9 10 9 10 11 9 Chayote 14 13 14 15 16 16 13 14 14 14 15 14 16 14 14 15 12 15 15 15 10 12 11 10 9 12 12 12 11 13 9 11 10 10 10 8 8 7 13 10 11 11 11 10 11 11 Cucumber 13 12 12 13 12 12 10 11 11 12 12 13 13 11 13 14 12 12 12 11 11 12 11 9 10 12 13 12 12 10 11 12 12 12 10 10 10 11 12 12 10 12 11 14 12 12 Lettuce 10 11 12 7 9 9 9 10 9 10 9 9 10 6 8 8 7 8 10 7 7 8 11 8 7 8 7 8 9 8 8 8 8 8 9 6 8 10 9 8 9 12 9 9 10 Onion 12 12 15 13 13 13 13 15 12 12 12 13 12 10 13 12 13 11 11 14 13 14 13 11 12 14 15 12 15 14 13 13 12 13 12 13 11 13 15 13 12 13 13 14 16 13 Jalapeno pepper 19 17 15 12 12 15 16 16 16 19 14 12 15 11 11 13 14 12 15 19 14 18 16 27 18 15 15 14 17 15 14 16 16 16 13 15 14 15 18 15 16 15 19 19 18 Poblano pepper 24 24 20 24 21 23 20 21 23 23 19 23 25 18 16 21 24 19 17 22 19 22 20 18 19 20 21 18 21 23 17 20 21 18 16 20 19 21 27 25 20 22 24 25 27 20 Serrano pepper 21 24 21 21 21 21 21 22 19 20 19 19 23 19 20 19 19 19 19 24 19 24 22 20 20 21 22 18 21 19 34 22 21 33 31 32 32 24 28 25 21 22 25 25 28 23 Potato 11 13 18 15 10 17 16 16 14 14 14 15 16 12 15 15 15 14 14 17 15 15 16 15 14 15 18 14 16 15 15 15 15 16 13 14 13 15 17 15 15 16 15 15 16 15 Tomatillo 18 16 18 20 20 17 19 21 19 19 20 20 19 16 15 19 13 14 20 16 16 16 10 17 19 17 20 17 18 9 13 14 15 14 17 15 18 19 17 14 17 17 21 20 14 Tomato ‐ beefsteak 18 20 19 18 16 19 20 21 15 16 17 17 17 13 19 15 19 19 19 19 15 18 15 14 22 19 17 18 20 18 19 21 20 18 17 23 19 17 19 18 18 17 22 27 22 20 Tomato ‐ roma 14 13 17 12 14 14 13 12 13 13 12 14 15 12 12 12 13 12 12 14 15 14 14 11 12 14 13 11 15 11 12 15 15 13 13 12 11 11 15 13 13 14 14 13 15 15 Zucchini 15 15 14 15 13 16 19 20 13 14 15 16 14 13 14 15 15 14 14 17 13 15 14 12 14 15 15 14 16 16 13 15 14 14 13 13 14 15 18 18 14 17 14 18 18 14 Salt 9 10 10 8 8 7 10 10 5 8 8 7 7 7 10 6 8 8 8 7 8 7 6 6 7 9 7 8 7 8 7 8 6 6 9 6 7 7 6 6 4 8 6 7 Sugar ‐ 1kg pack 19 17 15 16 17 15 16 15 15 16 17 16 16 16 16 16 22 19 15 14 15 15 16 13 11 21 20 21 16 19 16 22 17 16 14 13 13 14 16 Sugar ‐ 2kg pack 16 16 16 15 17 15 17 17 17 15 17 17 19 16 19 17 18 22 18 19 17 12 16 15 16 17 16 15 14 16 16 16 20 17 15 15 15 13 14 16 15 14 15 Cooking oil ‐ Mixed 28 28 28 23 24 22 25 27 27 25 26 27 26 27 22 23 24 23 22 24 24 24 26 27 25 25 25 24 22 24 24 21 24 23 22 22 23 22 26 21 22 22 22 24 Cement ‐ 50kg bag 146 141 179 114 132 123 114 135 157 154 99 114 120 102 104 118 128 100 106 115 100 124 113 109 112 110 113 103 120 115 107 108 109 104 98 118 111 110 139 117 104 117 136 125 137 117 21 Table 5. Average price dispersion in cities Northern border Northwest Northeast Center north Center south South MX HMO CHIH LEON MOR MXL ACU MAT HUA MVA DGO MTY TAM QRO TLAX COR CAM OAX LAP JUA TPC CUL TOR JIM FRE AGS COL CTZ GDL TEP ACA IGU TUL TOL CUE PUE VER SAT TAP TEU CHE VSA MID JAC SLP TIJ Beans 27 25 20 26 20 29 16 19 24 32 21 26 19 20 28 27 19 32 27 20 23 22 21 18 17 13 18 32 18 14 19 20 23 21 19 22 17 18 23 18 17 13 13 21 18 22 Rice 27 12 17 14 14 15 8 15 6 19 14 13 5 14 15 13 7 10 15 10 4 15 11 8 9 15 22 6 8 17 10 14 15 11 14 12 17 16 8 13 15 14 8 16 Eggs ‐ per piece 7 5 5 7 6 12 6 8 4 5 5 8 8 3 7 3 6 6 5 3 7 5 7 4 8 6 10 8 8 Eggs ‐ 12 pack 8 7 6 6 4 5 7 6 3 3 5 17 8 8 6 5 7 Eggs ‐ 18 pack 5 4 6 6 5 6 8 5 10 7 Eggs ‐ 30 pack 8 6 7 7 10 11 12 13 4 8 8 5 4 6 8 10 7 3 14 8 Milk 9 5 10 14 7 8 8 8 8 11 13 15 5 5 5 15 10 9 9 8 4 9 13 12 9 8 10 5 6 5 6 8 8 7 13 9 12 4 8 8 6 10 10 7 Bolillo bread 18 27 42 35 45 8 27 40 38 25 45 11 38 34 15 35 20 30 13 42 22 21 36 45 32 33 15 26 24 33 10 8 23 20 13 0 0 35 26 22 2 12 26 10 Concha bread 23 14 8 12 10 14 20 19 21 20 5 21 9 14 5 12 14 7 13 13 18 10 15 7 22 16 15 11 13 16 16 10 12 18 14 13 29 10 25 22 19 11 14 15 15 10 Corn tortillas 14 12 16 15 15 15 12 11 18 18 11 16 14 13 10 12 13 12 14 13 9 13 12 13 9 14 16 10 12 18 5 11 19 13 8 18 12 13 16 10 14 15 21 16 16 8 Water ‐ 1lt bottle 22 9 7 2 11 7 12 26 17 20 13 20 18 9 7 25 8 18 12 14 10 11 6 16 10 11 8 29 23 16 Water ‐ 1.5lt bottle 4 13 2 12 12 7 10 8 6 7 4 15 20 14 16 7 13 13 7 0 7 8 17 12 18 25 17 10 4 14 23 7 10 6 14 Soda ‐ 2lt bottle 11 7 25 11 11 13 7 10 6 8 8 4 13 7 16 14 8 8 7 13 7 11 5 12 16 14 19 6 17 Soda ‐ 600ml bottle 3 10 11 2 3 5 13 3 3 4 9 5 5 7 3 9 4 2 8 7 4 12 9 11 5 9 4 7 7 7 4 5 7 1 5 9 13 11 9 0 10 8 14 8 9 Beef ‐ chop 19 9 13 18 2 20 4 18 15 9 11 3 13 4 9 14 4 6 13 6 6 4 8 5 9 30 18 22 21 23 12 18 5 24 17 Beef ‐ ground 16 18 13 22 34 27 14 7 26 14 23 21 19 24 33 31 20 17 15 16 16 17 25 7 16 11 20 16 12 13 9 16 9 17 13 24 20 10 18 20 19 7 6 20 19 11 Beef ‐ rib 28 17 12 5 32 24 33 7 7 10 19 19 23 8 6 14 18 25 35 18 23 6 27 42 4 18 4 13 12 4 8 23 8 6 11 23 Beef ‐ steak 12 8 12 23 12 21 12 5 17 14 21 25 18 10 14 22 14 13 11 9 12 7 10 10 12 11 12 10 9 10 3 10 8 9 5 12 12 13 13 20 15 9 12 14 16 8 Chicken ‐ whole 10 10 22 18 11 14 7 13 10 6 11 12 14 13 20 11 20 12 10 10 11 15 10 14 11 10 15 11 11 8 7 13 10 11 8 16 7 7 11 6 17 12 11 11 11 10 Chicken ‐ breast 12 12 28 19 8 17 11 10 13 20 15 15 19 5 7 25 10 14 12 3 6 20 6 11 12 9 10 21 10 Chicken ‐ roasted 13 15 22 16 24 8 9 13 4 2 21 11 23 19 19 28 7 6 17 14 13 16 22 16 23 13 21 25 14 5 19 18 9 23 24 30 5 9 7 36 30 11 19 Pork ‐ rib 19 20 19 17 24 21 19 16 27 9 19 32 19 16 17 22 19 13 14 17 19 8 18 11 17 11 10 22 12 9 15 16 11 14 6 12 13 18 18 15 17 6 10 19 21 19 Pork ‐ steak 11 11 12 4 14 10 12 11 12 6 16 11 15 7 14 14 8 9 8 10 11 9 12 3 12 6 6 8 6 5 4 5 8 9 9 10 6 7 21 9 5 9 10 8 10 Apple 12 9 7 9 36 19 10 11 9 16 14 29 11 27 16 12 6 12 9 27 7 10 9 8 20 15 8 16 14 18 4 21 10 12 5 11 13 15 Banana 18 16 11 12 15 17 26 29 23 25 20 28 19 10 19 22 22 17 17 27 17 18 29 5 33 23 33 20 18 20 10 16 18 20 10 41 30 13 12 27 16 11 14 26 33 25 Guava 12 12 12 11 10 9 20 16 15 15 24 23 17 12 24 15 12 29 19 27 17 18 23 6 29 8 14 12 17 17 7 19 18 22 8 15 11 10 19 12 9 8 20 13 20 27 Lime 18 16 19 17 18 14 16 14 13 11 17 18 18 17 18 20 17 16 18 16 19 12 19 14 19 14 15 19 23 12 13 20 16 25 11 17 14 4 33 17 15 26 21 19 20 18 Muskmelon 19 16 20 13 24 18 13 16 20 15 27 20 24 18 21 18 18 12 19 18 18 14 18 11 20 11 17 22 19 18 15 17 10 22 14 13 19 12 16 13 15 12 13 21 22 17 Orange 21 17 19 19 22 24 19 16 30 16 20 23 19 15 21 20 21 32 28 18 18 27 26 11 27 12 15 21 19 18 16 24 27 27 18 17 24 18 21 20 17 19 17 19 28 27 Papaya 19 15 12 17 18 20 26 19 15 11 23 18 20 14 21 23 21 19 24 32 14 17 24 12 25 24 12 29 23 17 14 17 19 24 13 26 21 15 13 15 17 24 28 22 23 17 Pear 12 9 13 19 13 14 14 17 10 17 12 18 10 13 16 13 8 13 14 10 10 19 7 10 6 11 15 12 10 9 14 11 13 10 11 7 12 10 12 9 12 14 Pineapple 11 16 13 9 14 17 16 13 14 11 19 17 21 11 20 24 14 5 19 13 15 12 17 10 15 5 15 12 10 19 33 12 17 14 7 13 11 11 15 16 20 11 Plantain 8 13 20 11 13 7 7 10 18 7 12 8 5 8 5 9 14 9 12 18 14 15 7 14 8 11 15 Watermelon 29 18 20 18 27 20 27 22 36 19 20 25 28 20 18 19 18 19 19 17 15 12 20 12 16 11 17 18 16 13 11 17 13 18 16 17 13 12 14 18 15 13 31 20 21 19 Avocado 17 13 18 17 16 19 17 14 18 17 15 18 17 10 14 21 16 8 15 14 15 9 16 7 12 10 14 22 13 14 12 15 14 15 8 17 13 17 18 16 13 8 12 17 24 15 Cabbage 15 9 30 21 29 21 18 7 16 12 18 60 13 23 31 36 14 32 26 22 19 32 23 15 13 12 38 7 11 10 7 31 16 14 26 14 27 10 28 Carrot 13 20 16 19 17 19 12 18 13 18 19 13 20 20 19 19 12 25 15 19 12 20 15 24 8 16 25 16 16 12 18 14 23 12 18 16 9 14 13 15 14 10 16 19 22 Chayote 14 19 21 13 18 13 15 15 14 10 20 27 13 13 27 18 17 20 24 17 18 11 25 15 29 8 22 19 14 24 8 20 14 19 5 24 28 13 16 15 17 12 12 24 18 23 Cucumber 19 16 18 9 16 12 15 17 21 18 16 19 14 12 16 15 18 20 18 15 19 7 14 12 19 7 15 13 14 13 11 17 11 16 9 24 19 9 17 15 15 13 31 14 22 23 Lettuce 12 20 20 11 13 21 12 16 16 13 19 16 8 17 20 35 15 19 13 17 8 28 21 17 18 12 9 10 19 21 11 6 18 5 21 16 7 13 14 24 Onion 19 22 21 19 24 20 20 22 27 25 24 26 23 23 28 25 22 23 26 22 28 22 32 20 36 24 35 29 22 29 26 24 24 26 19 28 28 22 17 24 15 16 20 29 28 28 Jalapeno pepper 9 8 16 9 20 13 18 20 15 18 11 13 17 12 21 18 16 9 12 9 16 8 11 19 16 23 11 21 21 24 6 14 9 13 17 18 12 12 12 20 Poblano pepper 22 22 18 12 23 14 15 17 16 14 25 16 13 22 32 20 14 21 24 16 21 13 20 11 19 10 17 21 17 14 18 20 21 27 8 16 19 11 16 15 18 12 15 14 17 22 Serrano pepper 22 18 19 17 21 18 19 17 16 19 26 21 18 26 23 22 22 22 23 19 21 11 17 14 18 13 20 30 21 25 7 21 23 22 22 23 23 14 19 13 18 17 17 21 18 21 Potato 29 30 15 18 25 17 13 17 22 21 18 22 22 15 24 22 20 12 21 16 21 16 24 9 23 20 22 21 16 22 21 20 25 24 21 20 24 11 14 17 13 14 18 24 19 21 Tomatillo 21 28 19 13 23 24 21 14 18 17 15 19 29 30 21 28 36 8 12 21 23 19 19 7 15 13 18 10 27 29 26 27 20 22 10 13 15 18 19 19 18 17 29 Tomato ‐ beefsteak 13 14 18 20 16 15 16 24 10 18 23 16 11 27 19 15 16 18 10 26 12 24 12 15 15 18 12 21 10 12 22 7 14 13 12 10 8 14 21 Tomato ‐ roma 28 27 21 23 26 25 19 33 25 21 26 24 24 17 27 31 23 19 26 20 21 16 24 17 24 18 21 32 18 24 10 18 23 27 18 22 17 20 16 14 17 10 18 24 22 20 Zucchini 23 19 21 23 28 17 18 20 25 18 22 21 20 14 24 22 19 20 26 17 21 14 26 15 25 13 22 24 20 16 21 20 21 24 14 24 26 14 22 20 24 13 35 17 28 24 Salt 13 15 14 4 6 15 10 10 5 22 6 12 8 9 17 Sugar ‐ 1kg pack 9 14 7 8 13 8 7 7 5 5 19 10 8 10 9 10 17 16 14 9 11 3 9 8 5 2 1 14 3 16 8 Sugar ‐ 2kg pack 10 6 7 10 11 12 9 11 12 9 14 10 11 11 22 7 15 8 7 25 7 10 12 16 8 12 8 9 6 8 7 18 15 Cooking oil ‐ Mixed 7 7 10 3 3 6 12 2 5 6 5 4 3 6 8 4 7 3 5 5 6 11 8 5 11 7 12 7 5 5 8 Cement ‐ 50kg bag 1 1 1 1 2 2 2 3 3 2 2 9 3 8 6 6 22 Table 6. Average price volatility in cities Northern border Northwest Northeast Center north Center south South MX HMO CHIH LEON MOR MXL ACU MAT HUA MVA DGO MTY TAM QRO TLAX COR CAM OAX LAP JUA TPC CUL TOR JIM FRE AGS COL CTZ GDL TEP ACA IGU TUL TOL CUE PUE VER SAT TAP TEU CHE VSA MID JAC SLP TIJ Beans 6 7 7 7 6 6 6 6 7 5 8 5 8 8 6 6 7 6 6 6 5 6 7 5 6 6 7 6 4 3 4 5 5 4 4 5 3 4 6 6 6 5 4 5 5 5 Rice 4 6 2 9 5 3 4 9 4 4 3 6 4 5 5 4 5 3 8 6 3 5 4 4 5 6 5 7 5 9 3 3 2 3 3 6 3 3 6 7 2 4 8 5 4 4 Eggs ‐ per piece 2 4 9 6 6 8 5 5 6 7 5 8 5 8 6 5 7 4 7 9 9 7 8 6 4 5 7 7 6 6 6 3 6 Eggs ‐ 12 pack 4 10 9 5 7 6 10 4 7 8 3 4 7 6 6 13 5 3 7 6 6 6 7 8 8 6 6 3 9 18 10 4 6 Eggs ‐ 18 pack 1 11 5 7 4 5 6 8 7 9 5 3 8 5 13 9 10 3 11 9 8 10 8 5 6 5 6 1 4 7 7 Eggs ‐ 30 pack 8 8 9 9 8 8 8 7 5 6 6 8 10 11 7 6 8 2 2 5 9 8 6 3 7 8 8 7 6 9 9 7 5 6 9 6 Milk 1 1 2 0 2 2 1 1 1 3 1 1 2 1 1 1 2 1 1 1 1 1 2 1 2 1 1 1 1 1 1 1 1 2 1 2 1 1 1 2 2 2 2 1 2 1 Bolillo bread 2 1 2 1 2 2 3 2 3 1 1 1 1 2 1 2 1 1 1 1 2 1 1 1 1 1 1 1 2 1 0 1 1 0 1 2 0 2 3 3 1 2 2 2 2 Concha bread 1 2 3 3 5 3 2 1 3 2 2 4 2 1 2 2 3 3 1 2 2 1 2 1 1 1 2 2 4 3 3 2 3 3 2 2 2 2 2 3 4 2 2 3 2 3 Corn tortillas 1 2 2 3 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 1 1 2 2 2 2 1 2 2 2 2 3 3 2 2 3 2 2 2 2 2 2 Water ‐ 1lt bottle 5 1 6 2 2 2 5 4 4 1 4 2 4 8 5 4 10 2 1 0 2 4 2 0 3 2 2 10 0 10 1 15 4 4 10 12 4 1 12 5 14 2 9 4 4 3 Water ‐ 1.5lt bottle 5 1 4 1 3 2 2 3 1 2 1 4 4 1 1 1 3 3 1 1 3 6 6 2 7 2 3 2 2 4 4 1 5 0 4 4 8 3 3 6 3 5 Soda ‐ 2lt bottle 4 5 1 8 3 1 3 2 3 7 6 4 7 2 5 1 4 1 2 0 8 4 1 4 1 4 2 1 4 1 1 6 2 1 7 5 7 5 Soda ‐ 600ml bottle 1 3 1 1 3 1 2 1 2 1 2 1 1 2 1 2 1 1 1 1 2 1 2 1 1 1 1 1 1 4 4 1 2 2 1 1 1 1 2 1 1 1 2 2 4 2 Beef ‐ chop 4 1 6 4 4 8 1 3 7 7 3 2 1 6 4 12 2 3 1 2 4 2 2 3 2 3 5 1 4 3 5 1 4 2 3 3 2 5 2 4 3 7 2 6 4 Beef ‐ ground 4 5 4 7 4 4 3 3 3 5 5 5 3 5 3 3 7 5 7 3 3 3 5 3 4 5 3 6 7 6 4 5 5 5 4 5 5 6 6 4 6 4 4 3 5 4 Beef ‐ rib 4 4 5 6 5 5 5 7 2 8 7 6 6 2 10 4 5 2 2 4 3 5 4 2 2 3 4 5 4 2 2 1 4 2 1 6 5 7 3 6 5 2 2 4 4 Beef ‐ steak 3 3 4 4 5 5 3 4 4 3 6 3 5 2 4 4 5 3 2 3 2 3 2 2 3 2 3 4 5 5 2 4 3 3 3 2 4 3 3 5 5 2 4 4 4 3 Chicken ‐ whole 5 7 5 5 8 8 4 5 8 3 4 3 5 5 6 5 7 3 6 3 6 7 5 5 9 6 6 6 10 13 5 7 6 7 8 5 4 8 8 11 8 8 7 6 9 6 Chicken ‐ breast 8 7 1 6 3 2 4 1 6 1 7 1 1 5 5 8 1 6 3 4 3 7 3 11 7 1 3 9 4 8 7 3 4 11 10 5 9 15 2 5 7 Chicken ‐ roasted 1 2 1 3 2 1 1 1 1 1 4 2 1 1 3 2 1 2 1 2 1 2 2 1 2 1 2 2 2 2 1 3 2 1 2 2 2 2 2 4 3 4 1 1 2 2 Pork ‐ rib 4 4 4 4 4 4 4 3 4 3 4 5 5 4 3 4 5 3 4 4 4 4 3 3 2 6 3 4 5 3 5 5 2 5 2 4 4 4 4 5 5 3 3 3 5 5 Pork ‐ steak 6 4 6 8 6 6 3 5 5 8 4 3 4 5 3 5 4 7 6 5 3 5 3 2 4 5 3 8 5 8 4 2 4 6 1 5 4 5 6 5 4 5 6 5 4 5 Apple 14 12 13 12 10 15 19 12 5 14 8 12 6 12 12 13 20 14 12 11 12 9 9 6 10 15 4 5 9 10 4 8 3 8 7 4 10 6 9 6 9 5 11 8 Banana 16 14 12 17 12 18 11 13 13 12 18 16 19 14 12 19 21 9 15 9 13 11 14 12 12 13 13 17 12 16 13 16 17 13 14 10 7 12 12 10 15 10 16 10 14 12 Guava 13 13 10 9 9 12 9 11 7 9 12 12 8 9 13 9 12 11 10 9 13 8 10 13 13 15 13 12 10 14 14 14 12 13 9 9 10 8 12 12 10 8 16 9 11 10 Lime 22 18 20 22 21 23 17 18 19 20 22 22 17 20 19 22 19 19 21 21 20 18 19 17 18 18 18 19 20 24 15 18 19 17 14 21 15 12 20 20 17 21 29 19 23 17 Muskmelon 21 22 18 20 21 20 14 18 22 24 19 19 18 22 18 20 18 18 17 13 15 17 14 15 13 14 14 18 18 20 18 15 18 18 18 18 15 16 16 15 20 20 22 16 15 17 Orange 20 14 17 20 20 21 14 16 20 17 21 18 15 14 16 18 21 17 14 14 16 15 14 12 15 13 17 17 14 18 15 16 16 16 14 16 14 15 16 18 16 13 20 14 18 15 Papaya 13 13 14 16 11 15 12 13 12 13 13 13 11 9 14 13 14 13 14 13 15 15 13 13 15 14 14 15 12 13 12 14 14 14 13 13 13 12 14 11 13 17 18 11 14 13 Pear 12 19 12 15 11 11 8 13 9 14 18 12 9 8 12 12 16 11 12 9 10 9 12 8 12 9 11 10 10 13 11 9 8 9 8 6 7 9 13 10 10 15 13 13 13 10 Pineapple 15 15 18 17 14 11 13 18 13 18 14 14 11 8 14 14 15 16 11 11 13 13 12 11 13 11 13 15 14 15 16 16 15 14 16 14 15 15 10 16 17 19 18 15 14 12 Plantain 6 7 11 2 4 9 6 5 12 6 6 2 11 3 7 5 7 6 9 9 5 8 7 8 12 8 8 8 8 5 8 9 4 9 8 4 9 9 5 10 11 14 8 Watermelon 24 21 21 23 23 23 13 23 27 25 19 23 19 20 20 20 19 18 17 15 14 17 16 12 16 11 14 19 16 20 15 15 16 15 12 14 12 15 17 18 15 12 24 12 18 13 Avocado 14 12 15 18 14 16 14 14 16 17 16 16 12 12 16 14 14 12 12 11 13 11 12 10 13 12 11 13 10 12 11 12 10 11 10 11 11 11 14 13 14 11 10 12 15 10 Cabbage 17 17 27 18 10 14 5 4 15 20 10 15 10 37 13 15 17 10 12 10 16 15 13 11 11 9 10 7 18 12 15 13 14 11 5 11 9 15 13 11 11 22 10 11 14 Carrot 8 11 18 18 12 15 10 11 13 13 15 14 10 10 14 15 14 7 12 12 17 13 12 10 11 12 15 13 12 15 8 13 12 11 6 10 9 9 14 12 9 13 10 12 12 11 Chayote 18 21 19 18 9 12 18 14 20 17 18 15 16 16 15 15 18 14 11 14 20 19 23 17 21 22 20 23 17 18 16 21 25 20 9 19 23 17 17 19 20 17 14 22 22 17 Cucumber 24 24 24 25 22 24 21 20 27 26 20 23 21 21 28 24 16 25 19 11 20 23 14 12 20 16 16 23 18 21 18 22 26 19 10 17 17 13 20 18 24 19 19 18 23 16 Lettuce 23 18 18 23 16 18 15 12 18 21 17 15 23 10 23 11 6 9 12 12 14 12 9 13 7 13 13 10 14 11 9 11 11 8 9 10 6 13 5 13 9 5 8 13 10 Onion 18 18 18 25 19 22 16 18 18 23 20 22 17 20 21 20 20 20 17 16 18 16 15 16 13 17 17 20 16 17 16 17 17 16 14 17 15 16 18 17 16 17 17 17 17 16 Jalapeno pepper 17 21 21 24 16 19 12 15 18 18 21 26 14 20 15 21 20 15 18 10 15 14 17 8 14 21 14 19 17 17 11 15 16 15 12 14 11 16 20 15 12 14 9 24 16 Poblano pepper 18 17 17 19 16 20 18 21 17 14 20 19 14 19 22 19 14 20 19 15 18 18 17 19 17 15 16 20 15 19 16 18 19 19 17 17 15 14 15 14 18 12 16 12 17 18 Serrano pepper 19 18 15 20 14 20 17 14 14 17 17 19 13 15 17 20 19 19 19 16 17 17 14 17 17 15 17 19 16 19 17 16 18 15 15 17 16 14 16 15 15 16 14 14 13 16 Potato 13 14 16 19 16 16 9 11 18 16 15 18 16 17 16 19 19 9 11 11 12 10 12 10 11 14 11 16 12 12 10 12 11 11 7 12 9 10 12 12 10 12 13 11 12 10 Tomatillo 20 21 19 20 14 19 15 14 14 18 17 11 24 16 20 15 18 20 17 16 19 19 22 19 20 19 18 18 22 13 20 20 20 17 16 16 12 18 17 16 19 22 14 18 19 Tomato ‐ beefsteak 24 33 29 35 30 33 24 28 31 21 30 34 23 21 27 24 28 28 27 24 20 22 27 24 23 19 22 31 22 28 25 23 26 25 31 19 21 26 24 24 27 28 19 13 21 23 Tomato ‐ roma 25 26 25 34 30 32 20 25 30 27 29 24 22 26 24 33 26 18 22 17 21 21 21 19 23 23 22 27 20 25 18 18 21 22 21 23 19 22 23 17 22 17 22 22 23 20 Zucchini 24 23 18 24 22 22 19 19 22 22 22 23 18 24 23 23 18 20 22 18 18 18 22 20 23 19 20 22 18 21 18 19 20 19 17 20 19 16 17 15 17 16 21 12 23 19 Salt 4 1 0 1 5 1 6 2 1 9 1 1 1 1 2 1 6 1 1 1 4 1 1 1 3 2 2 3 2 5 2 1 2 1 5 2 5 3 3 3 13 6 9 3 Sugar ‐ 1kg pack 4 4 7 5 8 3 6 5 5 6 5 5 5 5 5 5 2 5 4 4 3 5 5 4 5 5 3 3 4 6 3 4 6 6 4 6 5 5 Sugar ‐ 2kg pack 7 6 5 8 7 6 5 4 8 5 4 9 2 5 4 10 6 1 5 4 5 6 5 5 6 6 5 4 6 6 4 5 2 5 7 7 6 6 6 5 8 5 5 Cooking oil ‐ Mixed 1 1 3 6 3 5 1 3 1 3 2 5 2 1 2 4 2 1 2 4 5 3 1 1 4 2 2 1 4 5 1 3 2 2 2 4 4 3 2 4 3 2 1 2 Cement ‐ 50kg bag 2 1 1 3 2 2 2 3 1 4 2 4 1 1 2 2 2 3 2 2 2 2 2 2 2 2 2 3 2 2 3 2 2 2 3 3 2 2 2 3 3 2 3 5 1 2 23 Table 7. Pass-through magnitude in cities Northern border Northwest Northeast Center north Center south South MX HMO CHIH LEON MOR MXL ACU MAT HUA MVA DGO MTY TAM QRO TLAX COR CAM OAX LAP JUA TPC CUL TOR JIM FRE AGS COL CTZ GDL TEP ACA IGU TUL TOL CUE PUE VER SAT TAP TEU CHE VSA MID JAC SLP TIJ Beans 42 58 65 71 40 63 46 12 21 44 30 25 26 20 26 13 Rice 0 131 0 13 0 0 0 0 14 18 62 0 0 40 0 44 11 0 0 8 0 23 0 Eggs ‐ per piece 41 40 42 42 38 41 40 55 44 23 45 36 43 43 42 54 46 29 45 48 46 41 Eggs ‐ 12 pack 35 0 53 0 17 0 0 24 16 13 13 4 27 0 15 23 19 28 0 0 0 0 18 Eggs ‐ 18 pack 0 0 0 0 0 0 0 35 38 0 44 32 39 17 0 0 0 29 Eggs ‐ 30 pack 20 22 18 27 18 20 12 0 40 0 0 0 0 0 33 0 27 39 46 28 47 48 0 Milk Bolillo bread Concha bread Corn tortillas Water ‐ 1lt bottle Water ‐ 1.5lt bottle Soda ‐ 2lt bottle Soda ‐ 600ml bottle Beef ‐ chop 77 48 61 58 69 55 16 79 0 69 25 45 63 60 105 57 59 64 37 60 97 12 88 57 72 67 61 72 129 70 51 74 25 69 82 19 68 79 54 82 73 70 Beef ‐ ground 71 100 49 58 89 99 55 67 26 66 65 49 51 73 71 36 72 70 79 32 85 44 42 61 82 61 52 65 89 65 63 57 66 58 86 61 76 86 70 89 78 67 Beef ‐ rib 70 79 68 45 59 83 59 88 53 58 84 78 28 40 74 61 38 75 55 81 48 70 75 93 85 89 84 76 95 88 58 Beef ‐ steak 58 53 61 64 52 83 64 64 66 73 44 62 59 69 57 56 58 62 66 59 59 71 61 53 74 57 67 57 63 68 54 56 56 66 60 61 67 57 64 71 67 76 71 73 65 68 Chicken ‐ whole 93 129 53 103 107 83 108 100 151 156 158 42 110 126 86 112 138 113 109 136 93 96 97 128 95 49 115 80 132 138 98 90 112 100 94 96 116 144 97 27 93 110 103 Chicken ‐ breast 111 98 405 60 65 178 138 116 97 116 128 81 107 117 118 140 177 112 110 114 67 119 94 98 124 67 113 112 23 87 109 97 96 Chicken ‐ roasted 159 62 107 15 117 62 70 73 57 84 101 99 53 111 106 108 192 102 77 59 115 95 86 171 93 80 140 65 90 91 78 112 84 111 80 110 25 68 78 97 Pork ‐ rib 20 38 43 36 41 33 33 0 40 39 33 36 31 33 26 16 40 36 59 32 30 21 34 20 40 37 38 21 40 26 28 17 43 30 34 33 29 40 38 29 Pork ‐ steak 36 28 33 31 29 28 34 25 40 10 25 20 29 23 27 21 24 24 36 27 34 47 22 23 0 19 14 16 22 17 32 51 22 32 20 24 33 18 28 30 27 Apple 41 45 71 39 53 28 32 23 56 92 60 54 39 48 0 37 58 30 40 33 63 45 48 56 45 36 58 44 72 52 59 40 41 24 48 18 70 52 50 Banana 90 114 59 171 99 151 155 120 126 49 150 111 155 159 191 133 52 317 119 140 36 109 118 0 115 65 123 132 156 134 157 130 124 0 117 139 116 47 98 101 98 174 117 108 Guava Lime 34 34 32 35 39 35 40 31 32 37 38 39 35 33 35 35 36 39 37 37 36 35 38 32 34 31 31 36 43 38 34 35 38 34 29 47 40 41 59 43 39 35 8039 52 35 Muskmelon 52 56 24 42 48 37 23 22 37 32 50 46 41 76 52 42 29 40 34 17 17 26 22 29 20 26 18 39 17 13 17 18 22 19 11 22 22 21 20 15 18 14 1819 18 20 Orange 42 33 0 0 40 35 22 31 31 0 0 36 28 0 30 28 0 0 27 28 19 0 0 0 0 40 35 28 30 29 0 33 0 29 0 35 30 30 53 57 32 42 5630 43 30 Papaya 36 54 31 50 29 53 44 43 49 28 41 44 35 42 45 44 41 44 51 42 46 70 42 53 59 42 60 47 58 61 67 63 74 65 63 45 68 52 56 22 58 55 3847 52 67 Pear 38 63 19 52 33 0 27 36 20 30 28 40 27 33 21 20 0 15 28 33 29 38 37 32 33 36 14 42 48 65 36 23 52 40 17 25 23 27 46 31 34 65 3334 45 50 Pineapple 0 0 0 26 0 0 0 0 0 0 0 29 0 0 0 20 0 0 0 16 25 16 17 17 0 14 0 21 0 28 0 26 30 51 0 27 0 0 29 42 54 5224 0 26 Plantain 69 48 0 0 85 37 106 98 0 61 0 0 114 80 110 50 197 46 82 59 0 168 112 95 93 99 132 48 80 112 14887 0 95 Watermelon 42 52 23 43 44 41 12 25 41 27 42 38 38 57 33 39 29 30 26 18 13 26 21 23 24 20 17 30 24 31 22 20 28 25 16 22 22 26 19 24 22 15 2622 21 21 Avocado 96 90 89 101 74 92 86 78 88 73 93 103 91 95 96 93 93 89 93 81 84 95 88 81 98 83 86 99 85 86 88 81 80 84 86 93 82 75 83 77 80 61 5775 89 86 Cabbage 37 27 46 49 35 16 42 0 30 94 28 0 19 31 0 23 39 19 32 21 0 0 13 30 27 0 41 14 33 26 11 24 20 19 13 0 26 3926 30 30 Carrot 0 52 25 42 28 36 33 29 44 0 42 40 33 31 37 39 36 36 40 41 28 42 32 24 37 35 32 39 36 37 33 36 37 35 35 38 41 38 38 26 34 34 4343 35 31 Chayote 38 57 50 35 15 33 53 43 53 0 48 41 48 40 46 44 47 42 40 40 56 77 65 58 67 62 51 72 57 52 59 49 98 72 26 63 87 66 51 48 63 47 3967 70 60 Cucumber 38 50 36 42 28 37 0 39 52 30 21 40 43 38 32 41 11 17 18 0 18 37 9 0 33 19 21 28 31 15 29 28 49 29 18 11 18 11 19 14 37 26 024 16 30 Lettuce 28 28 9 13 12 15 10 11 15 14 14 13 21 8 17 0 0 0 8 0 8 8 13 0 0 10 7 9 7 0 11 5 0 6 0 0 14 0 10 6 0 0 12 6 Onion 36 38 26 35 32 32 29 29 32 20 29 34 30 25 35 30 40 25 26 26 33 30 32 33 25 21 32 37 31 35 28 28 28 26 26 30 29 30 32 31 30 24 2929 27 27 Jalapeno pepper 50 66 61 121 58 81 38 43 60 56 79 99 46 91 78 89 73 67 73 32 39 53 58 41 71 35 74 72 79 48 59 71 67 58 63 72 65 61 48 32 36 22 26 69 Poblano pepper 48 56 38 86 61 92 74 69 50 47 79 91 59 85 97 86 66 98 84 63 57 80 71 67 76 64 57 91 69 69 74 75 80 83 71 68 69 63 64 51 73 41 52 51 74 79 Serrano pepper 45 56 31 93 45 93 60 45 52 46 68 81 42 56 68 90 68 79 77 46 48 65 45 53 66 59 55 80 73 78 43 55 73 48 48 44 42 55 51 45 46 48 44 39 41 62 Potato 19 14 0 12 0 0 0 0 15 Tomatillo 66 73 43 81 46 77 47 33 38 73 70 44 63 64 80 49 71 72 72 30 85 73 66 74 65 78 69 78 98 78 93 87 80 68 81 53 76 69 55 68 78 55 63 85 Tomato ‐ beefsteak 47 56 48 72 43 66 47 53 58 47 62 54 44 25 46 48 51 65 45 44 25 49 53 37 47 39 22 53 49 55 36 34 63 55 62 34 40 62 40 32 50 101 35 0 30 45 Tomato ‐ roma 42 43 33 39 36 40 29 41 52 39 47 38 35 42 29 41 28 23 31 28 28 33 34 25 36 38 28 32 28 27 23 27 30 23 26 35 22 24 24 15 28 14 24 25 20 31 Zucchini 22 29 11 15 0 13 7 9 23 16 11 11 12 19 0 13 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7 0 0 0 0 0 10 0 Salt Sugar ‐ 1kg pack 103 0 51 40 69 51 53 42 53 58 107 38 71 54 23 33 71 48 46 49 44 48 55 51 56 31 53 51 Sugar ‐ 2kg pack 141 71 60 75 0 42 71 0 54 17 23 34 43 62 57 0 50 71 77 67 133 82 63 57 80 57 73 87 76 0 65 61 62 Cooking oil ‐ Mixed 0 37 37 10 17 0 74 38 0 30 8 31 18 16 14 32 Cement ‐ 50kg bag 24 Table 8. Pass-through speed of adjustment in cities Northern border Northwest Northeast Center north Center south South MX HMO CHIH LEON MOR MXL ACU MAT HUA MVA DGO MTY TAM QRO TLAX COR CAM OAX LAP JUA TPC CUL TOR JIM FRE AGS COL CTZ GDL TEP ACA IGU TUL TOL CUE PUE VER SAT TAP TEU CHE VSA MID JAC SLP TIJ Beans 9 16 0 11 7 7 13 10 10 18 11 7 9 15 21 7 Rice 52 74 47 15 17 60 46 19 36 47 54 13 30 34 68 16 11 54 36 54 38 35 11 Eggs ‐ per piece 8 4 8 3 3 13 0 5 5 22 4 4 8 5 4 6 3 0 4 5 0 2 Eggs ‐ 12 pack 20 49 9 63 67 58 35 13 37 45 47 41 59 0 51 7 49 67 34 24 54 84 5 Eggs ‐ 18 pack 0 48 20 89 19 33 17 34 24 60 20 8 15 52 64 99 27 9 Eggs ‐ 30 pack 15 17 26 5 15 11 9 20 11 72 79 28 34 61 12 30 21 5 5 7 0 22 75 Milk Bolillo bread Concha bread Corn tortillas Water ‐ 1lt bottle Water ‐ 1.5lt bottle Soda ‐ 2lt bottle Soda ‐ 600ml bottle Beef ‐ chop 21 26 23 21 3 45 0 5 72 18 19 6 0 6 65 7 4 0 0 11 6 4 3 8 14 12 6 8 8 6 0 7 0 6 37 36 25 15 30 0 11 4 Beef ‐ ground 0 4 0 13 17 2 8 5 11 10 9 7 12 7 8 17 8 17 0 8 6 0 9 4 6 8 15 9 4 9 8 6 7 3 6 7 4 5 8 6 11 2 Beef ‐ rib 11 10 10 30 13 16 36 17 31 0 23 5 7 8 9 24 33 4 13 4 6 22 11 35 5 23 7 0 0 7 4 Beef ‐ steak 5 5 5 9 5 6 4 7 7 9 18 4 11 2 12 6 6 7 3 4 0 6 0 0 4 3 4 5 6 9 2 6 3 3 5 0 4 4 4 7 7 0 4 3 4 2 Chicken ‐ whole 38 30 35 43 16 33 24 30 16 0 17 23 51 19 26 21 17 15 36 15 16 43 24 27 24 46 63 27 47 26 18 32 23 23 41 38 67 52 48 27 24 58 16 Chicken ‐ breast 43 74 28 23 9 14 6 33 24 48 47 28 21 33 27 36 12 65 83 30 61 21 53 60 22 26 75 37 31 73 16 43 48 Chicken ‐ roasted 12 23 10 26 2 5 33 17 23 17 18 19 35 31 12 10 14 12 8 11 13 23 19 18 26 19 17 22 49 28 7 18 24 32 69 18 40 68 31 0 Pork ‐ rib 15 9 0 11 8 14 6 69 9 9 7 7 18 7 8 8 8 7 0 8 20 4 21 10 5 11 7 7 7 11 7 9 21 5 6 8 9 5 10 6 Pork ‐ steak 12 12 15 10 7 9 13 51 7 11 22 10 8 18 22 18 14 7 12 5 0 5 20 8 38 12 34 23 5 9 13 0 14 19 15 23 7 19 10 9 6 Apple 57 48 18 43 59 23 27 18 19 15 39 38 34 21 146 67 54 29 53 40 17 39 41 59 34 9 45 18 6 20 13 16 18 28 23 33 54 19 0 Banana 19 16 15 19 15 29 16 15 14 37 22 19 21 13 13 10 31 26 12 35 26 15 15 20 26 40 17 49 13 15 16 9 20 41 19 26 17 46 30 20 46 10 13 9 Guava Lime 0 0 0 0 0 0 0 0 0 10 0 0 0 0 0 0 0 0 0 0 0 0 0 9 0 0 0 0 0 21 0 0 0 0 0 0 0 0 17 19 0 19 17 0 15 0 Muskmelon 29 40 43 24 29 25 22 23 33 62 15 16 22 20 13 21 18 25 26 22 34 26 19 46 17 25 20 21 28 41 33 17 28 24 37 25 25 47 24 24 29 88 48 23 23 20 Orange 19 18 20 13 13 9 9 13 13 20 9 9 8 10 9 6 13 10 6 8 11 10 9 11 8 8 9 8 8 11 6 9 7 8 8 13 7 9 7 10 9 14 15 9 9 0 Papaya 17 10 28 23 20 17 23 21 19 26 14 16 18 22 26 14 17 33 22 27 26 19 17 32 22 24 18 19 17 22 13 15 15 14 17 16 12 20 17 22 18 43 33 15 18 11 Pear 29 34 19 37 40 68 16 20 30 31 29 23 29 27 17 10 23 34 23 17 25 18 17 29 21 13 22 16 16 24 19 19 22 16 44 42 11 29 19 39 20 36 35 42 20 12 Pineapple 13 16 40 19 10 8 20 19 14 41 11 21 10 14 10 10 26 28 14 11 13 10 17 17 14 14 18 11 31 17 11 13 10 36 15 12 14 7 18 22 29 25 12 30 6 Plantain 28 52 41 0 22 38 8 33 52 59 10 16 7 7 0 22 7 33 19 53 31 0 12 10 16 22 8 14 18 13 31 28 46 4 Watermelon 30 26 30 22 32 18 41 30 33 47 17 21 23 25 21 17 15 29 26 21 35 26 21 29 16 29 26 24 20 31 31 20 25 20 41 16 18 24 16 30 19 27 28 17 19 13 Avocado 18 13 22 16 19 14 10 16 15 32 15 17 11 11 15 11 19 24 11 12 19 13 9 17 13 16 10 9 12 17 13 17 10 9 13 11 9 18 16 20 12 21 24 13 19 8 Cabbage 26 54 43 22 24 19 20 35 17 15 14 105 33 24 21 25 22 27 30 10 20 23 27 20 28 23 29 29 19 12 42 19 27 43 16 43 33 34 11 13 31 Carrot 65 26 39 30 34 18 17 15 13 35 21 13 15 28 18 16 13 20 18 15 20 13 17 20 13 21 16 9 8 12 7 10 7 7 8 0 0 10 8 15 0 15 6 8 9 0 Chayote 26 28 20 25 18 11 22 17 32 29 24 15 20 43 15 12 22 19 22 18 32 16 18 30 22 29 33 19 22 34 28 34 28 22 34 24 29 26 24 25 23 39 41 24 19 17 Cucumber 37 30 34 32 23 24 32 43 43 45 22 24 24 32 31 24 26 30 16 14 36 31 18 20 22 29 26 24 28 32 37 43 37 24 19 24 29 29 22 29 35 40 42 23 23 23 Lettuce 42 42 34 52 37 26 32 20 16 29 25 24 53 26 31 36 31 26 27 19 16 21 46 15 29 37 19 31 32 29 27 25 36 27 19 26 21 19 19 31 26 32 19 18 Onion 6 0 7 7 6 8 10 10 6 14 5 7 6 9 7 6 7 8 8 8 10 7 0 13 8 11 12 7 6 13 8 6 7 7 8 7 8 8 7 8 7 11 8 8 10 6 Jalapeno pepper 51 47 56 65 30 56 40 33 47 34 47 71 38 41 47 54 59 58 50 44 41 20 49 63 66 35 51 41 65 33 32 51 35 45 36 29 57 45 46 58 40 23 76 25 Poblano pepper 22 17 9 28 22 19 18 13 28 14 31 19 29 25 17 14 13 26 16 17 32 33 11 51 17 27 11 14 13 32 20 22 10 12 17 30 25 19 18 19 26 23 29 20 21 0 Serrano pepper 28 36 22 39 28 53 32 29 48 39 39 35 35 40 32 49 32 49 42 25 30 32 25 51 38 29 34 38 24 34 18 24 32 58 41 25 28 30 21 18 30 47 27 25 17 26 Potato 20 15 21 14 14 11 20 13 10 Tomatillo 26 35 42 38 29 41 40 34 34 45 33 24 78 33 18 24 23 17 35 27 23 36 37 33 48 31 29 30 24 37 0 19 0 26 15 27 38 30 33 33 31 29 33 0 Tomato ‐ beefsteak 32 60 32 41 40 36 34 34 36 32 32 25 31 34 30 26 26 50 30 28 40 31 41 36 47 34 37 38 40 49 31 38 48 32 45 30 27 37 31 37 35 56 36 45 41 33 Tomato ‐ roma 28 26 23 34 33 27 21 21 27 35 27 22 25 34 23 25 24 20 21 23 28 27 20 28 21 34 29 24 23 23 22 23 21 20 23 21 17 21 31 20 22 22 28 21 28 21 Zucchini 37 30 37 27 13 21 36 38 37 47 18 21 16 25 22 18 19 20 23 33 22 21 30 35 28 29 28 20 24 30 22 25 28 25 29 21 26 25 37 21 37 39 52 24 27 24 Salt Sugar ‐ 1kg pack 18 66 12 10 8 5 7 12 0 9 16 6 8 4 24 33 0 17 0 6 10 27 5 0 12 11 20 3 Sugar ‐ 2kg pack 31 12 5 9 31 9 21 0 5 27 29 21 5 5 4 23 6 6 6 6 6 3 5 6 3 11 5 5 3 42 9 3 2 Cooking oil ‐ Mixed 13 24 14 13 22 29 0 51 38 16 19 33 10 114 19 38 Cement ‐ 50kg bag 25 Table 9. Pass-through magnitude asymmetry in cities Northern border Northwest Northeast Center north Center south South MX HMO CHIH LEON MOR MXL ACU MAT HUA MVA DGO MTY TAM QRO TLAX COR CAM OAX LAP JUA TPC CUL TOR JIM FRE AGS COL CTZ GDL TEP ACA IGU TUL TOL CUE PUE VER SAT TAP TEU CHE VSA MID JAC SLP TIJ Beans 0 0 0 ‐19 0 0 0 0 0 0 ‐20 0 0 0 0 0 Rice 0 0 0 22 24 80 ‐31 ‐23 0 49 ‐65 0 0 0 0 0 0 61 0 0 ‐34 0 21 Eggs ‐ per piece 0 15 0 0 0 0 0 0 0 0 20 22 0 16 0 0 14 0 0 19 0 20 Eggs ‐ 12 pack 0 0 0 0 246 ## 81 ‐46 0 0 0 ‐70 ## 0 ‐60 0 0 0 0 0 ## 0 0 Eggs ‐ 18 pack ## 0 0 117 0 0 0 0 ‐99 0 ‐41 ‐27 ‐50 0 ## 0 0 0 Eggs ‐ 30 pack 0 0 ‐64 0 ‐33 0 0 ‐71 0 122 0 0 0 0 ‐42 0 0 0 0 0 0 0 92 Milk Bolillo bread Concha bread Corn tortillas Water ‐ 1lt bottle Water ‐ 1.5lt bottle Soda ‐ 2lt bottle Soda ‐ 600ml bottle Beef ‐ chop 62 0 ‐44 0 0 ‐48 0 0 0 0 ‐33 0 0 0 0 0 ‐14 ‐7 0 ‐37 0 0 0 0 ‐31 0 0 0 ‐19 0 ‐10 0 0 0 0 ‐33 ‐86 0 0 0 0 ‐13 Beef ‐ ground 0 0 0 0 0 ‐14 0 0 0 0 0 0 0 0 0 0 ‐19 ‐35 0 0 0 0 0 ‐14 0 16 ‐20 ‐13 ‐17 0 0 0 12 0 0 ‐18 0 0 0 ‐24 0 ‐6 Beef ‐ rib 0 ‐32 0 0 0 0 30 ‐41 0 0 0 ‐12 0 0 ‐49 0 0 0 0 0 ‐6 0 ‐20 ‐45 ‐13 ‐91 ‐29 0 0 0 0 Beef ‐ steak 0 0 0 ‐30 14 ‐23 0 0 ‐13 0 23 0 ‐20 0 0 0 0 0 ‐10 0 0 0 0 0 0 12 0 0 0 0 0 0 0 ‐7 ‐11 0 ‐10 0 0 ‐25 ‐16 9 9 ‐19 ‐12 ‐5 Chicken ‐ whole ‐57 0 ‐34 ‐57 ‐29 ‐47 0 0 0 0 ‐39 0 0 0 0 0 0 0 0 29 0 0 0 ‐78 ‐33 0 ‐59 0 0 0 0 0 0 ‐47 0 0 ‐43 0 0 0 0 0 0 Chicken ‐ breast 0 0 0 ‐22 0 0 0 ‐58 26 0 0 0 0 ‐36 ‐54 0 0 0 0 ‐67 ‐89 0 ‐49 ‐56 0 0 0 0 0 ‐49 0 0 ‐41 Chicken ‐ roasted 0 ‐66 29 0 0 0 ‐45 0 34 0 0 0 ‐83 ‐59 0 ‐30 0 0 0 0 0 0 0 0 36 0 ‐26 0 ‐70 61 15 0 0 ‐76 0 0 ‐93 117 ‐61 0 Pork ‐ rib 0 0 13 0 0 0 0 0 0 0 11 0 0 0 0 0 0 0 0 0 0 0 0 0 ‐16 ‐29 0 ‐12 0 0 0 0 ‐31 ‐9 0 ‐18 ‐25 0 0 0 Pork ‐ steak ‐15 0 0 14 0 0 0 ‐55 0 0 0 0 0 0 0 ‐43 0 ‐21 0 ‐11 17 0 ‐44 ‐19 ‐47 ‐14 ‐74 0 0 ‐25 0 0 22 ‐40 ‐23 0 0 0 ‐15 ‐19 0 Apple 0 ‐35 0 36 ‐67 0 0 0 0 ‐28 ‐66 0 0 0 0 0 0 0 ‐85 0 0 0 ‐28 0 ‐47 0 ‐51 0 0 0 0 0 0 0 0 47 0 24 0 Banana 0 0 0 0 0 0 ‐25 0 0 0 ‐32 0 0 0 ‐42 0 0 0 0 0 0 0 0 0 43 33 0 ‐28 0 0 0 0 0 0 ‐52 0 0 ‐59 0 ‐21 0 0 0 0 Guava Lime 0 0 0 00 0 0 0 0 0 0 0 0 0 0 0 0 ‐31 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ‐30 0 0 0 38 0 0 0 0 0 0 Muskmelon 0 0 0 ‐27 34 0 0 0 0 ‐64 0 0 0 0 0 ‐32 0 0 0 0 0 0 0 0 0 ‐38 0 0 0 0 0 0 0 0 0 ‐34 0 0 0 0 ‐36 ## ‐38 0 0 0 Orange 0 0 0 00 0 0 0 0 0 0 0 14 0 0 0 17 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 17 Papaya 0 0 0 410 34 23 28 0 0 0 45 42 37 32 0 32 23 19 35 0 0 34 0 27 28 35 28 33 23 22 36 0 0 0 0 0 0 19 0 0 0 0 24 0 23 Pear 0 26 0 0 ‐98 37 ‐50 17 49 0 0 0 0 28 0 0 15 0 0 0 0 0 22 0 0 0 27 0 0 0 0 21 30 22 ‐38 0 0 ‐27 0 0 0 46 0 ‐48 0 23 Pineapple 0 0 ‐52 00 0 0 0 0 ‐51 0 0 0 0 0 0 0 0 0 0 0 0 0 ‐17 0 ‐18 0 0 ‐27 ‐36 0 0 0 ‐74 0 0 0 15 0 0 ‐31 0 0 0 0 Plantain 0 ‐76 0 0 ‐38 0 0 0 ‐98 67 ‐26 ‐34 0 0 0 ‐26 0 0 0 ‐84 0 0 0 0 0 0 0 0 0 33 0 0 ‐56 15 Watermelon 0 0 0 0 ‐40 0 ‐29 ‐38 ‐33 0 ‐29 0 0 0 0 ‐30 0 0 0 ‐29 0 0 0 0 ‐30 0 0 0 ‐29 ‐39 ‐31 0 ‐28 0 ‐44 0 0 ‐47 ‐25 ‐63 0 0 ‐49 0 0 0 Avocado 29 24 0 00 0 0 0 0 ‐37 0 0 0 22 0 0 61 0 0 0 0 36 0 0 22 0 0 29 0 23 24 0 0 0 0 0 0 21 0 0 0 0 0 0 39 0 Cabbage 0 0 ‐42 0 ‐20 0 0 ‐38 0 0 0 ## ‐34 0 0 0 32 23 ‐27 0 0 0 0 ‐43 ‐53 0 ‐33 0 ‐19 0 0 ‐52 ‐33 0 0 0 ‐35 0 0 0 0 Carrot 133 0 0 00 0 0 0 0 0 ‐37 0 0 0 33 22 0 36 43 19 0 0 40 0 0 0 0 0 0 0 0 0 0 20 22 14 0 0 0 0 0 0 0 0 0 17 Chayote 0 0 0 41 0 31 0 0 0 0 0 0 29 0 41 43 0 22 0 32 0 31 31 31 21 0 0 0 38 0 57 0 0 23 0 0 0 0 0 0 47 0 33 37 0 24 Cucumber 0 31 0 00 29 0 0 0 0 0 0 0 45 32 0 0 0 0 0 0 0 43 42 27 27 37 35 0 0 50 45 49 0 35 0 0 ‐25 23 37 39 0 0 0 0 38 Lettuce 0 0 ‐52 00 0 0 0 0 0 0 0 0 0 29 0 0 0 0 0 0 0 0 0 0 ‐29 0 31 0 0 0 0 0 0 0 0 29 0 0 0 ‐50 0 ‐26 0 Onion 0 0 0 ‐17 0 0 0 0 ‐16 ‐19 0 ‐15 0 0 0 ‐16 0 0 0 0 ‐23 ‐15 ‐13 0 ‐22 0 ‐24 0 0 ‐28 0 0 0 0 0 0 0 0 ‐14 0 0 0 0 0 ‐16 0 Jalapeno pepper 0 0 0 26 0 ‐55 0 ‐43 0 0 36 0 0 0 0 0 ‐50 0 52 0 0 41 0 0 0 0 32 0 0 0 0 39 0 0 0 0 0 0 39 0 0 0 0 0 Poblano pepper 0 0 0 29 0 0 0 0 27 0 0 0 0 ‐43 0 0 ‐59 ‐45 0 0 0 38 ‐37 0 0 37 0 0 0 0 0 0 0 0 0 0 36 0 0 0 ‐39 27 ‐48 0 0 0 Serrano pepper 0 0 0 0 33 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 21 0 0 0 0 41 0 0 0 ‐19 0 0 0 0 0 0 0 0 0 0 ‐42 0 35 0 0 Potato 0 0 0 19 0 0 0 0 0 Tomatillo 35 0 0 0 0 0 0 0 0 ‐24 0 0 0 0 0 0 35 39 0 0 54 0 41 0 ‐33 52 0 72 0 0 0 0 61 52 64 0 0 0 0 43 0 0 49 53 Tomato ‐ beefsteak 0 0 0 ‐50 0 ‐31 ‐31 0 0 0 0 ‐42 0 0 0 0 0 0 0 ‐19 0 0 0 54 ‐53 0 0 0 0 0 0 ‐31 0 0 0 0 0 0 ‐30 0 0 ‐84 ‐39 0 0 0 Tomato ‐ roma 0 25 0 0 0 0 21 0 0 0 0 0 0 0 0 30 0 21 0 30 0 0 0 35 0 0 0 0 0 0 0 0 0 25 0 0 0 0 ‐27 0 0 0 0 0 ‐22 24 Zucchini 0 0 36 0 0 0 36 ‐41 0 ‐42 0 0 16 0 0 0 0 0 0 0 ‐30 0 0 0 0 0 0 ‐27 0 0 0 0 0 0 0 0 0 0 0 ‐24 0 ‐24 ‐78 0 0 0 Salt Sugar ‐ 1kg pack 0 0 ‐37 0 0 0 0 ‐36 0 ‐19 0 0 0 ‐14 0 0 0 ‐38 0 0 0 ‐57 0 0 ‐31 0 0 0 Sugar ‐ 2kg pack 0 0 0 0 95 0 ‐32 0 0 0 ‐30 0 ‐24 0 ‐22 0 12 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Cooking oil ‐ Mixed 0 ‐33 0 0 0 0 0 0 0 ‐39 ‐30 0 0 0 0 0 Cement ‐ 50kg bag 26 REGRESSION RESULTS We exploit data granularity to estimate regressions of indicators at good-city-outlet level, which more are disaggregated than the indicators shown in the heat maps. First, we study which fixed effect characteristics of markets. Then, we run regressions to determine how markets respond to a large retailer’s entries and to COFECE sanctions. Fixed effect regressions Fixed effects (FE) regressions provide another way to analyze how markets are behaving. Table 10 shows the amount of variation (R-squares) explained by each kind of fixed effects by themselves on pooled regressions that include all goods. For example, when running a regression of dispersion with only good fixed effects, these explain about 9% of the variation of such indicator. Following the same intuition, it can be seen that the characteristics of goods play an important role in explaining the variation of dispersion and volatility, something already seen in the heat map. City characteristics play a more important role in the magnitude of pass-through. Outlet-type fixed effects seem to be less relevant. We also notice that time fixed effects explain relatively little of the variation in dispersion and volatility. This means that these indicators do not change substantially throughout the time period covered by this series. Table 10. R-squared values of fixed effect regressions Dispersion Volatility Differential PT magnitude PT speed PT symmetry Good FE 8.79% 51.35% 65.94% 1.19% 4.26% 4.26% City FE 3.91% 2.66% 0.84% 3.77% 1.73% 1.20% Outlet FE 1.46% 2.55% 1.40% 2.12% 0.72% 0.85% Time FE 0.15% 0.45% 1.19% ‐‐ ‐‐ ‐‐ Total (sum) 14.31% 57.02% 69.36% 7.08% 6.71% 6.31% R‐squared of fixed effect regressions containing only the controls described in each row. The row "Total (sum)" is for illustrative reasons only, as it adds the r‐squareds of separate FE regressions and does not necessarily coincide with pooled FE regressions containing all of these controls. Fixed-effects specifications. We also estimate fixed effect regressions with several types of dummies, as opposed to the results shown in the table above. These specifications are intended to associate differences in the patterns of indicators to markets across time periods in a more precise way. Our specifications include City, Good, Outlet and Time dummies in order to analyze their relevance in explaining the variation of the behavior of each of the three price indicators—dispersion, pass-through and volatility. (7) Equation (7) is the pooled specification we estimate. Variable y denotes the indicator that is being regressed and subscripts g, c, o, t refer to varieties of good, city, outlet and time periods, respectively. Variables d denote dummies for each category. 27 Table 11. City fixed effects results (1) (2) (3) (4) (5) Dispersion Volatility PT magnitude PT speed PT asymmetry City FE (Mexico City excluded) Mexicali ‐3.017*** 2.120*** ‐0.054*** 0.156*** 0.023*** Northern border Tijuana ‐2.569*** 1.772*** 0.052*** 0.162*** 0.049*** La Paz ‐1.514*** 0.621*** ‐0.170*** 0.175*** 0.008 Cd. Acuna ‐2.107*** 1.230*** 0.021* 0.056*** ‐0.030*** Cd. Juarez ‐1.433*** 0.537*** ‐0.057*** 0.090*** ‐0.026*** Matamoros ‐1.962*** 1.651*** 0.009 0.094*** 0.080*** Tepic ‐3.247*** 0.164* ‐0.025*** 0.098*** ‐0.005 Northwest Culiacan ‐3.741*** 0.861*** ‐0.003 0.079*** 0.014* Hermosillo ‐2.412*** 1.670*** 0.013 0.176*** 0.027*** Huatabampo ‐3.673*** 1.120*** ‐0.158*** 0.204*** 0.079*** Torreon ‐2.440*** 2.037*** ‐0.023*** 0.084*** 0.023*** Monclova 1.807*** ‐0.205* ‐0.088*** 0.007 ‐0.077*** Chihuahua ‐0.911*** ‐1.189*** ‐0.067*** 0.060*** 0.092*** Northeast Jimenez ‐4.066*** 1.400*** 0.061*** 0.117*** 0.079*** Durango ‐3.174*** 1.361*** ‐0.009 0.084*** 0.130*** Monterrey ‐0.924*** 2.123*** 0.051*** 0.078*** 0.001 Tampico ‐1.968*** 1.897*** ‐0.167*** 0.122*** ‐0.014*** Fresnillo ‐6.696*** 0.719*** 0.004 0.146*** 0.058*** Aguascalientes ‐2.149*** 0.979*** ‐0.009 0.081*** 0.006 Colima ‐3.848*** ‐1.226*** ‐0.049*** 0.091*** ‐0.036*** Leon ‐2.624*** 0.639*** ‐0.067*** 0.051*** 0.025*** Center‐North Cortazar ‐6.113*** ‐0.501*** ‐0.031*** 0.056*** ‐0.058*** Guadalajara ‐1.109*** ‐0.142*** ‐0.030*** 0.038*** 0.008* Tepatitlan ‐9.216*** ‐0.400*** ‐0.058*** 0.079*** ‐0.033*** Morelia ‐4.830*** 0.408*** ‐0.038*** 0.105*** 0.056*** Jacona ‐8.472*** ‐0.685*** ‐0.071*** 0.056*** ‐0.089*** Queretaro ‐3.014*** 0.747*** ‐0.042*** 0.103*** 0.011* San Luis Potosi ‐3.310*** 1.210*** 0.067*** 0.083*** 0.040*** Acapulco ‐3.169*** ‐0.364*** ‐0.027*** 0.074*** 0.034*** Iguala ‐6.064*** 1.174*** ‐0.032*** 0.179*** 0.118*** Tulancingo ‐6.413*** ‐0.530*** ‐0.058*** 0.050*** 0.074*** Center‐South Toluca ‐2.362*** ‐0.328*** ‐0.057*** 0.078*** ‐0.022*** Cuernavaca ‐5.639*** 0.712*** ‐0.061*** 0.113*** 0.028*** Puebla ‐3.033*** 0.187*** 0.040*** 0.062*** 0.028*** Tlaxcala ‐4.991*** ‐0.074 0.027*** 0.109*** 0.039*** Veracruz ‐3.393*** 0.134 ‐0.099*** 0.091*** 0.035*** Cordoba ‐2.456*** ‐0.663*** 0.012* 0.054*** 0.070*** San Andres Tuxtla ‐6.188*** ‐1.218*** ‐0.058*** 0.093*** 0.009 Campeche ‐2.991*** ‐0.353*** ‐0.087*** 0.081*** 0.048*** Tapachula ‐4.030*** ‐0.534*** ‐0.152*** 0.136*** 0.065*** Oaxaca ‐3.405*** 0.752*** ‐0.048*** 0.135*** 0.034*** South Tehuantepec ‐4.617*** ‐0.702*** ‐0.086*** 0.156*** 0.117*** Chetumal ‐5.892*** 1.248*** ‐0.235*** 0.133*** 0.056*** Villahermosa ‐3.988*** ‐0.648*** ‐0.071*** 0.091*** 0.051*** Merida ‐0.039 0.744*** ‐0.082*** 0.083*** 0.046*** Constant 15.264*** 4.343*** 0.178** 0.284*** 0.022 Observations 412,017 663,508 4,475 4,475 4,475 R‐squared 0.149 0.646 0.340 0.238 0.043 Clustered standard errors by city (not shown in table). *** p<0.01, ** p<0.05, * p<0.1 Regressions include additional controls: Good FE, Outlet type FE for all regressions; Time FE for columns (1) and (2). Tables 11 to 14 show fixed effect results of pooled regressions of the indicators studied in this paper. All regressions include additional controls than the ones shown in the tables, in order to obtain FE results after controlling for other factors that if omitted, may generate noise to the estimations. 28 Table 12. Good fixed effects results (1) (2) (3) (4) (5) Dispersion Volatility PT magnitude PT speed PT symmetry Good FE (Rice excluded) Grain Beans 5.177*** 2.287*** 0.359*** ‐0.129*** 0.128 Eggs ‐ per piece ‐6.342*** 4.565*** 0.197** ‐0.180*** ‐0.070 Eggs & diary Eggs ‐ 12 pack ‐7.624*** 1.547*** ‐0.026 0.067 0.021 Eggs ‐ 18 pack ‐5.881*** 1.465*** ‐0.055 ‐0.020 0.339* Eggs ‐ 30 pack ‐6.128*** 3.262*** 0.079 ‐0.101* 0.082 Milk ‐5.400*** ‐2.132*** ‐‐ ‐‐ ‐‐ Bolillo bread ‐2.454 ‐1.855*** ‐‐ ‐‐ ‐‐ Bread & tortilla Concha bread ‐0.581 ‐1.072*** ‐‐ ‐‐ ‐‐ Corn tortillas ‐4.979*** ‐1.039*** 0.179 ‐0.242*** 0.038 Water ‐ 1lt bottle ‐3.842*** 0.415 ‐‐ ‐‐ ‐‐ Beverages Water ‐ 1.5lt bottle ‐2.847** ‐1.186*** ‐‐ ‐‐ ‐‐ Soda ‐ 2lt bottle ‐3.863*** ‐0.470 ‐‐ ‐‐ ‐‐ Soda ‐ 600ml bottle ‐6.574*** ‐1.371*** ‐‐ ‐‐ ‐‐ Beef ‐ chop ‐0.773 0.463 0.345*** ‐0.149*** 0.044 Beef ‐ ground 2.889** 1.322*** 0.541*** ‐0.185*** 0.096 Beef ‐ rib 1.151 1.066*** 0.483*** ‐0.138*** 0.131* Beef ‐ steak ‐2.475*** 0.692** 0.455*** ‐0.195*** 0.066 Meat Chicken ‐ whole ‐3.867*** 2.938*** 0.933*** 0.013 0.092 Chicken ‐ breast ‐1.539 2.504*** 0.988*** 0.071 0.125 Chicken ‐ roasted ‐0.852 ‐0.734** 0.627*** ‐0.046 0.074 Pork ‐ rib 2.093** 0.805** 0.163** ‐0.117*** 0.055 Pork ‐ steak ‐4.404*** 1.411*** 0.118 ‐0.107** 0.086 Apple ‐0.020 6.752*** 0.308*** 0.062 0.043 Banana 2.086** 10.257*** 1.022*** ‐0.060* 0.061 Guava 0.757 7.667*** ‐‐ ‐‐ ‐‐ Lime 0.774 16.107*** 0.271*** ‐0.225*** ‐0.004 Muskmelon 1.188 13.603*** 0.247*** ‐0.003 0.024 Fruits Orange 2.056** 12.744*** 0.163** ‐0.171*** ‐0.019 Papaya ‐0.220 10.321*** 0.387*** ‐0.049 ‐0.053 Pear ‐1.835** 6.472*** 0.163** ‐0.062 ‐0.006 Pineapple ‐0.893 9.226*** 0.028 ‐0.133*** 0.032 Plantain ‐2.169** 3.825*** 0.340*** 0.012 0.051 Watermelon 1.678* 13.346*** 0.158** 0.040 0.056 Avocado ‐0.101 9.546*** 0.745*** ‐0.122*** ‐0.042 Cabbage 4.590*** 8.601*** 0.127 ‐0.029 0.086 Carrot 1.044 8.809*** 0.209*** ‐0.041 0.031 Chayote 2.438** 14.167*** 0.401*** 0.021 ‐0.048 Cucumber 0.616 16.188*** 0.137* 0.020 ‐0.097 Lettuce 0.530 9.937*** ‐0.004 0.072** 0.094 Vegetables Onion 3.564*** 14.935*** 0.217*** ‐0.216*** 0.067 Jalapeno pepper 0.484 13.548*** 0.515*** 0.179*** ‐0.046 Poblano pepper 0.255 14.193*** 0.573*** ‐0.033 0.043 Serrano pepper 1.233 13.552*** 0.466*** 0.035 0.024 Potato 0.714 9.731*** ‐0.060 ‐0.084* 0.072 Tomatillo 0.867 14.418*** 0.553*** ‐0.018 ‐0.106* Tomato ‐ beefsteak 1.672* 20.217*** 0.290*** 0.084* 0.042 Tomato ‐ roma 3.369*** 19.924*** 0.212*** ‐0.043 ‐0.008 Zucchini 2.791*** 17.311*** ‐0.024 0.009 0.034 Salt ‐0.065 ‐2.239*** ‐‐ ‐‐ ‐‐ Sugar ‐ 1kg pack ‐3.706*** 0.646* 0.316*** ‐0.133*** 0.104 Other Sugar ‐ 2kg pack ‐2.990*** 0.202 0.368*** ‐0.186*** ‐0.056 Cooking oil ‐ Mixed ‐6.371*** ‐0.667** 0.062 ‐0.030 0.085 Cement ‐ 50kg bag ‐10.822*** ‐3.395*** ‐‐ ‐‐ ‐‐ Constant 15.264*** 4.343*** 0.178** 0.284*** 0.022 Observations 412,017 663,508 4,475 4,475 4,475 R‐squared 0.149 0.646 0.340 0.238 0.043 Clustered standard errors by city (not shown in table). *** p<0.01, ** p<0.05, * p<0.1 Regressions include additional controls: City FE, Outlet type FE for all regressions; Time FE for columns (1) and (2). 29 The results of city fixed effects regressions are provided in Table 11. The excluded city in this case is the capital. Some clear patterns emerge. In terms of price dispersion, prices in Mexico City exhibit more dispersion in general and lower dispersions tend to be concentrated in cities in central states. This finding is not surprising since it is likely that our definition of markets, in a large city like Mexico City includes many markets. As stated before, our empirical definition of markets was likely to be more aggregated than one using cross-price elasticities would have determined. Also, cities with more volatile prices tend to be located in the north. The magnitude of pass- through in Mexico City is generally higher than in most cities, but the speed of adjustment is lower. The characteristics of goods matter in understanding other aspects in the functioning of markets. Goods that are less perishable and easier to transport, such as beverages, cement, salt, sugar and eggs have the lowest dispersions after controlling for other factors. In contrast, fruits and vegetables exhibit similar or relatively higher dispersion than rice, the excluded term. As suggested by the heat maps, the most volatile prices are for fruits and vegetables, while prices for meat are the most sensitive to changes to international prices. Table 13 displays the results of the outlet FE regressions. Dispersion and volatility in supermarkets are in general higher than in other types of stores, but also in turn absorb shocks of international prices to a greater extent and at a more rapid pace. Table 14 shows semester FE results. This was done in order to identify how dispersion and volatility have evolved through time. The results indicate that dispersion is modestly increasing. Coefficients are relatively small but significant. Dispersion has increased by less than 1%. Volatility has been comparatively more stochastic across time; the coefficients suggest that volatility has decreased slightly by the end of the period. Table 13. Outlet fixed effect results (1) (2) (3) (4) (5) Dispersion Volatility PT magnitude PT speed PT symmetry Outlet FE (Supermarket excluded) Specialized store ‐1.911*** ‐4.834*** ‐0.096*** ‐0.071*** ‐0.074*** (0.452) (0.305) (0.026) (0.015) (0.018) Public market ‐3.041*** ‐5.240*** ‐0.102*** ‐0.062*** ‐0.071*** (0.437) (0.309) (0.015) (0.011) (0.015) Convenience store ‐1.671*** ‐4.080*** ‐0.064** ‐0.049*** ‐0.051** (0.458) (0.346) (0.024) (0.017) (0.023) Informal market ‐2.228 ‐4.452*** ‐0.134*** ‐0.012 ‐0.044* (1.403) (0.626) (0.024) (0.028) (0.022) Warehouse club ‐3.212*** ‐4.373*** ‐‐ ‐‐ ‐‐ (0.704) (0.612) ‐‐ ‐‐ ‐‐ Constant 15.264*** 4.343*** 0.178** 0.284*** 0.022 (0.793) (0.278) (0.069) (0.039) (0.062) Observations 412,017 663,508 4,475 4,475 4,475 R‐squared 0.149 0.646 0.340 0.238 0.043 Clustered standard errors by city in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Regressions include additional controls: City FE, Good FE for all regressions; Time FE for columns (1) and (2). 30 Table 14. Semester fixed effect results (1) (2) Dispersion Volatility Semester FE (First available semester excluded) 2010 ‐ 2nd semester 0.007 ‐‐ (0.100) ‐‐ 2011 ‐ 1st semester 0.492*** 0.393*** (0.113) (0.076) 2011 ‐ 2nd semester 0.487*** 0.174 (0.159) (0.146) 2012 ‐ 1st semester 0.235 ‐0.297** (0.147) (0.127) 2012 ‐ 2nd semester 0.449*** ‐0.694*** (0.152) (0.135) 2013 ‐ 1st semester 0.702*** ‐0.142 (0.166) (0.160) 2013 ‐ 2nd semester 0.606*** 0.231 (0.184) (0.141) 2014 ‐ 1st semester 0.787*** 0.716*** (0.206) (0.155) 2014 ‐ 2nd semester 0.574*** ‐0.413** (0.208) (0.166) 2015 ‐ 1st semester 0.841*** ‐0.095 (0.229) (0.186) 2015 ‐ 2nd semester 0.685*** ‐0.861*** (0.214) (0.197) Constant 15.042*** 5.394*** (0.823) (0.322) Observations 412,017 663,508 R‐squared 0.148 0.645 Clustered standard errors by city in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Regressions include additional controls: City FE, Good FE and Outlet type FE Market structure and competition regressions The tables that follow contain the results of the effects of the large retailer’s entries and COFECE sanctions on market efficiency. We include the results of split regressions by goods groups and geographical regions to have a clearer idea of how markets might be responding to such variables. Specifications related to the large retailer’s store entries. We use the cumulative entries of stores since the beginning of the period of study to measure the effect of the large retailer in Mexican markets. Equation (8) shows the specification, which also includes FE dummies. This means that the value of variable large retailer indicates the total of openings in city c until time period t. We have the possibility of running regressions using the large retailer’s third brand store entries only, the first brand stores entry only, or the effect of any of the large retailer’s brand stores entering in markets. In regressions where dependent variables are pass-through indicators, the variable large retailers refers to the total openings of stores by city during the period of study. This is because pass-through indicators are calculated for the complete period 2010-2015 and thus do not allow for time variation. 31 (8) Table 15. Effect of large retailer’s entries in markets Sugar, Salt & Eggs & milk Vegetables Cooking oil Beverages All goods Bread & tortilla Grains Meat Fruits Dependent variable: Price dispersion National 0.043 ‐0.073 ‐0.043 ‐0.039 0.170** 0.128** 0.038 0.039 0.190* Northern border 0.164 4.768 ‐0.487 0.717 ‐2.373** 0.528 ‐0.371 0.139 2.329** Northwest ‐0.569 ‐4.422 1.940** 4.273 1.501 ‐3.878** ‐1.246 0.011 ‐14.490** Northeast ‐0.132 ‐1.328 ‐0.128 ‐0.368 ‐0.533* ‐0.079 ‐0.024 ‐0.198 1.851*** Center north ‐0.014 ‐1.599** ‐0.178 0.371 ‐0.205 ‐0.069 ‐0.123 0.140 ‐0.289 Center south ‐0.397* 0.999 ‐0.306 0.622 1.892** ‐0.164 ‐0.923* ‐0.597* 0.085 South ‐0.264 ‐5.043*** 1.364 ‐0.451 0.425 ‐0.418 ‐1.019 0.559 2.055 Mexico City 0.014 ‐0.078 0.026 ‐0.059 0.215 0.080 0.072 ‐0.063 ‐0.131 Dependent variable: Price volatility National ‐0.051* 0.024 0.000 ‐0.047** ‐0.063** 0.008 ‐0.091* ‐0.071 ‐0.004 Northern border ‐0.319 ‐0.546 ‐0.662 ‐0.473* ‐0.841 ‐0.491*** ‐0.160 ‐0.210 ‐0.975* Northwest ‐2.031** ‐0.625 ‐0.723 0.448 ‐1.685** ‐1.106** ‐3.009** ‐2.870* ‐1.530 Northeast ‐0.098 0.092 ‐0.226** ‐0.036 ‐0.026 0.070 ‐0.204 ‐0.023 ‐0.173 Center north ‐0.026 0.034 0.039 0.046 0.078 0.091 ‐0.189 ‐0.074 0.156 Center south ‐0.277** 0.514* ‐0.092 0.255* ‐0.083 ‐0.237 ‐0.464* ‐0.430** ‐0.123 South ‐0.151 0.735 0.191 ‐0.641** ‐0.528 ‐0.471 ‐0.155 ‐0.217 0.895 Mexico City ‐0.067*** ‐0.128* 0.002 ‐0.125** 0.069 0.046* 0.033 ‐0.239*** 0.028 Dependent variable: Pass‐through magnitude National 0.331* ‐0.547 ‐0.171 ‐‐ ‐‐ 0.564*** 0.190 0.303 0.490 Northern border 2.818* 15.759* 1.086 ‐‐ ‐‐ 0.129 3.073* 3.464 14.813 Northwest 11.370* ‐12.538 12.221** ‐‐ ‐‐ 3.385 16.049** 11.989* ‐33.578*** Northeast 1.329 5.087 0.352 ‐‐ ‐‐ 2.100 ‐0.003 1.431 4.684** Center north ‐1.189 ‐9.934* 1.577 ‐‐ ‐‐ 0.780 ‐1.856*** ‐1.801 ‐1.666 Center south 1.832*** 1.348 3.934* ‐‐ ‐‐ 3.194*** 2.660*** 0.543 ‐2.210 South 3.143** ‐2.166 6.736 ‐‐ ‐‐ ‐1.848 6.648 2.594* 8.227 Mexico City* 0.144 ‐0.482 ‐0.160 ‐‐ ‐‐ 0.416 ‐0.324*** 0.283 ‐0.257 Dependent variable: Pass‐through speed of adjustment National ‐0.581*** ‐1.123** 0.747* ‐‐ ‐‐ ‐0.678*** ‐0.822*** ‐0.333** ‐0.784** Northern border 2.580** 10.352** 7.848** ‐‐ ‐‐ ‐0.142 4.703** 1.873 48.870*** Northwest 5.229 ‐5.744 ‐9.713 ‐‐ ‐‐ ‐0.276 6.255 3.627 60.723*** Northeast ‐0.163 0.628 1.395 ‐‐ ‐‐ 0.224 ‐0.936 ‐0.044 ‐0.453 Center north ‐0.253 ‐2.462 3.158* ‐‐ ‐‐ ‐0.847 ‐0.364 ‐0.258 3.805* Center south ‐1.514 ‐11.359** ‐6.173** ‐‐ ‐‐ ‐2.120 ‐1.407 ‐0.166 ‐7.616 South 1.576** ‐7.238 23.333 ‐‐ ‐‐ 1.141 0.617 3.507*** ‐6.780 Mexico City* ‐0.644*** ‐0.362 1.230*** ‐‐ ‐‐ ‐0.725* ‐0.876** ‐0.441*** 0.260 Dependent variable: Pass‐through magnitude symmetry National 0.207* 1.438 2.549* ‐‐ ‐‐ 0.547** 0.272 ‐0.244 0.590 Northern border ‐1.272 ‐7.712 4.088 ‐‐ ‐‐ 2.006 ‐0.681 ‐3.168** 12.448 Northwest ‐2.174 ‐11.639 ‐1.710 ‐‐ ‐‐ ‐13.141* 2.288 ‐3.527* 14.334 Northeast 0.965 8.974 5.041 ‐‐ ‐‐ 0.004 ‐0.257 1.117 4.579 Center north ‐1.550* ‐1.638 ‐3.616 ‐‐ ‐‐ ‐0.118 ‐1.171 ‐2.829* 0.493 Center south 0.670 11.599 7.292 ‐‐ ‐‐ 1.037 ‐0.150 ‐0.508 ‐10.060 South ‐0.477 19.598 ‐21.797 ‐‐ ‐‐ 3.711 ‐5.904*** 2.856 ‐5.049 Mexico City* 0.112 ‐0.253 3.126** ‐‐ ‐‐ 0.438 0.237 ‐0.206 0.000*** Clustered standard errors by city (not shown in table). *** p<0.01, ** p<0.05, * p<0.1 The table presents the results of split regressions by good groups and geographical regions. Each cell contains the coefficient β of the respective split regression according to Equation (8). β captures the effect of an additional Wal‐Mart store on the dependent variable after controlling for good, city, outlet and period FE. *In Pass‐Through regressions, the Mexico City region also includes the neighboring cities Toluca, Cuernavaca and Puebla, to allow for geographical variation. 32 Table 15 shows the results of large retailer’s entries in cities.12 In the pooled regression (i.e. in the “National” and “All goods” cell in the table), it seems that price dispersion is not affected by having additional stores in the markets of cities. Split regressions show that price dispersion is reduced after the large retailer’s stores are opened for some goods or regions. This is especially the case for the Center-South region and for grains in some areas. The large retailer’s apparent lack of effect in price dispersion might seem to be puzzling, but it is important to have in mind that these regressions estimate the effect of the large retailer in different markets within a city, that is, indicators at the triad-level of good-city-outlet. Thus, it is possible that a new large retailer store in a city might have dissimilar effects in its different markets: it might affect price dispersion in supermarkets, but perhaps not too much in convenience stores, for example. However, the large retailer appears to have a greater effect in lowering dispersion at city level markets (i.e. dispersion of prices irrespective of type of store).13 The effect of the large retailer’s entry on price volatility is more noticeable. On average, the large retailer’s stores reduce volatility, which means that prices tend to be more stable in markets within a city (as well as at city level). The effect is more general in northern cities and in the capital. Perishables are the most affected goods. As expected, the extent of price transmission (pass-through magnitude) increases when the large retailer enters a market. This effect is greater in the North. The speed of adjustment seems to be reduced, but this seems to be driven by stores in Mexico City. Specifications related to changes in market competition. We construct a dummy variable, COFECEgc, valued zero in time periods before a sanction and one after the sanction. Dates of sanctions vary by industry, (i.e. the type of good), which means that zero and one values differ by good. Equation (9) shows the specification used. Only goods affected by a COFECE sanction were included in regressions, thus, the interpretation of results should be seen as a before/after exercise. Pass-through regressions were not estimated for COFECE sanctions, due to the lack of time variation in pass-through indicators. (9) Table 16 shows the results of COFECE sanctions. In general, dispersion and volatility are reduced after COFECE fines businesses. This implies that COFECE sanctions are effective and indeed seem to be correcting competition problems in some extent. However, when analyzing regressions by good type, effects seem to vary, suggesting that the effect of fines may vary by industry. 12 We present the results of entries of Wal-Mart supercenters only (not other types such as Aurrerá or Superama), but results with complete Walmex stores or Aurrerá stores are very similar to Wal-Mart entries. 13 Regressions not shown in document. The specifications are exactly as regressions at good-city-outlet level, but for more aggregated indicators at good-city level only. 33 Table 16. Effect of COFECE sanctions affected by All goods Avocado sanctions Cement Poultry Tortilla Eggs Soda Dependent variable: Price dispersion National ‐0.560** ‐0.671 ‐0.617* ‐0.182 0.867** 0.770** ‐0.025 Northern border 0.000 0.571 0.235 1.410 ‐0.165 1.292 0.121 Northwest ‐2.067* ‐0.360 ‐0.753 1.374 1.001 1.011 ‐0.931*** Northeast ‐0.913 ‐0.807 0.685 ‐0.597 0.533 1.205 0.254 Center north ‐0.337 ‐0.809** ‐1.098** ‐0.101 0.953 0.026 0.012 Center south ‐1.108** ‐0.093 ‐0.795 ‐1.222 0.297 1.448 ‐1.910*** South 0.729 ‐3.058* ‐1.956*** ‐2.569 2.372 ‐0.466 0.000*** Mexico City ‐0.354 ‐0.725*** ‐1.284*** 1.330*** 1.840*** ‐0.123 1.189*** Dependent variable: Price volatility National ‐1.205*** ‐1.979*** ‐1.086*** ‐0.288 0.140 ‐1.600*** ‐0.498*** Northern border ‐0.814* ‐1.084** ‐0.768** ‐1.010 ‐0.225 ‐0.835* ‐0.536 Northwest ‐0.832 ‐0.261 ‐1.390 0.383 ‐0.471 ‐1.051 ‐0.778 Northeast ‐1.376*** ‐2.497*** ‐0.967*** ‐0.050 0.448 ‐1.802** ‐0.243 Center north ‐1.240*** ‐1.731*** ‐1.139*** ‐0.194 0.388 ‐1.563** ‐0.313 Center south ‐1.426*** ‐2.654*** ‐1.239*** ‐0.326 0.098 ‐2.359*** ‐0.598* South ‐1.184*** ‐1.920** ‐1.044*** ‐0.844 ‐0.019 ‐1.736* ‐0.729* Mexico City ‐0.586*** ‐1.675*** ‐1.150*** ‐0.687*** 0.751*** ‐1.469*** ‐0.919*** Clustered standard errors by city (not shown in table). *** p<0.01, ** p<0.05, * p<0.1 The table presents the results of split regressions by good groups and geographical regions. Each cell contains the coefficient β of the respective split regression according to Equation (9). β captures the effect of the COFECE sanction on the dependent variable after controlling for good, city, outlet and period FE. CONCLUSIONS Our goal in this descriptive study was to develop and explore a set of measures to examine the relative efficiency (functioning) of markets by looking at the behavior of prices for well- specified commodities. We examined the behavior of outlet-level retail prices for common commodities sold throughout Mexico in order to understand the functioning of markets. This examination of prices was done by looking at three main indicators; namely, price dispersion, price volatility, and price transmission (speed, completeness and symmetry). Once we designed the indicators, we provide patterns and conditions for each across commodities, regions and time. Depending on the indicator, we found clear regional and commodity- specific effects. In other words, some commodities were inherently more volatile than others in most regions of Mexico and some regions were inherently more volatile when we analyzed the behavior of a specific commodity. We then used fixed effects models to more precisely identify the extent to which commodity, location, outlet, and temporal variation explain the behavior of these price indicators. The results of these models confirmed what the descriptive statistics showed; namely, while there is heterogeneity across products, locations matter tremendously. So, the price conditions and trends of something quite common like maize can vary across locations and outlets. In sum, there is a need to better understand these patterns and we can conclude that Mexico is not one, well-integrated national market if we use these indicators to guide us. Finally, we tested whether we could predict changes in these indicators (increased efficiency) by looking at three changes expected to have a positive effect on the functioning of markets; changes in competition and the entry of the large retailer’s stores into the local retail market. 34 We showed that these changes did in fact affect the changes in the proposed market efficiency indicators in the way theory would predict. This supports our assertion that these indicators are likely to be good benchmarks to assess the relative efficiency (functioning) of markets. While validating the indicators developed to test whether they behave as theory would predict with the entry of the large retailer, another important finding should not be lost; namely, the large retailer’s entry into retail markets seems to lower prices and therefore likely increases economic welfare. While our findings are suggestive, and our goal was never to test this hypothesis with the kind of rigor that it merits, this indicative finding supports the work of Atkin et al (2016). For policy makers concerned about how changes in policies translate to prices paid by vulnerable groups or in poorer areas, the findings of this study provide support for efforts to monitor markets through the use of these indicators. For example, for policy makers worried about the distributional effects of liberalizing trade, they can usefully benefit from examining these micro-level price indicators to help them identify and understand the challenges certain markets may face, and to design, target and evaluate the effectiveness of efforts to help groups of people and households that transact in markets and locations most negatively affected by changes in policies. In addition, these indicators provide information about relative competition levels. Competition authorities can test and develop these indicators further and use them in their arsenal to monitor relative competition as a first, cursory assessment of competition. Doing this may save these authorities a lot of time and effort and possibly make them more effective. A yet to be published paper by Falcao Bergquist (2016) also suggests that competition among traders (or lack thereof) has similar traces with respect to price behavior in for maize in Kenya.14 In conclusion, the big message of this research is about the heterogeneity of how markets function, even for common and well-defined products. There are clear, non-random patterns to the behavior of these indicators. This is sobering because the singular focus of macro-level statistics on such things as price stability (a very worthy aim) may mask a lot of variation on the micro-level behavior of prices. 14 Falcao Bergquist, L (2016) Pass-through, competition and entry in agricultural markets: Experimental evidence from Kenya, Job Market Paper. 35 APPENDIX Table A1. Cointegration and unit root tests DFGLS Unit root test DFGLS Cointegration test Total of No trend Trend series p<0.01 p<0.05 p<0.10 p>0.10 p<0.01 p<0.05 p<0.10 p>0.10 p<0.01 p<0.05 p<0.10 p>0.10 Grains Beans 8% 16% 20% 56% 1% 6% 9% 84% 1% 4% 3% 92% 224 Rice 25% 13% 9% 53% 13% 12% 9% 66% 9% 14% 5% 72% 151 Eggs ‐ per piece 5% 27% 24% 44% 2% 5% 20% 73% 2% 7% 9% 81% 137 Eggs ‐ 12 pack 21% 17% 19% 43% 20% 12% 8% 60% 9% 21% 8% 62% 116 Eggs Eggs ‐ 18 pack 19% 20% 12% 50% 12% 10% 12% 66% 9% 14% 10% 66% 86 Eggs ‐ 30 pack 17% 11% 23% 49% 10% 6% 7% 77% 1% 16% 10% 73% 125 Tortilla Corn tortillas 1% 2% 3% 94% 0% 1% 0% 99% 0% 0% 2% 98% 176 Beef ‐ chop 28% 23% 20% 29% 2% 4% 2% 92% 8% 15% 8% 69% 168 Beef ‐ ground 30% 36% 16% 18% 1% 1% 1% 97% 10% 8% 6% 76% 186 Beef ‐ rib 31% 20% 8% 41% 1% 4% 1% 95% 7% 12% 10% 72% 169 Beef ‐ steak 33% 39% 13% 16% 2% 0% 3% 95% 4% 7% 9% 79% 223 Meat Chicken ‐ whole 59% 16% 9% 16% 9% 5% 6% 80% 23% 19% 12% 45% 187 Chicken ‐ breast 54% 9% 10% 26% 10% 6% 7% 77% 23% 14% 14% 49% 134 Chicken ‐ roasted 45% 18% 15% 22% 5% 5% 2% 87% 14% 13% 13% 60% 164 Pork ‐ rib 46% 16% 13% 25% 4% 4% 6% 87% 6% 12% 12% 70% 197 Pork ‐ steak 41% 30% 11% 18% 4% 4% 4% 88% 12% 19% 8% 61% 178 Apple 56% 16% 8% 20% 29% 5% 9% 57% 24% 11% 14% 51% 167 Banana 71% 10% 7% 12% 39% 25% 11% 25% 46% 28% 10% 16% 189 Lime 70% 10% 4% 16% 74% 5% 3% 18% 66% 12% 4% 18% 217 Muskmelon 79% 8% 3% 9% 67% 12% 8% 13% 68% 14% 5% 14% 185 Fruits Orange 46% 22% 13% 18% 28% 28% 14% 31% 55% 15% 9% 21% 228 Papaya 77% 11% 4% 8% 27% 20% 23% 30% 51% 28% 9% 12% 211 Pear 54% 13% 8% 25% 15% 16% 15% 53% 27% 28% 12% 32% 195 Pineapple 42% 22% 22% 15% 17% 23% 15% 45% 26% 32% 14% 28% 151 Plantain 36% 14% 10% 40% 20% 13% 3% 64% 20% 11% 14% 55% 152 Watermelon 55% 12% 9% 25% 34% 20% 12% 35% 61% 10% 8% 21% 234 Avocado 69% 13% 6% 12% 33% 33% 13% 21% 17% 28% 18% 37% 214 Cabbage 65% 9% 6% 20% 48% 16% 8% 28% 37% 20% 13% 29% 147 Carrot 65% 17% 9% 10% 30% 23% 18% 30% 22% 35% 15% 29% 210 Chayote 62% 17% 8% 13% 45% 23% 13% 18% 56% 18% 9% 17% 179 Cucumber 61% 14% 7% 18% 47% 14% 14% 25% 73% 6% 4% 17% 175 Lettuce 70% 10% 4% 16% 42% 14% 14% 30% 67% 11% 5% 17% 184 Vegetables Onion 55% 29% 7% 8% 37% 43% 9% 11% 7% 29% 22% 42% 211 Jalapeno pepper 77% 6% 6% 12% 57% 19% 5% 19% 73% 7% 8% 13% 159 Poblano pepper 84% 5% 4% 7% 73% 12% 4% 11% 59% 22% 5% 14% 207 Serrano pepper 77% 7% 7% 9% 57% 19% 8% 16% 76% 6% 3% 15% 209 Potato 14% 8% 18% 60% 11% 16% 24% 49% 6% 9% 9% 75% 216 Tomatillo 82% 5% 5% 8% 81% 5% 4% 10% 62% 23% 1% 13% 173 Tomato ‐ beefsteak 72% 11% 6% 11% 67% 8% 7% 18% 60% 15% 8% 17% 168 Tomato ‐ roma 85% 3% 3% 9% 79% 3% 5% 13% 74% 7% 6% 14% 246 Zucchini 80% 6% 4% 10% 75% 9% 4% 12% 76% 9% 2% 13% 216 Sugar ‐ 1kg pack 28% 18% 13% 40% 2% 2% 6% 89% 2% 9% 6% 83% 126 Other Sugar ‐ 2kg pack 31% 21% 13% 35% 2% 5% 2% 92% 3% 6% 4% 87% 131 Cooking oil ‐ Mixed 9% 18% 8% 65% 3% 5% 5% 88% 5% 4% 5% 86% 148 Note: The table summarizes the results of Dickey Fuller GLS tests performed for assessing cointegration and stationarity of the price series of commodities. Columns denote percentages of the total of series.The null hypothesis of the DFGLS test is that the series in question has a unit root. For the case of cointegration, we apply the DFGLS test to the residuals of long‐run regressions of the domestic price series with their respective international prices. 36 REFERENCES Abbott III, TA (1994) “Observed price dispersion: Product heterogeneity, regional markets, or local market power?” Journal of Economics and Business, 46(1): 21-37. DOI: 10.1016/0148-6195(94)90019-1. Aker, JC (2010) “Information from markets near and far: Mobile phones and agricultural markets in Niger.” American Economic Journal: Applied Economics, 2(3): 46-59. Ardeni, PG (1989) “Does the Law of One Price really hold for commodity prices?” American Journal of Agricultural Economics, 71(3): 661–669. DOI: 10.2307/1242021. Atkin, DG, Farber, B, and Gonzalez-Navarro, M (2015) "Retail Globalization and Household Welfare: Evidence from Mexico." CEPR Discussion Papers 10593, C.E.P.R. Discussion Papers. URL: https://www.dropbox.com/s/3gfnu6vhtm9rozt/retail_globalization_AFG.pdf Atkin, DG, and Donaldson, D (2014) “Who's getting globalized?: The size and nature of intranational trade costs.” mimeo. URL: https://dl.dropboxusercontent.com/u/2734209/Atkin_Donaldson_WGG_paper.pdf. Baron, DP (1970) “Price uncertainty, utility, and industry equilibrium in pure competition.” International Economic Review 11(3): 463-480. Article Stable URL:http://www.jstor.org/stable/2525325. Barrett, CB (1996) “On price risk and the inverse farm size–productivity relationship.” Journal of Development Economics, 51(2): 193-215. Basker, E (2005) " Selling a cheaper mousetrap: Wal-Mart's effect on retail prices." Journal of Urban Economics, 58(2): 203-229. Blomberg, SB, and Engel, RC (2012) “Lines in the sand: Price dispersion across Iraq’s Intranational Borders before, during, and after The Surge.” The Journal of Law and Economics,55(3):503-538.DOI: 10.1086/666586. Bresnahan TF, and Reiss, PC (1991) “Entry and competition in concentrated markets.” The Journal of Political Economy, 99(5):977-1009. Burdett, K, and Judd, KL (1983) “Equilibrium price dispersion.” Econometrica, 51(4):955- 969. Burstein A, Neves, JC, and Rebelo, ST (2003) “Distribution costs and real exchange rate dynamics during exchange-rate-based stabilizations.” Journal of Monetary Economics, 50(6):1189-1214. 37 Burstein, A, Eichenbaum, M, and Rebelo, ST (2007) “Modeling exchange rate passthrough after large devaluations.” Journal of Monetary Economics 54(2):346-368. Çağlayan, M, Filiztekin, A, and Rauh, MT (2008) “Inflation, price dispersion, and market structure.” European Economic Review, 52(7):1187-1208. DOI: 10.1016/j.euroecorev.2008.01.003. Campa, JM, and Goldberg, LS (2005) “Exchange rate pass-through into import prices.” Review of Economics and Statistics, 87(4):679-690. Campa JM, and Goldberg, LS (2006) “Distribution margins, imported inputs, and the sensitivity of the CPI to exchange rates.” NBER Working Paper Series 12121. Crucini MJ, Telmer, CI, and Zachariadis, M (2005) “Understanding European real exchange rates.” American Economic Review, 95 (3):724-738. Deschamps, R (1973) “Risk aversion and demand functions.” Econometrica, 41(3): 455-465. Article Stable URL:http://www.jstor.org/stable/1913370. Dixon, R (1983) “Industry structure and the speed of price adjustment.” Journal of Industrial Economics, 32(1): 25-37. Article Stable URL: http://www.jstor.org/stable/2097984 Domberger, S (1979) “Price adjustment and market structure.” Economic Journal, 89(353): 96-108. Article Stable URL:http://www.jstor.org/stable/2231409. Eckard, EW (2004) "The 'Law of One Price' in 1901.” Economic Inquiry, 42(1): 101-110. Eichenbaum, M, Jaimovich, N, and Rebelo, ST (2011) “Reference prices, costs, and nominal rigidities.” American Economic Review, 101(1): 234-262. Engel, C (1999) “Accounting for U.S. real exchange rate changes.” Journal of Political Economy, 107(3): 507-538. Engel, C, and Rogers, JH (1996) “How wide is the border?” American Economic Review 86(5):1112-1125. Engel, C, and Rogers, JH (2004) “Euro's price dispersion.” Economic Policy, 19(39): 347– 384. DOI: 10.1111/j.1468-0327.2004.00126.x Engle, RF, and Granger, CWJ (1987) “Co-integration and error correction: Representation, estimation, and testing.” Econometrica, 55(2): 251-276. DOI: 10.2307/1913236. Fackler, PL, and Goodwin, BK (2001) “Spatial price analysis.” In Handbook of Agricultural Economics, 971-1024, vol 1B, ed. B Gardner and G Rausser. Amsterdam: Elsevier. Fackler PL, and Tastan, H (2008) “Estimating the degree of market integration.” American Journal of Agricultural Economics, 90(1): 69-85. 38 Fama, E (1970) “Efficient capital markets: A review of theory and empirical work.” The Journal of Finance, 25(2): 383-417. -Stable URL: http://dx.doi.org/10.1111/j.1540- 6261.1970.tb00518 Finkelshtain, I, and Chalfant, J (1991) “Marketed surplus under risk: Do peasants agree with Sandmo?” American Journal of Agricultural Economics, 73(3): 557-567. Article Stable URL:http://www.jstor.org/stable/1242809. Finkelshtain, I, and Chalfant, J (1997) “Commodity price stabilization in a peasant economy.” American Journal of Agricultural Economics, 79(4): 1208-1217. Article Stable URL:http://www.jstor.org/stable/1244278. Fishman, C (2006) The Wal-Mart Effect : How the World's Most Powerful Company Really Works and How It's Transforming the American Economy, Penguin. Gagnon, E (2009) “Price Setting during Low and High Inflation: Evidence from Mexico” The Quarterly Journal of Economics, 124(3):1221-1263. Stable URL: http://www.jstor.org/stable/40506256 Gluschenko, K (2011) “Price convergence and market integration in Russia.” Regional Science and Urban Economics, 41(2):160-172. DOI: 10.1016/j.regsciurbeco.2010.12.003. Goldberg, LS, and Campa, JM (2010) “The sensitivity of the CPI to exchange rates: Distribution margins, imported inputs, and trade exposure.” Review of Economics and Statistics, 92(2): 392-407. Goldberg, PK, and Hellerstein, R (2008) “A structural approach to explaining incomplete exchange-rate pass-through and pricing-to-market.” American Economic Review, Goyal, A (2010) “Information, direct access to farmers, and rural market performance in Central India.” American Economic Journal: Applied Economics, 2(3): 22-45. Hanoch, G (1977) “Risk aversion and consumer preferences.” Econometrica, 45(2): 413-426. Article Stable URL:http://www.jstor.org/stable/1911218. Hellerstein, R (2008) “Who bears the cost of a change in the exchange rate? Pass-through accounting for the case of beer.” Journal of International Economics, 76(1): 14-32. Jackson, III W (1997) “Market structure and the speed of price adjustments: Evidence of non-monotonicity.” Review of Industrial Organization, 12(1): 37-57. Article Stable URL:http://dx.doi.org/10.1023/A%3A1007700703448. Jensen, R (2007) “The digital provide: Information (Technology), market performance, and welfare in the South Indian fisheries sector.” Quarterly Journal of Economics, 122(3): 879-924. Iacovone, L, Javorcik, BS, Keller, W and Tybout, R (2011) “Supplier responses to Wal- Mart's invasion of Mexico.” NBER Paper No. 17204 URL:http://dx.doi.org/10.3366/w17204. 39 Kehoe, PJ, and Midrigan, V (2007) “Sticky prices and sectoral real exchange rates.” Federal Reserve Bank of Minneapolis, Working Paper 656. Stable URL: http://www.minneapolisfed.org/research/wp/wp656.pdf Klenow, PJ, and Kryvtsov, O (2008) “State-dependent or time-dependent pricing: does it matter for recent US inflation?” Quarterly Journal of Economics,123 (3):863-904. Lee, I (2010) “Geographic price dispersion in retail markets: Evidence from micro-data.” Journal of Macroeconomics, 32(4): 1169-1177. DOI: 10.1016/j.jmacro.2010.06.007. McCann, F (2012) “Conflict, market (dis)integration and price dispersion: Evidence from Somali microdata.” mimeo URL: https://editorialexpress.com/cgi- bin/conference/download.cgi?db_name=CSAE2012&paper_id=637. Nakamura, E (2008) “Pass-through in retail and wholesale.” American Economic Review, 98(2): 430-37. DOI: 10.1257/aer.98.2.430. Nakamura, E, and Steinsson, J (2008) “Five facts about prices: A reevaluation of menu cost models." The Quarterly Journal of Economics, 123(4):1415-1464. Nakamura, E, and Zerom, D (2010) “Accounting for incomplete pass-through.” Review of Economic Studies, 77(3):1192-1230. Oi, WY, (1961) “The desirability of price instability under perfect compeition.” Econometrica, 29(1): 58-64. Stable URL: http://www.jstor.org/stable/1907687. Olken, B, and Barron, P (2009) “The simple economics of extortion: Evidence from trucking in Aceh.” Journal of Political Economy, 117(3): 417-452. Parker, C, Ramdas, K, and Savva, N (2013) “Is IT enough? Evidence from a natural experiment in India's agriculture markets.” Available at SSRN:http://dx.doi.org/10.2139/ssrn.2353771. Portugal-Pérez, A, and Wilson, JS (2008) “Trade costs in Africa: Barriers and opportunities for reform.” Policy Research Working Paper 4619, World Bank. Stable URL: http://elibrary.worldbank.org/doi/book/10.1596/1813-9450-4719. Pratt, JW (1964) “Risk aversion in the small and in the large.” Econometrica, 32(1/2):122- 136. Pratt, JW, Wise DA, and Zeckhauser, R (1979) “Price differences in almost competitive markets.” The Quarterly Journal of Economics, 93(2): 189-211. Rotemberg, JJ (1983) “Aggregate consequences of fixed costs of price adjustment.” American Economic Review, 73(3): 433-436. Salop SC, and Stiglitz, JE (1977) “Bargains and ripoffs: A model of monopolistically competitive price dispersion.” The Review of Economic Studies, 44(3):493-510. Article Stable URL:http://www.jstor.org/stable/2296903. 40 Salop, SC (1979) “Monopolistic competition with outside goods.” Bell Journal of Economics, 10(1): 141–56. Sandmo, A (1971) “On the theory of the competitive firm under price uncertainty.” American Economic Review, 61(1): 65-73. Article Stable URL:http://www.jstor.org/stable/1910541. Schmitz, A, Shalit, H, and Turnovsky, SJ (1980) “Producer welfare and the preference for price stability.” American Journal of Agricultural Economics, 63(1):157-160. Article Stable URL:http://www.jstor.org/stable/1239823. Shepard, A (1991) “Price discrimination and retail configuration.” Journal of Political Economy, 99(1): 30-53. Stahl, DO (1989) “Oligopolistic pricing with sequential consumer search.” American Economic Review, 79(4): 700-712. Stewart H, and Davis, DE (2005) “Price dispersion and accessibility: A case study of fast food.” Southern Economic Journal, 71(4): 784-799. Stigler, G (1961) “The economics of information.” Journal of Political Economy, 69(3): 213- 225. Turnovsky SJ, Shalit H, and Schmitz, A (1980) “Consumer’s surplus, price instability, and consumer welfare.” Econometrica, 48(2):135-152. Vaughan, D (2013) “An Analysis of the Process of Disinflationary Structural Change: The Case of Mexico.” Banco de México, Working Papers, No. 2013-12. Article Stable URL: http://www.banxico.org.mx/publicaciones-y- discursos/publicaciones/documentos-de-investigacion/banxico/%7BF48F72D2-6B27- 4AC6-1AB8-8E8B1A52B010%7D.pdf Varian, HR (1980) “ A model of sales.” American Economic Review, 70(4): 651-659. Ysusi, C (2010) “Consumer Price Behavior in Mexico Under Inflation Targeting: A Microdata Approach.” Banco de México, Working Papers, No. 2010-09. Article Stable URL: http://www.banxico.org.mx/publicaciones-y- discursos/publicaciones/documentos-de-investigacion/banxico/%7B07F3BEC2-B899- E776-52AC-3ED4E983236E%7D.pdf 41