Redistricting in Mexico: progress and Limitations

According to ^{Handley and Grofman (2008)}, redistricting is "the process by which lines on maps get drawn partitioning a territory into a set of discrete electoral constituencies from which one or more representatives are to be elected" (p. 3). This process is fundamental for democracy, because it outlines the physical space of political representation, where the bond between legislators and citizens is established.

The mixed-member electoral system of Mexico's legislatures combines two principles of representation: majority in single-member districts and proportional representation in plurality districts (^{Shugart and Wattenberg, 2001}). The lower house of Mexico's congress -the Cámara de Diputados or lower House of Congress- is composed of 300 districts of the first type, which elect three-fifths of the house, and five districts of the second -known as circunscripciones plurinominales- from where the other 200 members are elected. The same logic applies to the 32 local legislatures, whose numbers and proportions vary from one state to another (^{Balkin and Orta, 2004}). Our note addresses only the case of majority districts (^{Palacios Mora and Tirado Cervantes, 2009} provide an evaluation of the proportional representation districts).

In order to balance demographic changes among districts, Mexico renewed its congressional maps for the 1997 (replacing the map that had been in use since 1979) and 2006 federal elections (^{Trelles and Martínez 2012}; ^{Lujambio and Vives 2008}; ^{Magar et al., 2015}). The IFE (the Federal Electoral Institute, predecessor of the INE),² which drew up both maps, also completed a third redistricting process in 2013, on time and as planned, for the 2015 federal election. But amid doubts and tensions over an electoral reform that was being hammered out at the same time -and which would be adopted some months later- the EMB's executive board decided at the last minute to put off adopting the new district map.

With some technical differences, the 2013 redistricting effort followed the same process as the previous ones (^{Trelles et al., 2015}). The process, as it is officially supposed to be implemented, can be summarized in four phases (see Chart 1). It begins with the distribution of 300 districts among the states and Mexico City depending on the relative populations from the most recent census. Next, a Technical Committee appointed by the EMB's executive board -INE's Consejo General- develops and implements an optimization algorithm that automatically produces a preliminary map for each state. Each proposal is then submitted for two subsequent rounds of reviews by the parties. Acceptance of suggested modifications depend on the extent to which the modification improves the value of the plan, as calculated using the score function by which the maps are evaluated. The process ends when the Technical Committee selects a final proposal and submits it for the approval of the EMB's executive board.³

Source: Chart prepared by the authors.

Chart 1 The redistricting process in Mexico 

In the academic sphere, important contributions have been made to the study of Mexico's electoral geography. Authors like ^{Martínez Assad (1990)}, ^{Molinar (1990)} and ^{Emmerich (1993)} analyze how changes in the electoral geography have affected electoral results and the balance of power in the country. ^{Gómez Tagle and Valdés (2000)} and ^{Islas (2007)} point to the connections between socioeconomic traits in the population, their geographic distribution, and citizens' voting preferences. Along the same lines, ^{Sonnleitner (2013a)} stresses the social, collective and territorial characteristics of voting patterns in Mexico. ^{Sonnleitner (2001}, ²⁰¹³⁾ and ^{González (2008)} argue that despite regulatory changes in the last two decades, Mexico's indigenous population remains politically unrepresented, because of redistricting processes that are carried out with no regard to the diffe rences, and even antagonisms, among geographically proximate indigenous communities.⁴

This literature offers detailed and interesting descriptions of the territorial breakdown of the vote, but has not undertaken an in-depth discussion of issues such as the criteria used by the EMB for boundary delimitation (like ^{López, 2006} or ^{López and Soto, 2008}), the optimization models used by the electoral board (as in ^{Rincón García et al., 2015}) or parties' intervention with the redistricting authority (as in ^{Trelles et al., 2015}). There is fertile ground for ongoing analysis of the intersection between technical, regulatory and methodological aspects in redistricting, as well as the political impact that these various dimensions have on the generation of electoral maps. We believe that the availability of analyzable information is a prerequisite -ambitious, yet obtainable- for pursuing this line of research.

The systematic study of boundary delimitation criteria becomes of primordial importance in light of partisan gerrymandering. This means the practice, more or less surreptitious but widespread in the world, of designing districts to benefit one or more identifiable groups, such as parties, incumbents or racial groups (^{Cox and Kata, 2002}; ^{Jackman, 1994}; ^{Johnston, 2002}; ^{Magar et al., 2015}; ^{Otero, 2003}).⁵ Partisan participation in the redistricting process is well documented, but to evaluate the degree to which parties are able to affect electoral maps we need to know more about their incentives, their strategies, and the ways they interact with the emb. In studying redistricting in Northern Ireland, ^{Rossiter, Johnston and Pattie (1998)} argue that the neutrality of independent electoral commissions does not deter political parties from trying to influence their decisions to safeguard their own interests, and they present evidence that, to a significant extent, they succeed. For these authors, the degree of partisan influence in redistricting depends on three factors: a) the clarity of regulatory criteria; b) the profile of the members of the organization in charge of drawing the map; and c) the dynamics between that organization and political parties. The discussion contained in this note yields another two factors: d) the transparency and accessibility of all information on the redistricting process, and e) the existence of means of communication -and verification- between the public and the organization in charge of boundary delimitation.

Mexico is no exception when it comes to partisan influence in electoral regulation. In the same way to ^{Rossiter, Johnston and Pattie (1998)}, ^{Estévez, Magar and Rosas (2008)} argue that IFE's impartiality does not stem from the independence of its board members, but rather the system of party checks and balances within the institution. They describe the IFE as a collegiate body whose members are carefully selected with expectations that they will conduct themselves in accordance with the interests of the parties that appointed them in the Cámara de Diputados. In addition to this, a substantial portion of the EMB executive board's decisions -including redistricting- can only be audited by the parties. While it is fundamental that parties retain the ability to monitor this process, opening it broadly to society would improve representation, which is highly desirable. We elaborate on the concept of participation later on in this paper. But first, we show that the conditions are not yet in place for agents without direct presence in the board to track, evaluate or, much less, participate in the redistricting process.

Transparency, Accountability and Civic Participation

Admittedly, Mexico's EMB has made a tremendous effort and has achieved a great deal in working toward an open data policy during the (almost) twenty years of history of that body. A simple visit to the webpage reveals the substantial data now available.⁶ In terms of redistricting, although almost all the cartographic inputs are available online (^{INE, 2015b}; ^2015c) this occurs only on an ex post basis and once the process has concluded. Incredibly enough, because of the nature of redistricting, and the technical complexities it entails, the efforts made so far to open up the information and make this process more transparent have been insufficient.

Boundary delimitation requires a formidable volume of information. Among the most relevant are demographic figures of various kinds, the location of geographic features throughout the national territory, or the digitalized geography. It is very difficult, if not impossible, to evaluate a decision if we cannot replicate the process by which it was taken (^{Smith, 2015}). Table 1 lists seven major aspects we consider necessary and sufficient to replicate the redistricting process. It includes data bases, miscellaneous cartography, the mathematical formulas used, the existing laws and regulations, the party maps and alterations, the electoral results, and the specialized software.⁷ Thirty-six public agencies are responsible for producing and distributing it.

Table 1 Availability of information for redistricting. 

Source: Table prepared by authors. * In many states, the cartography is available only in photographic form and cannot be read by GIS software for manipulation. ** Authors' classification. High: information is not online and the institution has no means of public distribution; Medium: information is not online but the institution has a way to distribute it; and Low: the information is available online. *** Municipalities that cannot be fragmented because of their geographic nature, density or population type. **** Tool available in commercial software and in open-source software.

Some of the information is from sources outside the electoral board, such as INEGI's census data. The estimations of traveling time on the roads and highways within each district come from the Ministry of Communications and Transportation (SCT); the location and concentration of indigenous population is prepared by the cdi. Another substantial portion of the information comes from the INE itself, like the description of the optimization algorithms (see ^{Magar et al., 2015} for a critical discussion of the automated redistricting process), the applicable laws and regulations, or the cartography of the country's geographical blocks (secciones electorales, the basic redistricting units, equivalent to census tracts in the US). Finally, each one of the 32 local EMB's -known as Local Public Organizations (Organismos Públicos Locales, or OPLES after the 2014 electoral reform)- concentrates information on the various legal and administrative criteria applicable to that state. Although redistricting processes do not formally include the analysis of previous election results in any of its stages by the Technical Committee or the EMB's executive board, we assume that party modifications are made primarily on this basis, which implies that votes are key information for evaluating the maps proposed by the various parties.

As the table also indicates, some information is not available in accessible formats. There are some data that, despite being declassified by the board, are hard to come by.⁸ Further, there are various types of specialized software needed to process part of the data. For example, analyzing the data requires a software for manipulating geo-referenced data in GIS format. There are many options available, both commercial and open-source (Wikipedia lists a dozen under "geographic information systems"). Analyzing the data also requires a copy of the system developed by the board specifically for redistricting. This software performs various tasks: prepares proposed maps with a combinatorial optimization algorithm (known as simulated annealing or honeycomb optimization); allows for the selection of "seed secciones" as a starting point for the automated redistricting process; or calculates measures for evaluating different scenarios using quantitative indicators. Finally, if interaction with a third party is required (whether the electoral board, a political party or an external individual) there must be a platform to "socialize" the proposals and receive comments and changes.

Making the missing information available to the public will clearly help making redistricting processes more transparent. But given the nature of the redistricting process and its various phases, it is not enough in and of itself. The electoral board must also make a significant effort to harmonize and integrate the information so that it is easier to use. Below we describe three conditions for ensuring that transparency and open data translate into effective accountability (^{Smith, 2015}; ^{Ferreira da Cruza et al., 2015}).⁹ The conditions are ordered from least to most demanding, and can be understood as three stages: the higher the stage the redistricting guidelines reach, the greater the probability that transparency will translate into effective accountability.

Signs of Opacity

In this section we use three concrete examples to illustrate why it is important to meet the conditions of open data, replicability and participation. These examples point to some of the areas in which the existing open data policy is insufficient to ensure accountability.

Example 1: The Subnational Level

We have already discussed some of the problems and challenges of federal-level redistricting. But the shortcomings we detected pale in comparison to the local redistricting effort in most states. Before INE became responsible for redistricting local districts in the 32 states (with the 2014 electoral reform), most states failed even to meet the open data condition, the most basic of our list. In the majority of states, the task of drawing up local district boundaries fell exclusively to local electoral boards, although in some cases the legislative branch of the state played a central role in that process (^{Trelles and Martínez, 2007}; ^{López and Soto, 2008}; ^{Lujambio and Vives, 2008}). Today, it is still extremely difficult to obtain information about the adoption of local districts. Even obtaining the local electoral district maps is challenging. Photographic images of local districts are available in many states on the local EMB's websites, but the digitalized files -such as the shapefiles- that enable the maps to be analyzed on a basic level, are unavailable.¹¹ Without that information, it is virtually impossible to gauge the consistency and objectivity of the technical, regulatory and methodological criteria applied in each local redistricting process, or to document the extent to which the process has been politicized.

This was largely the reason why major parties agreed, as part of the 2014 electoral reform, that local redistricting -as well as electoral regulation in general- should be left in the hands of a single EMB at the national level. Because every state has its own legal framework and socio-demographic conditions, INE began a dialogue with the Local Public Organizations (OPLES), which are subdivisions of the board that replaced the state level EMB's, to reduce the tensions arising from the mismatch between local interests and the constitutional regulatory framework. The INE began work on redrawing electoral boundaries for the local congresses early in 2015.¹²

Another area of opacity that has not been addressed from a policy standpoint is the political effect of using different methods and technical criteria for redistricting at the federal and local level.¹³ In other words, in the past the EMB has not systematically used the same criteria and methods to redraw electoral boundaries at the federal level, and although it is now responsible for boundary delimitation at the local level as well, it has not used the same criteria for redrawing local legislative districts. For local redistricting the INE has decided to use a different optimization method than what it used in the federal process, and to take into account fewer criteria.¹⁴ For local processes, in contrast to the 2005 and 2013 federal redistricting processes, it chose to use not four but only two criteria in the optimization algorithm: population balance and geometric compactness (^{INE, 2015}; ^{Trelles et al., 2015}). The population component was assigned twice the weight of compactness in the cost function.

The INE has not justified these changes, either technically or in terms of the regulation, and given the lack of local-level information available, it is almost impossible to evaluate the political impact that the various decisions by the electoral board had on the electoral maps of each state. At the same time, the ambiguity in the selection of criteria and the lack of efforts to raise public awareness about the redistricting process pose tremendous challenges in terms of transparency. Why is it that the algorithms, components of the model, and weighting changed if the redistricting laws have not been modified in the past few years? How do these changes affect the configuration of local legislatures? Has any party (or parties) benefited systematically from these changes? Why are the criteria, their weighting, and methodology for redistricting local congresses different from those used in previous federal redistricting processes? What has been the political effect of the malapportionment induced by the differing criteria adopted in local boundary delimitation? One cannot derive satisfactory answers from the information that the EMB makes currently available.

Example 2: Partisan Influence

Since 1996 political parties have played an active role in formulating observations and counterproposals in the process of drawing up federal electoral districts. Because the condition of replicability has not been met, it is much harder to evaluate how much the interaction between the parties and the electoral board affects the configuration of electoral boundaries. In the most recent federal redistricting process that was held before the 2015 federal election (which was drawn up according to schedule but was never adopted), Leonardo Valdés -then President of the EMB's executive board- emphasized parties' active participation during the boundary delimitation process:

"The enormous participation by political party representatives with the National Surveillance Commission and Local Surveillance Commissions has already been accurately described: in effect, there are more than 540 observations presented by these representatives during the boundary delimitation process, and in many cases observations improved the proposed district plans according to the established rules. In all such cases, the Technical Committee accepted these recommendations and incorporated them into the scenarios, particularly in the Third Scenario which is today being considered by this General Council." (^{IFE, 2013b}).

In effect, the parties presented more than twice as many counterproposals as in the 1996 and 2005 redistricting processes; 236 were included to the first scenario, which was generated by an automated combinatorial optimization process, 157 of which were formulated in the Local Surveillance Commissions and 79 in the EMB's National Surveillance Commission. The counterproposals served as an input for generating the second scenario, and the parties formulated 308 new counterproposals, 139 of which were local and 79, national. On average, the seven parties represented in the IFE presented 75 counterproposals each (^{Trelles et al., 2015}).

What changes did the parties propose? What interests lay behind the counterproposals? How different were the proposals from the automated scenario? Which were accepted and which rejected? Were legal and regulatory criteria followed systematically? Which party proposed more scenarios, and what were the implications? Was the original value associated to the cost function improved? Did the board evaluate all counterproposals with the same criteria? What party submitted the most successful (accepted) proposals? And, in more general terms, did the parties' intervention introduce any bias in the districts? Unfortunately, with the information available to the public, no citizen, organization or stakeholder can answer these questions. Replicating and evaluating the different scenarios requires not only the IFE/INE software, but also access to the operationalization of the optimization algorithm (i.e. the software source code), the elements used to calibrate the model, the random selection of a seed -sección electoral- as a starting point for automated optimization, the various simulations that were conducted to determine the first scenario, and each of the 544 counterproposals submitted by the parties. Without a serious effort to meet the open data and replicability conditions, the proposed maps cannot be properly evaluated.

Example 3: Indigenous Representation and the Absence of Participation Mechanisms

Authors like ^{Sonnleitner (2013)} have pointed out the jarring contrast between the electoral board's initial effort to draw indigenous-majority districts, and the perennial lack of representation for Mexico's indigenous population in political life. Meeting the condition of participation would contribute tremendously to closing this gap.

The 2001 constitutional reform forced the electoral authorities to include indigenous population for the first time in the federal redistricting process facing the 2006 election.¹⁵ Although this was the first significant attempt to include this minority community, there are many aspects that can still be improved to strengthen political participation by these groups in public life. Grouping indigenous people together -as the board does- by census count during redistricting, without taking into account differences across indigenous communities or encouraging their participation, does not guarantee these groups have access to political representation or that they are able to build or strengthen the ties with their political representative.¹⁶

There has been little debate over how the representation of the indigenous population might be improved through redistricting, and we believe there are viable alternatives for addressing this need. A first step would be to recognize that including minority groups has various levels and options. As shown in Chart 2, a first level of participation is reached when the regulatory framework recognizes the need to safeguard the representation and participation of certain groups. In Mexico, this happened with the 2001 constitutional reform. The second level of participation is indirect inclusion. This happens, for example, when the board takes into account only the total number of inhabitants that were reported to speak an indigenous language, but does not take into account any other w-demographic factor or a direct consultative mechanism with the communities to obtain additional information. The third level refers to the degree in which the population belonging to a minority group is aware that its rights are safeguarded in order to guarantee their political representation and participation. The fourth level requires mechanisms for people to participate and communicate their needs to the electoral board. The fifth level is reached when the electoral board develops mechanisms for formally incorporating the opinions and needs of minority groups into the redistricting processes.

Source: Chart prepared by the authors.

Chart 2 Levels of participation in the redistricting process 

By this standard, Mexico is right now at the most basic levels of participation. First, the regulatory framework guarantees certain groups' right to representation; second, there is a mechanism for indirectly including the indigenous population.¹⁷But to move on from these initial levels of participation (indirect inclusion) to direct participation (the fourth and fifth levels), an intervention is required, which would include: a) education campaigns to ensure the people are informed and aware that their right to representation is safeguarded and that they can participate in these processes, and b) the mechanisms and tools needed for citizens to participate.¹⁸ In the following paragraphs we will describe how open-source software and web-based mapping technology offer a possible solution for making the redistricting process transparent and enabling the public to participate in that process.

The Use of Open Source Software and Public Mapping Platforms as a Possible Solution

New information technologies have brought the ability to meet the conditions of open data, replicability and participation within our grasp.¹⁹ We conclude this research note by introducing District Builder, a public mapping web based platform that has tremendous potential for overcoming the limitations and problems encountered in federal and local redistricting processes. District Builder offers a way to replicate any boundary delimitation exercise and a mechanism for citizens to participate openly in these processes.²⁰

The platform operates from the cloud, eliminating the need for users to acquire or install new software on their personal computers. District Builder can be hosted on the server of the electoral board, an academic institution, a non-governmental organization, or even by renting a space on one of the many available commercial services (i.e. Amazon Cloud). Users have one-click access to all the information involved in the redistricting process (which meets the open data condition), have a user-friendly tool to replicate, evaluate, compare and create electoral districts (which meets the replicability condition), and can formulate suggestions and observations to the electoral board (which meets the participation condition).

Users can create districts departing from a totally blank scenario, or they can visualize and edit existing federal and local maps preconfigured in the system. The public may also be given access to the electoral maps of previous years, to conduct comparisons or analyze the political effect that historical changes have had in the country's electoral cartography -the platform is highly versatile. This type of system also enables users to share information, store their proposals, and download all files in easily accessible formats that are readable by any digitalized geographic information system (GIS). Furthermore, the platform has the potential to become an extremely effective tool during INE's redistricting processes -both federal and local-, which involve many phases and participants, to reduce the existing communication gap between the public and electoral authorities, meeting the three conditions that guarantee accountability and making possible a transition to a public participation that is both direct and inclusive.

Figures 1 and 2 show a screenshot of the electoral cartography of the State of Mexico in District Builder. Both represent, on different scales, the portion of the state that borders on northern Mexico City. Figure 1 shows the borders of some federal districts with municipal subdivisions. Figure 2 shows the same district division, but with subdivisions at the level of sección electoral. The platform changes the map layers from municipalities to secciones automatically when zooming in with the mouse (similar to the way the Google Maps interface works).

Source: Figure prepared by authors. Note: The image shows the web interface for the platform, with a view of the 40 electoral districts of the State of Mexico and its sub-divisions at the municipal level (with the most densely populated municipalities shown in darker colors). At the top of the screen are the tools for editing and analyzing the map. On the right hand side there is a calculation tool with the values associated to the number of inhabitants, number of voters, and gender ratios in each district.

Figure 1 Sample screenshot from home page of District Builder. 

Source: Figure prepared by authors. Note: The image shows the web interface for the platform, with a zoom-in view of the 40 electoral districts of the State of Mexico and its sub-divisions at the sección electoral level (with the most densely populated of these shown in darker colors). At the top of the screen are the tools for editing and analyzing the map. On the right hand side there is a calculation tool with the values associated to population, contiguity, and compactness.

Figure 2 Screenshot of electoral cartography of the State of Mexico at the sección level in District Builder. 

In both images we can see, to the right of the screen, a calculator with the values associated to the map evaluation criteria that have been incorporated into the system. The platform allows for the incorporation of data on any variable at the sección level. In Figure 1, for example, the calculator reports total population, registered voters and the ratio of men to women in each district on the map. Figure 2 shows the population, with an indicator of contiguity and geometric compactness index for each district. All the socio-demographic information associated with the census results (number of inhabitants, income, education, gender, age); the values associated to the components used in the redistricting process (like percentage of indigenous population, geographic contiguity and continuity, geometric compactness and inter-municipal travel times); the previous electoral results (including any indicator of electoral competitiveness); or indicators generated by other institutions (crimes, kidnapping, extortion) can be viewed in this space.

To modify the map, the platform allows the user to select a redistricting unit or sección electoral (or group of them) and manually drag them from one district to another. The system automatically updates the new district boundaries, recalculates the values for the data the user is interested in, and displays them in the calculator. With this, any user can conduct geo-spatial analysis, modify maps and evaluate the effects of these changes without the need of a mapping specialization background. The platform also permits scenarios to be published and shared: it saves all changes made on the server and generates a link so the user can share the map through e-mail or social media. With this, various users can interact, work as a team, and formally submit their proposals to the electoral board in order to safeguard the interests of certain communities or groups. If, for example, a map that is being proposed by the electoral board is splitting a specific territory into two districts, or if it would place it in the same district with another rival -and more numerous- community, the platform would allow any user to communicate this situation to the emb, so it can make an informed decision before adopting the map.

The platform also allows users to select a group of variables and export them to a database for analysis with whatever tool the user prefers. The calculator that shows the figures has the potential to substantially increase the level of transparency surrounding the process, because users must undertake a dialogue (especially with the authority in charge of redistricting) objectively and based on quantitative values. District Builder makes it possible, among many other things, to measure the political effect of counterproposals submitted by political parties during the boundary delimitation process.

Finally, the platform allows users to check whether their district plan meets the legal criteria. The tool filters each plan, checks that it complies with restrictions (for example, the population deviation in each district is limited to certain percentage) and ranks the plan alongside district plans proposed by other users (including plans suggested by political parties and the first scenario produced by the automated process). This classification is public, based on criteria established a priori by the electoral board, and enables the redistricting effort to meet the conditions necessary for the fifth level of participation -direct and formal- described in Chart 2.

To sum up, this type of tool opens a window of opportunity for any citizen to take part in the redistricting process, communicate his or her needs and interests and, at the same time, gives the electoral board more information through crowdsourcing, as well as tools for evaluating and comparing counterproposals using objective, automated criteria. In other words, public mapping platforms are a possible solution for ensuring that transparency translates into effective accountability and for transitioning from the most basic to the highest levels of civic participation.

Conclusions

Although redistricting in Mexico has not been highly politicized in the past, it cannot be said to be a transparent process, and this lack of transparency generates unnecessary political tensions in both the short and long term. Looking ahead to the 2018 election, for example, it is not clear what optimization method, criteria or weighting the INE will use to redraw the boundaries of 300 federal districts. The districts that will be used in the 2018 election will, for the first time, be the battleground where elected representatives will try to build closer ties with the electorate in order to stay in office for up to three consecutive periods. In theory, these districts will serve to reelect legislators in subsequent elections, in 2021, 2024 and 2027.

The absence of legislative reelection in the past, and the centrality of political parties in the Mexican electoral system, explains -in part- the absence of conflict and politization around the redistricting processes to date. For many years, Mexican legislators have cultivated loyalties with the party leadership, but not with their constituents. In turn, Mexican voters, lacking a bond with their legislators, vote mainly along party lines, not to punish or reward the individual performance of their district representative (^{Dworak, 2003}; Godoy, 2007). Starting in 2018, however, legislators can be expected to try to build a more ambitious electoral connection with citizens in order to retain their seat for subsequent terms, and to show more interest in the changes that the electoral geography will go through approximately every decade.

The new normative framework allowing legislative reelection, as well as the potential political consequences of adopting different methodologies and criteria for redistricting, are fertile ground for new research. The framework can also be used as a key to make redistricting a more transparent process and offer new mechanisms that allow the public to be informed and participate in the process. This would help generate new paths of communication -so far non-existent- between citizens, their representatives, political parties and the authority.

The Mexican regulatory framework, from the Constitution to the regulations and bylaws of public institutions, already recognize access to public information as a basic right (INE, 2015a). In other words, there is a broad-based acceptance among the institutions of the Mexican State regarding the importance of having effective systems for accessing public information; and there is an explicit recognition of the importance of generating and making socially useful information available to citizens. The new online mapping technologies and open-source software offer to make this possibility a reality. For the first time, Mexico can meet international open-government standards and transition from a society in which information is public because the laws make it so, to one in which information is transparent, available, and usable by any interested party.