The Mexican Tourism Sector as a Driver of Shared Growth
Public Disclosure Authorized
Public Disclosure Authorized
Public Disclosure Authorized
Public Disclosure Authorized
70076
Nelly Aguilera César Velázquez Alejandro Montesinos
November 18, 2009 1
Executive Summary
It has been argued that tourism could help reduce poverty and promote shared growth more than other industries. However, there is little empirical evidence on this. The goal of this paper is to estimate (with the data available) the impact of tourism as a factor of shared growth at the local or municipal level in Mexico. This country provides an excellent research example due to its socioeconomic characteristics, the fact that tourism is an important economic activity and that, in spite of not having an ideal database to prove empirically that tourism is associated to shared growth, there are reliable databases that combined may provide insightful information. The first aspect of this study was to define the localities to be analyzed. We define tourist localities as those in which more than 20% of its working population is in sectors related to tourism using the Economic census of 2004. In some analysis, although not in the statistical ones, we include all “Pueblos Mágicos” (small typical towns that receive federal funding), and the city of Oaxaca. The analysis of the tourist destinations is done using several variables that try to capture shared growth and good socioeconomic conditions. These are: population without coverage of social security, workers earning less than two minimum wages, an index of marginality or development, and the Gini Index.1 Tourist destinations belong to different groups according to their type: beaches resorts, small beaches, Pueblos Mágicos, and other little towns. The main findings are the following. First, tourist destinations in general have better economic conditions that neighbor communities (page 17). Second, in general we find that growth in tourism (growth in tourism related employment) is associated to more employment, to lower percentage of population working in social security, and to better figures in the Human Development Index (HDI). It has also been found that growth in maquila and oil industries do not present these positive impacts. However, growth in employment is not reflected in a decrease of the percentage of population earning less than two minimum wages. This situation may be explained by the possibility that both, low skilled and skilled workers are being benefited by tourism. Growth in agriculture is only related to a slightly increase in total employment (page 26). 1
The Gini index is only estimated for a group of few cities due to data limitations. We use the “Index of marginality” of the CONAPO and the “HDI” of the UNDP.
2
Within tourist destinations, we observe that beaches resorts have better economic conditions than the rest of destinations. Archeological sites show the worst outcomes. In the middle there are small beaches, Pueblos Mágicos and other little towns (page 12). Fourth, we identified that indeed beaches resort are the destinations that are driven the outcomes mentioned in the previous lines (page 28, and in particular, we identified there are successful stories of tourism to promote employment and share growth, such as the case of Playa del Carmen and Tapalpa (page 21), although migration can be playing an important role in reshaping the socioeconomic conditions in massive tourism destinations that were built from scratch, such as Cancún and Playa del Carmen (pages 21-22).
3
The Mexican Tourism Sector as a Driver of Shared Growth Nelly Aguilera, César Velázquez, and Alejandro Montesinos** Index 1. Introduction .......................................................................................................................................... 5 2. The tourist sector in Mexico................................................................................................................. 7 2.1 International and domestic context..................................................................................................... 7 2.2 Tourism and the Mexican Economy................................................................................................... 8 3. Assessing the impact of tourism in shared growth ............................................................................. 12 3.1 Selection of tourist destinations........................................................................................................ 12 3.2 General performance of tourist destinations..................................................................................... 16 3.2.1 Cross sectional analysis ..........................................................................................................16 3.2.2 Performance over time ...........................................................................................................18 3.2.3 Two aspects to consider when assessing performance of tourism destinations .....................21 3.3 Performance of tourist destinations vis a vis the oil, maquila and agricultural communities .......... 23 3.3.1 Selection of comparable localities and descriptive statistics..................................................23 3.3.2 Statistical analysis ..................................................................................................................25 4. Conclusions ........................................................................................................................................ 28 5. References .......................................................................................................................................... 30 Appendix 1. Description of available databases..................................................................................... 32 Appendix 2. Social characteristics of tourist destinations...................................................................... 35 Appendix 3. Cross section comparison of tourist destinations vs. other communities .......................... 38 Appendix 4. Tourist destinations and change in the social variables..................................................... 51 Appendix 5. Migration and maturity of destinations.............................................................................. 62 Appendix 6. Regression results .............................................................................................................. 68 **
Nelly Aguilera is specialized in social and social security programs, health and labor markets. Currently, she is the Research Coordinator of the Inter American Conference of Social Security. César Velazquez if professor at Universidad Iberoamericana and has special interest in decentralization and municipal and state public finances. Alejandro Montesinos is a student of Toulouse School of Economics and during the development of this project was Research Assistant in the Coordination of the Inter American Conference of Social Security.
4
The Mexican Tourism Sector as a Driver of Shared Growth 1. Introduction There is empirical and theoretical literature that proves that tourism can generate economic growth, and studies that provide evidence in favor of the coined Tourism Led Growth Hypothesis (see for example Brida, Risso and Bonapace 2008 and Brida, Lanzilotta and Risso 2008). But the challenge still exists in order to prove whether economic growth through tourism can be translated into improved economic conditions for the majority of the population. In theory, tourism can help diminish poverty levels and promote shared growth more than other industries because: i) tourism may employ poor population since tourism is high intensive in labor; and entry level tourism jobs do not require high skills; ii) in order to attract tourism, governments may invest in infrastructure that could benefit the poor, and iii) there may be some spillovers effects or induced economic activity that can benefit the poor, for example the tourism industry has links to the informal sector that may provide opportunities for low skill workers (Siegel and Alwang 2005). However, there has been little direct evidence so far on the effect of tourism in stimulating shared growth and on how its impacts compare with those of other industries. The little evidence that exist is mainly based in case studies, that, while very rich to understand specific reasons that may explain the performance of a destination in particular, provide results with little statistical robustness and recommendations difficult to generalize (See for example Siegel and Alwang 2005). Thus, this study is an attempt to provide information that can help us close this information gap. In particular, the objective of this study is to analyze tourism as a factor of shared growth and to assess whether it has some positive spillovers. The findings of this study could be used to pursue further analysis in the topic and to justify public policies aimed at promoting shared growth through tourism. Here we are not analyzing other benefits tourism may have, such as providing an important amount of foreign currency, since this topic is not directly related to the distributional impacts we want to analyze. Neither we are analyzing the negative effects tourism may have.
5
We analyze the relation of tourism and shared growth by focusing on Mexico. This country provides an excellent research example for several reasons. First, it is a country with a large fraction of the population living in poverty conditions (according to the official measure of poverty, 18.7 millions of Mexicans or 18.2% of total population were in 2005 in “nutritional poverty”) and the income distribution is high unequal (the Gini index for the same year was around 0.5). Second, it is a country where tourism is important. Mexico is one of the top ten most visited countries in the world and the one with the highest number of international visitors in Latin America and where national tourism also plays a significant role. Third, Mexico has several destinations and of different types and tourist has different aspects, for example, beaches resorts and small beaches, archeological sites, on one side, and day trips, traveling tourism, vacations and retirement on the other. Fourth, while the total number of international visitors to Mexico has stabilized in the last years, which indeed has represented that Mexico lost market share, the behavior across the different destinations is heterogeneous, i.e., there are destinations that have grown significantly in last years, while others not, situation that allow us to identify an association between growth in tourism and variables that measure shared growth. Fifth, in spite of not having an ideal database to prove empirically that tourism is associated to shared growth, there are reliable databases that combined provide insightful information. The main findings are the following. First, tourist destinations in general have better economic conditions that neighbor communities (page 17). Second, in general we find that growth in tourism (growth in tourism related employment) is associated to more employment, to lower percentage of population working in social security, and to better figures in the Human Development Index (HDI). It has also been found that growth in maquila and oil industries do not present these positive impacts. However, growth in employment is not reflected in a decrease of the percentage of population earning less than two minimum wages. This situation may be explained by the possibility that both, low skilled and skilled workers are being benefited by tourism. Growth in agriculture is only related to a slightly increase in total employment (page 26). Within tourist destinations, we observe that beaches resorts have better economic conditions than the rest of destinations. Archeological sites show the worst outcomes. In the middle there are small beaches, Pueblos Mágicos and other little towns (page 12). Fourth, we identified that indeed beaches resort are the destinations that are driven the outcomes mentioned in the previous lines (page 28, and in particular, we identified there are successful stories of tourism to promote employment and share growth, such as the case of Playa del Carmen and Tapalpa (page 21), although migration can be 6
playing an important role in reshaping the socioeconomic conditions in massive tourism destinations that were built from scratch, such as Cancun and Playa del Carmen (pages 21-22). The organization of the document is as follows. The Section 2 shows the main highlights of the tourism industry in Mexico. Section 3 presents the analysis that identifies the impact of tourism in variables measuring shared growth and positive externalities: employment; poverty measured by share of population earning less than two minimum wages; share of population without entitlements to social security benefits; and human development index. Section 4 concludes. The annex presents the databases used in the study. 2. The tourist sector in Mexico 2.1 International and domestic context Mexico is an important country when talking about international visitors, measured by international arrivals. In 2008, Mexico had 22 million international arrivals, which represents approximately 2.7% of the world tourist travel. It leads Latin America in tourism. Nevertheless, in the last years there has been stagnation in the growth of visitors.2 As a consequence tourist market share is lower when compared to the 3.9% Mexico had in 1990 (IADB 2008). With respect to the demand, around 90% of Mexico’s international visitors come from the USA and Canada and more than 80% of visitors arrive in one of five airports: Mexico City, Cancun, Puerto Vallarta, Guadalajara, and Los Cabos. Europeans make up only around 5% of Mexico’s tourists, although it has been European investment in the Riviera Maya that may attract more tourists from that region. Mexico has also a huge domestic market that represents around 80% of tourist consumption. In 2005, 51 million national tourists registered hotel stays (Sectur 2008a). What is the importance of tourism in terms of contribution to GDP and employment is what we analyze in the next sub-section.
2
According to Sectur (2000), tourism in Mexico since the 1950s has had four stages: the initial period (periodo gestacional), the consolidation period, the culmination period and the transition period. The initial period (1950-1958) is characterized by simple sun, sand and surf destinations. Regarding national tourists, traditional sun, sand and surf destinations started their decadence (Veracruz, Tampico). The consolidation period (1958-1974) was immersed in the boom of the international tourism provoked by the world economic expansion. The ideal destination wanted to reflect the “American way of life” and produced large hotels with all the amenities. This is also a period of a huge expansion of the supply accompanied by a similar trend in the flights. The hotel chains grew a lot mainly because of easy access to credit by multilateral organizations as the World Bank and the IDB. This period also saw the take off of the “Centros Integralmente Planeados (CIPs). The CIPs has been the biggest tourism planning effort. The CIPs were controlled by Banco de Mexico (the Mexican Central Bank). In the culmination period (1974-1986), the study says the tourist model developed came to an end. The study does not clarify what is the transition period.
7
2.2 Tourism and the Mexican Economy Tourism sector’s contribution to GDP cannot be assessed directly as it does not constitute a sector of economic activity in the System of National Accounts (SNA). However, INEGI together with the Ministry of Tourism produced the Tourism’s Sector Satellite Account (TSA) as part of the SNA. The TSA was produced for the years 1993-2004 and 2004-2006 but data across accounts are not fully comparable, it is only yearly data and is national. For this reasons, when necessary in the document we will use Trade, Restaurants and Hotels as proxi for tourism. Taking gross added value as a proxy of GDP, it is possible to assess tourism sector performance. According to SECTUR (2008a) tourism in Mexico represented 8.2% of total GDP in 2006. Tourism in this sense represents a more important activity than mining, agriculture or construction which account for around 1.2%, 5.2% and 4.5% respectively of total GDP. We have mentioned that there has been a stagnation of international travels to Mexico in the last years. This is not yet reflected in the GDP shown in Table 1, but we can see that in the period 2003-2006, the tourism activity grew in Mexico, since annual rate of growth was higher than the same rate during the periods 1993-1998 and 1998-2003 (period that captures the recession of 2001 and the September 11 events). Although we do not have data after 2006, we can expect a worse performance of the tourism sector than the economy if the sector behaves across the business cycles as in the past (see Box 1). Table 1 Annual GDP Growth by Sector 1/ 1/ 1993-1998 1998-2003
2003-2006
2/
3.10 2.59 4.17 Total 2.27 0.47 3.76 Tourism n.a n.a 1.84 Agriculture n.a n.a 1.55 Mining 5.88 1.01 4.39 Manufacture n.a n.a 5.61 Construction n.a n.a 5.57 Electricity, gas and water 2.10 3.20 6.23 Trade, restaurants and hotels 6.18 6.09 4.99 Transport and communication n.a n.a 13.15 Financial services 1.24 1.02 2.76 Social and communal services Note: 1/ Data from Tourism Satellite Account 1993-2004, 2/ Data from Tourism Satellite Account 2004-2006. Data not fully comparable across accounts Source: Tourism Satellite Account 1993-2004, Secretaría de Turismo (2008a).
8
Box 1 Tourism and the global economic crisis The current situation of the global economy will have an adverse effect in the already lethargic Mexican tourism sector. This Box shows how the tourism sector behaves across business cycles in order to understand what we can expect in the following quarters. The following figure shows the quarterly cyclical component of the GDP and of the trade, restaurants and hotels sector since 1980 calculated using the Baxter-King filter. The next table shows the behavior of the GDP cyclical component by sector of economic activity. Several interesting features arise from the information in the figure and the table. First, over the 1990 – 2008 period there are three recessionary episodes: the 1995 financial crisis, the 2000-2003 crisis and the 2008- global crisis. Second, the trade, restaurants and hotels sector is highly correlated to total GDP and shows an even higher volatility: when an expansion occurs, the sector grows more than the GDP, when a contraction occurs it is deeper in the tourism sector. GDP Cyclical Component (% deviation from the trend, Baxter-King filter
Source: Own calculation using data from BIE/INEGI
Third, the trade, restaurants and hotels sector show the highest correlation to GDP among all sector that compose the GDP. Fourth, the correlation of GDP and trade, restaurants and hotels is simultaneous, i.e., there are not lead or lagged effects, even on quarterly basis. Behavior of GDP Cyclical Component by Division of Economic Activity (Baxter-King filter) Volatility Division of economic activity Absolute Relative
Correlation with total GDP x(t-4)
x(t-3)
x(t-2)
x(t-1)
x(t)
x(t+1)
x(t+2)
x(t+3)
x(t+4)
0.03
0.31
0.63
0.89
1.00
0.89
0.63
0.31
0.03
1.73
0.11
0.37
0.65
0.87
0.94
0.80
0.53
0.21
-0.04
2.77
-0.10
0.19
0.52
0.80
0.92
0.84
0.60
0.29
0.00
3.29
1.54
-0.12
0.16
0.48
0.75
0.89
0.84
0.64
0.38
0.16
2.92
1.36
-0.16
0.09
0.40
0.69
0.85
0.85
0.68
0.42
0.16
1.26
0.59
0.28
0.45
0.62
0.73
0.75
0.63
0.44
0.23
0.05
Mining Financial services Electricity, gas and water
2.21
1.03
-0.06
0.09
0.30
0.47
0.55
0.46
0.27
0.05
-0.12
1.07
0.50
0.34
0.48
0.59
0.61
0.53
0.32
0.07
-0.17
-0.36
1.97
0.92
0.22
0.33
0.41
0.43
0.37
0.22
0.04
-0.12
-0.22
Agriculture
2.01
0.94
-0.12
-0.06
0.05
0.19
0.30
0.34
0.30
0.19
0.04
Total Trade, restaurants and hotels
2.14
1.00
3.69
Construction
5.94
Manufacturing Transport, storage and communication Communal, social and personal services
Source: Own calculation using data from BIE/INEGI
9
This information helps us to understand why tourism has been highly affected during the crisis that started in September 2008, and that was aggravated by the AH1N1 episode in the country. On the positive side, if the sectors behave as in the past, the tourism sector will show a quick recovery after the economy starts growing.
Economic activities related to tourism generated on average 2.4 million jobs in 2006, representing 6.7% of total employment in the country. Out of this 6.7%, services account for 6%, and goods 0.7%. Within services, restaurants, bars, and night clubs account for 2.5. Employment in restaurants, bars and night clubs plus employment in accommodation represented 44% of the total jobs in tourism, which justify that in absence of information we use this two subsectors as a proxi for tourism in the analysis in the following sections (see Figure 1).
Figure 1 Tourism is above all a service industry (% share of the total employment)
8 6 4 2 Other services
Restaurants, bars and night clubs
Transport
Second Homes (imputation)
Accommodation
Services
Others
Handcrafts
Goods
Tourism
0
Source: SECTUR 2008a.
Finally, we can observe that labor productivity is higher in the tourist sector than the national average but lower than the productivity of manufacturing and transportation sectors industry. If labor productivity is high, at least in relative terms, we can expect that tourism can generate “high quality” employment, in the sense that it is a well-paid employment (see Figure 2).
10
Figure 2 Tourism is more productive in terms of value added than the national average, but well below manufacturing, 2004
Source: INEGI, Banco de Información Económica (BIE).
In sum, the Mexican tourism sector is important in international standards, measured by international arrivals, and it is an important sector regarding its contributions to the GDP. Given the fact that it is labor intensive and that it shows a high productivity, we can expect that growth in tourism be a factor that can contribute to more economic growth and employment (as assumed in the Plan Nacional de Desarrollo, see Box 2). Less is known however about the role of tourism to spur shared growth or to have other spillovers effects. This is what we analyze in the next section. Box 2 Tourism in the Plan Nacional de Desarrollo (PND) The PND 2007-2012 underlines the relevance of tourism in the economic growth of the country and states that the development of infrastructure and services must include those focused on increasing the capacities of the local population. The main goal of the PND is to make the country a leader in the sector through the diversification of markets, products and destinations; enhancing the competitiveness of Mexican tourism products to reach international standards. The PND relies on the following 6 strategies to achieve the latter: 1) making the sector a priority to attract new investments and to alleviate poverty; 2) increasing the competitiveness and the supply of the tourist opportunities without compromising the sustainability of the sector; 3) developing programs focused on the quality, satisfaction and security of the services provided; 4) promoting a better regulatory framework; 5) strengthening the actual markets and developing new ones; and 6) including the local populations in the process.
11
3. Assessing the impact of tourism in shared growth 3.1 Selection of tourist destinations The first step in the empirical analysis is the selection of the tourist destinations for study. In doing this we took two criteria. i) First we look at the importance of tourism in the municipality, measured as the percentage of persons working in tourism, specifically hotels and restaurants3, out of the total number of workers in the municipality based on the Economic Census 2004. The threshold was set at 20%. ii) Second, we included Pueblos Mágicos as posted in March 2009 in the SECTUR web page4. Pueblos Mágicos is a federal program that started in 2001 aimed at grating federal funding to some little towns in order to improve their tourist infrastructure. A Pueblo Mágico is characterized by its preserved colonial or typical architecture, which may or may not include major constructions such as cathedrals or convents. We made sure that we include different types of destinations in order to understand whether tourism impact varies with the type of destinations. In the graphical representation we included all Pueblos Mágicos, while in the statistical analysis we limited Pueblos Mágicos to those that comply with the 20% threshold. In the same line, in the individual graphical representations (mostly in the appendix) we also include Oaxaca city, a well known tourist destination that do not comply with the threshold of 20%. We classify the tourist destinations in five groups: major beach resorts, small beaches, archeological and colonial sites, Pueblos Mágicos and other little towns. Figure 3 shows the average data of the group of destination, while Figure 4 shows the location of the destinations in the map of Mexico. Table A2.1 in the Appendix 2 describes the individual data of the sample of the destinations.
3
Ideally, we would have liked to perform the selection based on the value added produced directly and indirectly by tourism. Nonetheless, in Mexico, gross municipal product is not available. We use the 2004 Economic Census since it is the last available. The definition of tourism used rules out the “trade sector” of the Economic Census because although “trade” is related to tourism, it is also related to other economic activities. If we had included “trade”, practically all the Mexican localities and municipalities would have been considered as tourist destinations. 4 Recently SECTUR updated the list of Pueblos Mágicos. We leave the analysis as it was in March 2009.
12
Figure 3 Main characteristics of different types of destinations
Figure 4 Tourist in Mexico clings to the coast and to lesser extend to typical towns and archeological sites
Source: own elaboration with data from INEGI
13
The sample has the following characteristics: Importance of tourism in the destination means, by definition, that 20% or more of registered jobs are tourism-related. We selected most – although not all - of destinations based on this criterion. In this respect, Pueblos Mágicos and Oaxaca deserve special mention. Most of Pueblos Mágicos do not meet the criteria. Indeed, only 7 Pueblos Mágicos do so. We have to take into account, though, that information from economic census – the data base used to classify destinations - is at the municipality level and that, as explained above, Pueblos Mágicos are small destination within municipalities. If we had had census data at the locality level probably many of Pueblos Mágicos would have met our criteria. Another reason why most of Pueblos Mágicos do not meet the criteria may be because the economic census does not collect data from street vendors, or informal businesses and the fact that informal and street vendor businesses make up a significant part of business activity in many of these towns, from street restaurants to crafts-sellers, both on the street and out of their homes. In the analysis performed we will separate Pueblos Mágicos from the rest of the destinations. Not shown in the graph, in the individual analysis performed (shown in the Appendix 2) we included Oaxaca, but it deserves special treatment. In Oaxaca only 12% of the workers have formal jobs in tourism. This is striking because Oaxaca is a well known and frequently-visited Colonial City very close to two very important archeological sites, Monte Alban and Mitla. Oaxaca is also the capital city of the state, indeed the only capital city considered in this review. As a capital city, Oaxaca hosts many different activities: education, government related and retail. In view of this, we have to examine the results from Oaxaca with caution. Total population. Our sample includes large, medium and small destinations. In general major beach resorts are large communities averaging 250,000 inhabitants. Huatulco is the exception, registering only 33,000 persons. The rest are in general small communities with Oaxaca (230,000), and the Pueblos Mágicos Bacalar (219,000), Todos los Santos (219,000) and San Cristobal de las Casas (166,000) being exceptions: small beaches averages 20,000 inhabitants, archeological sites 37,000, Pueblos Mágicos 66,000 and other little towns 42,000 inhabitants. The heterogeneity in the sample appears even within groups and could be very useful to identify different effects of tourism. Workers earning less than two minimum wages. As mentioned previously, given that we do not have reliable measures of income, we use percentage of workers earning less than two minimum wages as our proxi for poverty. The overall share of population earning less than two minimum wages is similar 14
to the share of population that suffers from assets poverty. Here we can see that destinations classified as major beach resorts register the lowest poverty indicators on average (39% earning less than two minimum wages), while destinations classified as archeological sites show the highest (81%). The rest, in order, from the lowest to the highest, are: small towns (44%), Pueblos Mágicos (53%), and small beaches (58%). This average disguises variability within groups, with the exception of the group of beaches resort. For example, the Pueblo Mágico Santiago in Nuevo León registers a percentage of 26% while Cuetzalan, another Pueblo Mágico registers 84%. Similar examples can be found in the rest of the groups, with the exception of the archeological sites, which all register a very high percentage of active population earning less than two minimum wages (See Table A2.1 in Appendix 2). Population without coverage of social security. Population without coverage of social security is also a proxi for well-being. For Mexico it has broadly proved that people entitled to social security benefits earn on average more than people without social security and approximately 95% of persons with social security live in urban communities, where in general there are better living conditions. It is not surprising that the correlation coefficient between percentage of population earning less than 2 minimum wages and population without coverage of social security is .72 Among our tourist destination types, the major beach resorts show the lowest percentage of persons without coverage of social security (40 at the locality level and 48 at the municipal level), and archeological sites show the highest percentage at 79% in the locality and 83% at the municipality level. In the middle but with a percentage of persons uncovered are Pueblos Mágicos, other little towns, other beaches and Oaxaca with 52% of uncovered rate. Marginality level. All beaches resorts – both major and small - show a low and very low levels of marginality with the exception of Huatulco which has a medium marginality level. All archeological sites show high levels of marginality. Marginality levels vary significantly among the other destinations. Seen at the locality level all destinations have low or very low levels of marginality (see Table A2.1 in Annex 2). Gini Index. We calculated Gini Index for based on employment surveys for Cancún, Acapulco and Oaxaca. The numbers are 0.37, 0.36 and 0.39. In sum, there are several interesting findings when analyzing tourist destinations: 1.
The sample is composed of destinations showing high heterogeneity in terms of type of
destination, size and economic conditions. 15
2.
In general, major beaches resorts show the best conditions, followed in several variables by
Oaxaca but we have to remember that Oaxaca is a capital big city that does not meet our criteria to qualify as tourist destination. Archeological sites show the worst conditions. 3.
Within groups there may be variation, with the exception of Archeological sites, which all show
similar conditions, i.e., all show very poor conditions. Understanding these differences could be very important to better understand tourism impacts. For example, within Pueblos Mágicos there are big communities such as Bacalar and Todos los Santos (with more than 200,000) and San Cristobal de las Casas (with more than 160,000) and very small ones such as Real de Catorce, Mazamitla y Real del Monte with around 10,000 inhabitants. Moreover, there are communities, like Mazamitla where 30% of the population works in tourism and Huamantla, a town that belongs to the metropolitan industrial area of Tlaxcala-Puebla, where only 4% of the population works in tourism. Indeed, the larger the Pueblo Mágico the lower the percentage of the population that works in tourism. We can observe within Pueblos Mágicos that tourism is not highly related to the outcome variables analyzed (percentage of population earning less than two minimum wages, percentage of population without coverage of social security, average wage and development conditions), and that indeed the size of the destinations is a more powerful predictor (see Table A2.2 in Annex 2). In the statistical analysis we will include only the Pueblos Mágicos that meet our 20% threshold. 3.2 General performance of tourist destinations Ideally we would like to know whether tourism improves the well being of the population. Although identifying causality in economics is a very difficult task, with the information at hand we can assess whether tourism is associated with better conditions for destinations’ populations. We will do so using two approaches. First we will compare cross section characteristics of tourism destinations to other geographical areas. Second, we will analyze time series data to understand how growth in tourism is associated with changes in other variables of interest. These two approaches will allow us also to group destinations according to their importance in terms of shared growth and external effects and to identify possible reasons for the differences among destinations. 3.2.1 Cross sectional analysis In this section we analyze how tourist destinations compare to other communities, to the state they belong and to the whole country. We do the analysis by each type of destination for our five variables of well being: population earning less than two minimum wages, population not covered by social 16
security, marginality index, average income of occupied population and Gini index (when possible). Figures in Appendix 3 show the comparisons. What the analysis tells us is: 1. Tourist destinations in general have a lower percentage of population earning less than two minimum wages than other communities. This is true because this measure is lower at the locality level than at the municipality and the state level. Remarkably enough, in this respect all of the major beaches resorts show a better performance level than the country as a whole. The rest of the destinations have a variable record, with the archeological sites showing a higher percentage of population under two minimum wages than national averages. 2. The pattern regarding population covered by social security show more or less the same pattern than our measure of poverty (people earning less than two minimum wages). This indicates that in tourist localities in general more people are covered by social security than at the municipality and state level. However the beaches resorts, major and small, take the lead and, in many cases, outweigh the other types of destinations where coverage is lower than at national level. 3. Tourist destinations in general show better well being conditions, measured by marginality indexes, than other close communities. As we can see, in almost all cases the marginality level of the tourist destination locality is lower than the marginality level of the municipality and the state. There are very few exceptions. 4. Data do not indicate a clear association between tourism and average income of occupied population. As can be seen, with the exception of major beach resorts, no clear pattern emerges to indicate that average income in the locality is higher than the average income at the municipal, state and national level. We have to take this information with care, in particular for Pueblos Mágicos since it is information collected prior to the start of the program. 5. Inequality in the tourist destinations is similar to inequality at the national urban level, although it is lower than the national one. The Lorenz curve and Gini Indexes of the tourist destinations in the sample closely reflect the national urban data and the situation in other cities. Although, the Gini Index at the national level measured by household data is slightly higher, 0.4731 (INEGI 2009).
17
3.2.2 Performance over time In this section we present information on the evolution of tourist destinations both in absolute terms and as compared to the state and the whole country. We included one additional variable in this analysis - growth of employed population – which should help us to understand any spillover effects on employment in the community. In order to present the data we have grouped destinations according their growth in employment related to tourism. We classify the destinations in quintiles: very good, good, medium, low and very low growth in employment in tourism (see Figure 5). This will suggest whether growth in tourism is a driver of shared growth (measured as population earning less than two minimum wages and HDI5) and whether growth in tourism has some spillover effects, measured by total employment in the destination. Figure 5 Growth rate in tourism employment at municipality level, 1999-2004
If the destinations that have grown also show improvements in the outcome variables, and/or if in the destinations that with negative growth also show worse outcome indicators, this may indicate, depending on the outcome variable analyzed, that tourism is a driver of shared growth and that has positive externalities. A formal statistical analysis is performed in the next section. Figures A4.1 contain the different figures that support these findings. The analysis indicates that: 1. Total municipal change in employment will be linked to an increase or decrease in tourism growth, effect that is more pronounced in destinations where tourism is very important (more than 25% of workers). Prime examples of this case are the three destinations that show the highest positive change in employment in tourism and the three destinations that show negative growth in tourism: Nuevo Vallarta, Calestúm, Playa de Carmen, Chichen Itzá, Uxmal and La Huerta.
5
We use in this case HDI developed by the Mexican charter of UNDP and not marginality index calculated by CONAPO because this last index can not be used for doing comparisons of data in different points of time.
18
2. In general there is not a clear pattern between growth in tourist destinations and rates of poverty reduction (percentage of population earning below two minimum wages), measured in comparison to their home state. As can be seen in Figure 6 within all five groups there are destinations that have reduced poverty more than their home state (negative numbers), but there are also destinations in which poverty has increased in relative terms to their home state (positive numbers)6. Nuevo Vallarta merits further discussion. It shows a slower reduction of the population earning less than two minimum wages than their home state, although we have to remember that indeed in these destination only 30% of the population earned less than two minimum wages in 2000 leaving little room for improvement.
Figure 6 Difference in change in percentage of population under two minimum wages between municipality and state, 20002005
3. With respect to population not covered by social security we can observe that there is not a clear pattern among groups and within groups. Sadly we observe that the national figures for the percentage of population not covered by social security did not change – for all practical purposes in the period of 2000-2005. Moreover, we observe that this is also the case in most of the tourist destinations. The tourist destinations registering a reduction in the percentage of population not covered by social security are exceptions. Not surprisingly, these destinations are Playa del Carmen, Cancún, Huatulco and Los Cabos. La Huerta is the only tourist destination that shows an important decrease in the percentage of population not covered by social security that belongs to a different group of destinations and indeed in terms of growth in employment in tourism belongs to
6
There is a possibility that when calculating the data INEGI extrapolates the data at the state level to data at the municipal level, but we could not confirm this.
19
the very low performance group. We should interpret this case with care since La Huerta is a very small community that can be affected seriously with any little change in work opportunities. 4. In general tourist destinations show a higher growth in the HDI than their state (and the country as a whole). Even in the cases that already had a high index for 2000 we observe a good performance (this is the case of Playa del Carmen). The following Figure supports this affirmation, where the destinations are ordered by rate of growth in tourism in the destination. Figure 7 Difference in change in HDI between municipality and state, 2000-2005
5.
There has been no significant change in inequality in the tourist destinations during the last four years. The same pattern holds at the national level. One possible reason is that four years is a short period of time to observe changes in inequality measures. Unfortunately the change in methodology in the employment surveys as explained above makes it impossible to calculate comparable Gini Indexes and Lorenz Curves. This is something that should be tackled in future research.
What have we learned from all this information? In our opinion two conclusions should be highlighted. First, tourism could be an economic activity that spurs shared growth and can have positive externalities when the tourist destination shows a positive and important growth, effect that is larger when tourism represents a significant percentage of employment in the community. If these two conditions hold tourism can have important marginal effect in reducing poverty and spurring employment. Unfortunately, with the data available we were unable to get evidence on the effect of tourism in the income distribution. For the case of Cancún, Acapulco and Oaxaca we could observe that the Lorenz curve did not change. Second, the effect of tourism in spurring shared growth and positive externalities varies across destinations. The understanding of the reasons that may explain the different effects should be a next step in the research agenda. 20
As an example, the following Figure shows the different destinations arranged according to the rate of growth of tourism and its effects (measured as the impact in generation of total employment in the municipality). The size of the dots measures the importance of tourism employment in the municipality. Figure 8 While most always positive, gains in total employment appear to be linked to increased tourism employment
3.2.3 Two aspects to consider when assessing performance of tourism destinations Before we proceed, it is important to discuss two issues that may be affecting the measures just presented. The first one is the case of internal migration. The second one relates to the maturity of the destination. A note on internal migration The socioeconomic indicators of a municipality or a state can be affected by migration. If immigrants in a region are less educated or have less income that the mean of it, it could be the case that the socioeconomic indicators are worse with respect the last estimation even if the government has implemented successful policies in terms of social development. In fact, such policies or the better socioeconomic conditions may be the cause of the inflow of people. In this sense, it is important for the analysis to figure out whether internal migration affects the results presented in this study.
21
The answer in general terms is no. Tables A.5.1 and A.5.2 in Appendix 5 show for 2000 the percentage of people earning less than 2 minimum wages and without social security for the entire population, for the people living in the same municipality in 1995 and for the people not living in the same municipality in 1995.7 From the data, we can observe than many of the municipalities analyzed benefited from migration, in particular for the variable of salary, in the sense that the indicators are better for the people not living in the municipality in 1995, but given the percentage of people in this condition with respect to total population, the indicators taken from all the population change almost insignificantly if we only estimated them with people living in the same municipality five years ago.8 For the cases destinations that have been “attraction places” like Los Cabos, Playa del Carmen and Puerto Vallarta, it could be the case that, at the same time, they have received people with higher and less human capital that the mean of the locality or municipality although at least for Playa del Carmen, it should be noted that immigrants have better socioeconomic conditions. Soloaga (2006) who estimated the education and income parts of the Human Development Index (HDI) of the UNDP for the Mexican states adjusting from migration, found that Quintana Roo for example, attracted people of higher income but also less educated people. He also found that migration only produced small changes in the index for almost all the states (less that 0.5%) and that the ranking of the states was almost the same. A note on maturity of destinations So far we have seen that many of the tourist destinations already show very good well being conditions. Moreover, we have seen that the impact of growth in tourism in inequality for example is nil. How can we reconcile both pieces of information? Tourism has or do not have impact on well being of the whole distribution? One possible answer to both questions is that the data analyzed (because it was the data available) already shows mature destinations, whose positive impact was observed in the past, or the data could not capture the impact of destination because they mature after the last observation recorded (especially important for Pueblos Mágicos). Table A.5.3 in Appendix 5 shows the year that indicates when the destination started as a tourism destination. It also includes the agencies that sponsor the 7
The tables and analysis were not done with information for 2005 because the INEGI 2005 Population Count does not have detailed data about migration. However, the main results should not be different since internal migration in the period 20002005 was smaller than in the period 1995-2000. 8 The results shown just considered the effect of people entering the municipalities but socioeconomic indicators are also affected by the people that left the destinations. This last fact cannot be analyzed because of a lack of data at the municipality level. However, Soloaga (2006) did the analysis at the state level and, as noted in the text, found similar results.
22
destination. As can be seen, most of the destinations included were developed well before the period we are analyzing. In the case of the Pueblos Mágicos, on the contrary, we have to wait some years to capture all the positive impact of the program. The only destination that is fully captured since its development is Playa del Carmen. In this sense, if Playa del Carmen is a valid representative of all destinations in the country, at least of the same type and size, we could argue that tourism indeed generated important shared growth and positive spillovers. 3.3 Performance of tourist destinations vis a vis the oil, maquila and agricultural communities We have shown that on average tourism destinations perform better in terms of “shared growth”: poverty and marginality reduction, than the nearby communities, than the states where each of them are located and the country as a whole. In terms of public policy, when localities compete for scarce resources it is important to understand these relationships. While tourism development does not rely on public sector outlays beyond standard infrastructure and access ways, such investments in communities with potential as tourist destinations can generate a significant pay back. In this section we assess how tourist destinations compare to oil, maquila, and agricultural sectors9. The importance of the sectors, measured by its contribution to GDP, total employment, and population (of those cities classified as maquila, oil, agriculture and tourism driven) is presented in Figure 910. Figure 9 Importance of different sectors
3.3.1 Selection of comparable localities and descriptive statistics The maquila locations where easy identified. The INEGI’s Banco de Información Económica (BIE) collects statistics of the Export Maquila Industry and indentifies the main cities of this sector. In our sample we kept those cities where workers in “maquila” are at least the 20% of total workers at the municipal level according to the economic census of 2004. Oil and agricultural cities chosen were those that had 20% of total working population working in oil and agriculture respectively. In the oil 9
Ideally we would have liked to include government activities, but the census data does not collect information on government. 10 We could not find reliable data on total agriculture employment, since most of it is informal.
23
industry we included extraction and processing. We have 6 oil cities, 12 maquila cities and 120 agricultural cities. Figure 10 shows the descriptive statistics of the destinations (the individual data for oil and maquila cities can be found in Table A.6.1 in Appendix 6). Oil cities are very heterogeneous. There are cities where oil employees account for more than half of total employment, and cities where this share is 21%. There are small and big cities, poor and rich cities (measured by percentage of population earning less than two minimum wages, marginality index and average wages). Maquila cities are in general cities where maquila – the assembly of imported parts into machinery and items for export - is significant to the local economy, are big cities, and not so poor: the percentage of the population earning less than two minimum wages is on average less than 30%, the population without coverage of social security reaches at most 47% and all cities have very low marginality indexes and an average wage around 3.8 times minimum wages. On the other hand, agricultural cities are small and poor: on average 67% of the population earns less than two minimum wages, 78% of the population has no social security coverage and the average income is 1.59 minimum wages. Data for the median income and the 25 percentile of individual communities (including tourism destinations) is presented in Figure 10. This corroborates the previous statements. Figure 10 Median income and 25 percentile income in communities in four sectors
Data from Figure 10 and Figure 11 allow us to see that in general Maquila cities have better economic conditions: higher HDI, higher average wage, and a lower percentage of the population earning less than two minimum wages and there is also a lower percentage of the population without coverage of social security. On the contrary, agriculture shows the worst outcomes. Tourism has lower indicators than oil cities, in general. But, how do these cities compare to tourist destinations in terms of inducing shared growth to benefit the poor and in term of having positive externalities, in terms of total 24
employment in the municipality, if the main industry in the city grows? This is what we try to answer in the following section. Figure 11 Description statistics of tourism and other sectors destinations
3.3.2 Statistical analysis In order to assess the impact of the specific industry, tourism, oil, maquila and agriculture in promoting shared growth and spillovers, we perform a series of regressions of the following form: ΔYi = α + ∑ β1 j ΔEDij + ∑ β 2 j D j + Xβ 3 j
j
Where €
measures change in the outcome variable in destination i measures change in employment in the respectively industry, tourism, oil, maquila and agriculture is a dummy variable that takes the value of 1 depending of the type of destination. The omitted variable is tourism. are control variables. The variables included are: marginality index in 1995, workers earning less than two minimum wages in 1995 and population without coverage of social security in 2000. 25
Table A.6.2 in Appendix 6 presents the results of the regression for each of the four outcome variables considering all the different industries: growth in total employed population, change in workers earning less than two minimum wages, change in population without coverage of social security and change in the marginality level. We perform OLS regressions11 in all cases. Figure 12 shows the value of the coefficients and whether they are statistically significant (blue) or not significant (red). In terms of the variable of interest, growth in employment in tourism is positively and statistically associated to growth in total employment in the municipality, to a decrease in the percentage of the population with social security coverage, and to an improvement in the HDI. It is worth noticing that this impact is inexistent for the other three sectors (only growth in employment in agriculture is associated to growth in total employment). Nevertheless, growth in employment in tourism is not associated to a lower share of the population earning less than two minimum wages. According to the coefficient and the significant levels, HDI is the only variable that is statistically significant: better socioeconomic conditions (measured by the HDI) is associated to a decrease in the percentage of the population earning less than two minimum wages and to a decrease in the percentage of the population without access to social security entitlements, nonetheless, is also associated to worse outcomes in the change in the same index, i.e., once the index is good in a community it is very difficult to improve it. The percentage of the population earning less than two minimum wages may also explain improvement in the percentage of the population earning less than two minimum wages, percentage of the population without access to social security entitlements, and HDI, but their effects are nil.
11
OLS: Ordinary least squares is a technique for estimating the association (parameter) between a dependent variable and independent variables in a linear regression model. This method minimizes the sum of squared distances between the observed responses in a set of data, and the fitted responses from the regression model.
26
Figure 12 Coefficients in the regression analysis of the external factors
In sum, regression results show that few variables explain changes in our outcome variables. If we ignore that some of this results may be explained because we used different time span in the calculation of different variables –after all we are merging data from different sources—we can conclude that there are no easy formulas to improve the well being of the communities. Nonetheless, one result should be taken with optimism: growth in tourism can increase growth in total employment, decrease the percentage of the population with out access to social security and improve the well being conditions of the persons living in the destinations (measured by the HDI). Now, there are some tourist destinations more effective than others in order to promote employment and shared growth? We will analyze this focusing our regressions only to tourist destinations and identifying the different types of tourism. The results can be found in Table A.6.3 in Appendix 6, the following Figure 13 posts the value of the coefficients and their significance, as in the previous figure.
27
Figure 13 Coefficients in the regression analysis of internal factors
As can be seen, growth in tourism in beaches resort is associated to higher growth in total employment, to an improvement in the change in the HDI and a decrease in the population with coverage of social security. Growth in tourism in small beaches is also associated to higher employment in the destinations. What this results tell us is that beaches resort is the destination that is driving the results of the impacts found in the previous regression. There are many possible factors explaining this circumstance: i) some of these destinations show a high growth; and ii) it is in beaches resort where more population is employed in the tourist sector, in relative terms. 4. Conclusions Reducing poverty and inequality in Mexico is a priority in the public agenda. Unfortunately we have seen that the country has not grown enough to attain this objective. Moreover, it is not clear the economic growth is translated in better economic conditions for the population. The World Bank is focusing its efforts to support those initiatives that not only generate growth but also benefit the majority of the population. Tourism could be an activity that can generate shared growth; after all it is a sector highly intensive in human capital. Nevertheless, there is little evidence on the effect of tourism in spurring shared growth and on how its impact compares vis a vis other important industries in the country. This study attempts to provide information that helps us to close this information gap. 28
The analysis provided in this study suggests that tourist destinations in general have better economic conditions than nearby communities. Taking into account the fact that most of the important destinations considered were developed well before the period analyzed in this study, one hypothesis that emerges is that indeed it was tourism that helped these communities to attain this level of development, as the example of Playa del Carmen -the destination for which we could capture the early development stages with the data available- evidences. Moreover, growth in tourism is associated with growth in total employment in the communities, lower percentage of the population with out social security entitlements, and better socioeconomic conditions, measured by the HDI. Nonetheless, growth in tourism is not associated to lower percentage of the population earning less than two minimum wages, our measure of poverty. This impact is absent in the oil, maquila and agricultural communities, although the maquila locations are richer than the average of the tourist destinations, measured by average wage. The study also helped us to identify different destinations in terms of the rate of growth of tourism and its impact to spur shared growth and positive externalities (total employment). We observe that indeed, the previous results can be driven by the performance of beaches resort. The understanding of why different destinations behave differently is a pending research agenda. The understanding of why communities differ in its growth and in its impact will allow us to identify public policies aimed at making this sector as dynamic as it used to be in the past, and more importantly, at identifying those initiatives that achieves better economic conditions for the population.
29
5. References Banco de México. Estadísticas. Available at: http://www.banxico.org.mx/tipo/estadisticas/index.html Brida, Juan Gabriel, Wiston Adrian Risso and Amarita Bonapace (2008) “The Contribution of Tourism to Economic Growth: An Empirical Analysis for the Case of Chile”. Mimeo School of Economics and Management, Free University of Bolzono. Available on line at http:ssrn.com/abstract=1298404 Brida, Juan Gabriel, Bibiana Lanzilotta and Wiston Adrian Risso (2008) “Turismo y Crecimiento Económico: el Caso de Uruguay”. Mimeo School of Economics and Management, Free University of Bolzono. Available on line at http:ssrn.com/abstract=1016074 CONAPO. Indicadores. Available at: http://www.conapo.gob.mx/ CONEVAL. (2007). Informe Ejecutivo de Pobreza en México 2007. México, D.F: Consejo Nacional de Evaluación de la Política de Desarrollo Social. Corona, R (1993). “Migración Permanente Interestatal e Internacional, 1950-1990”, Comercio Exterior, agosto: 750-762. Esquivel, Gerardo. (2009). “The Dynamics of Income Inequality in Mexico since NAFTA.” UNDP Research for Public Policy Inclusive Development Working Paper ID-02-2009. Inter-American Development Bank (IDB). “Turismo y Desarrollo en México”. Washington D.C., 2008. Available at: http://www.iadb.org.uy/res/pub_desc.cfm?pub_id=CSI-117&Language=Spanish Székely, Miguel. (2005). “Pobreza y Desigualdad en México entre 1950 y el 2004.” SEDESOL Serie Documentos de Investigación No.24. INEGI. Banco de Información Económica, BIE. Available at: http://dgcnesyp.inegi.gob.mx/ Secretaría de Educación Pública. ENLACE, Consulta de Base de Datos. Available at: http://enlace2008.sep.gob.mx/cons_bd.html Secretaria de Salud. Sistema Nacional de Información en Salud, SINAIS. Available at: http://sinais.salud.gob.mx/ Secretaría del Trabajo. Estadísticas del Sector. Available at: http://www.stps.gob.mx/DGIET/web/menu_infsector.htm Secretaria de Turismo (SECTUR). El empleo en el sector turístico de México. Mexico, 2002. Available at: http://datatur.sectur.gob.mx/pubyrep/emp/emp122002.pdf
30
Secretaria de Turismo (SECTUR). Estudio de gran visión del turismo en México: Perspectiva 2020. Mexico, 2000. Available at: www.sectur.gob.mx/work/sites/sectur/resources/LocalContent/14661/4/GranVision.pdf Secretaria de Turismo (SECTUR). Análisis del turismo. Mexico, 2008a. Available at: http://datatur.sectur.gob.mx/jsp/docs.jsp?trep=5 Secretaria de Turismo (SECTUR). Resultados de la actividad turística. Mexico, 2008b. Available at: http://datatur.sectur.gob.mx/jsp/docs.jsp?trep=5 Siegel, Paul B and Jeffrey Alwang (2005). “Public Investments in Tourism in Northeast Brazil: Does a Poor Area Strategy Benefit the Poor?” Latin America and Caribbean Region Sustainable Development Working Paper 22. The World Bank. Soloaga, I., G. Lara y F. Wendelspiess (2008).“Determinantes de la migración interestatal en México: 1995-2000 y 2000-2005. Mimeo, El Colegio de México. Soloaga, Isidro (2006). “Evaluación del impacto de la migración sobre el cálculo del índice de desarrollo Humano en México”. Mimeo, Universidad de las Américas, Puebla. World Economic Forum (WEF). The Travel & Tourism Competitiveness Report 2009. Switzerland, 2009. Available at: http://www.weforum.org/en/initiatives/gcp/TravelandTourismReport/index.htm
31
Appendix 1. Description of available databases In México there are different databases that can be used to analyze the performance of the tourist sector as a driver of shared growth. None of them is comprehensive and none provides all variables of interest. Still, with the use of all of them we can get a good idea of the importance of tourism to improve well being. This study uses four data bases at destinations levels: the economic census of 1999 and 2004; the population census of 2000 and the population counts of 1995 and 2005; the marginality indexes of CONAPO of 2000 and 2005 and several rounds of the national surveys of employment. The descriptions of each and, their respective advantages and limitations are shown in Table A.1.1.
Economic census (INEGI)
Population census (INEGI)
Population counts (INEGI)
Table A.1.1 Description of the Databases Used in the Assessment Description Variables Variables used Rounds included in the study available • Census to all • Employment • Employment • 1999 establishment • Some simple in the sector • 2004 s regardless indicators of of interest • (2009 round of their the P&L (tourism, taking place) compliance to (sales, agriculture, regulations COGS, maquila, and • Stratified by payroll, oil) industry expenditures) • Total • Des• Some simple employment aggregation at indicators of the the balance municipality sheet, assets level • Census to all • Income • Coverage of • 2000 households • Employment social security • Des• Coverage of aggregation at social the locality security level • Education • It is used to • Household calculate the and marginality community indexes infrastructure • Personal characteristic s: age, ethnicity • Interstate migration • Census to all households • Desaggregation at the locality level • It is used to calculate the
• Coverage of social security • Education • Household and community infrastructure
• Coverage of social security
• 1995 • 2005
Limitations • Data on value added not reliable • Only two observations • No data on street vendors
Data on reliable
income
Very limited analyzed
not
variables
32
marginality indexes
• Personal characteristic s: age, ethnicity • Interstate migration
Marginality indexes (CONAPO)
• Marginality indexes at the state, municipal and locality levels • Marginality indexes are composed of 9 indexes: % of illiterate population (15 years old o more): % of population without complete primary studies (15 years old o more): % occupants in dwellings without drainage; % occupants in dwellings without electric energy: % occupants in dwellings without running water; % of dwellings with some level of overcrowding; %occupants in dwellings with soil floor: % of population in localities with less than 5,000 inhabitants; % of occupied population with income until 2 minimum wages.
Marginality
Human Development Index – PNUD (HDI)
• Measured at the national, state and municipality level • Measures health (three variables), education (five variables and income (two
•
• •
Index
Marginality indexes Population Earning less than 2 minimum wages
Index
• 1995 • 2000 • 2005
Not comparable across time
• 2000 • 2005
33
National employment surveys (INEGI)
Human Development Index
variables) • Data from census, count and household surveyed • Quarterly survey on labor market variables • One year rotative panel • Includes 32 metropolitan areas • Change in methodology in 2005 • Index used to rank countries and municipalities by level of "human development", which usually also implies whether it is developed, developing or underdevelope d.
•
•
•
•
• •
Labor market participation per sector Labor market variables Sociodemographi c characteristi cs Index of population health and longevity Educational attainment Standard of living
GINI indexes and Lorenz curve for three tourist destinations (Cancún, Acapulco and Oaxaca)
• Life expectancy • Adult literacy rate • Primary, secondary and tertiary gross enrollment ratio • GDP per capita
ENEU 1990-2004 ENOE 2004-
• •
2000 2005
Out of the 32 metropolitan areas only three are tourist destinations according to our definition (see below) Information of data across surveys (ENEU vs. ENOE) is not comparable
•
•
Only focuses on three dimensions of capabilities. Not designed to assess progress in human development over a short-term period because two of its component indicators are not responsive to shortterm policy changes.
As can be seen, some of the information is available at the metropolitan area (which includes several municipalities), this is the case of the labor surveys. The economic census highest level of disaggregation is at the municipal level, while the population census and population counts are available at the locality level12 (several localities form a municipality). Tourism destinations can be metropolitan areas, like Cancún, or localities. In some cases when the destination is large, information about the municipality reflects the characteristics of the destination. On the other hand, in the case of the small destinations, like Pueblos Mágicos and small beaches, the information at the municipality level may be less accurate to reflect exactly what happens at the destination (locality) level. Summarizing, we will use different databases to assess the performance of tourist destinations. This approach will give us a good idea of what has happened in the recent years. Unfortunately, not all the variables that we wanted to analyze for all destinations are available. In particular, information from the employment survey is limited to Acapulco, Cancún and Oaxaca. The remaining 32 metropolitan areas surveyed are not considered tourist destinations. All databases are reliable in general as they are used to calculate the most important variables to follow the performance of Mexico. 12
According to population count, in Mexico there are 32 states, 2,454 municipalities and 187,983 localities.
34
Appendix 2. Social characteristics of tourist destinations Table A.2.1 Social characteristics of tourist destinations
State
% of workers in tourism at municipality level, 2004
Total population of destination, 2005
Workers earning less than two minimum wages, 2005
Population without coverage of social security, 2005
Marginali Average ty level, wage, 2005 2000
GINI Index , 2008
Human Develop ment Index, 2005
0.41
0.88
Beaches resort Cancún
Quintana Roo
25.64
572,973
26.01
35.81
Very low
3.92
Puerto Vallarta
33.96
220,368
24.09
42.47
Very low
3.86
0.88
Los Cabos
Jalisco Baja California Sur
32.21
164,162
15.74
35.66
Very low
4.01
0.88
Playa del Carmen
Quintana Roo
33.97
135,512
30.48
42.85
Very low
3.46
0.87
Nuevo Vallarta
Nayarit
37.18
83,739
34.53
53.76
Very low
3.19
0.85
Ixtapa
Guerrero
27.28
104,609
43.18
60.32
Low
2.99
0.83
Acapulco
Guerrero
20.59
717,766
59.81
54.27
Low
2.33
Huatulco
Oaxaca
38.52
33,194
50.63
65.61
Medium
2.71
0.80
Celestúm
Yucatán
25.2
1,626
71.5
83.39
Medium
1.53
0.79
La Huerta
Jalisco
30.42
20,161
40.62
70.23
Low
2.52
0.79
Puerto Escondido
Oaxaca
19.95
33,682
57.31
78.35
Medium
2.35
0.79
Tecolutla
Veracruz
44.41
24,258
73.32
89.2
High
1.2
0.77
Chichen Itzá
Yucatán
42.23
9,960
78.08
84.79
High
1.12
0.75
Uxmal
Yucatán
45.6
3,617
82.45
84.88
High
0.94
0.74
Palenque
Chiapas
23.68
97,991
81.51
80.1
High
1.29
0.73
Todos Santos
Baja California Sur
8.99
219,596
24.08
32.41
Very low
3.54
0.90
Santiago
Nuevo León
17.23
37,886
26.47
32.5
Very low
3.69
0.87
Tepotzotlán
México
4.23
67,724
36.39
50.88
Very low
3.01
0.85
Tepoztlan
Morelos
18.65
36,145
38.06
69.66
Low
2.51
0.85
Real del Monte
Hidalgo
22.03
11,944
58.85
60.21
Low
2.3
0.84
Bacalar
Quintana Roo
10.44
219,763
48.76
50.31
Low
2.62
0.83
Coatepec
Veracruz
7.02
79,787
56.72
58.04
Low
2.13
0.82
Parras
Coahuila
5.32
44,715
52.75
41.55
Low
2.54
0.82
Jerez
Zacatecas
10.36
52,594
53.99
70.68
Very low
2.32
0.82
Comala
Colima
19.6
19,495
58.42
71.46
Low
2.1
0.81
Mexcaltitán
Nayarit
9.86
84,314
63.66
68.55
Medium
2.06
0.81
Valle de Bravo San Cristóbal de las Casas
México
20.78
52,902
46.06
75.34
Low
2.39
0.80
Chiapas
12.76
166,460
64.81
71.43
Medium
2.31
0.80
0.39
0.82
Small beaches
Arqueological
Pueblos Mágicos
35
Taxco
Guerrero
7.11
98,854
57.21
77.82
Medium
2.26
0.79
Pátzcuaro
Michoacán
10.59
79,868
59.33
74.49
Low
2.14
0.79
Huamantla
Tlaxcala
3.76
77,076
72.1
78.94
Low
1.71
0.79
Tequila
Jalisco
11.21
38,534
39.82
63.57
Low
2.66
0.78
Mazamitla
Jalisco
30.41
11,671
40.23
80.41
Low
2.56
0.78
Bernal
Querétaro
8.09
34,729
60.92
83.2
Medium
2.16
0.78
Izamal Huasca de Ocampo
Yucatán
6.63
24,334
72.33
59.41
Medium
1.5
0.78
Hidalgo
28.65
15,201
72.17
90.99
High
1.55
0.78
Papantla San Miguel de Allende
Veracruz
8.47
152,863
64.48
80.65
High
1.61
0.77
Guanajuato
15.68
139,297
46.19
81.1
Medium
2.56
0.77
Alamos
Sonora
18.1
24,493
52.28
79.99
High
1.91
0.76
Cuitzeo
Michoacán
4.58
26,213
54.93
88.12
Medium
2.03
0.75
Dolores Hidalgo
Guanajuato
6.69
134,641
54.32
82.75
Medium
2.3
0.75
Real de Asientos
Aguascalientes
14.54
40,547
47.47
73.02
Medium
2.17
0.75
Real de Catorce
San Luis Potosí
29.37
9,159
75.36
64.2
High
1.21
0.75
Cosalá
Sinaloa
7.57
17,813
59.16
84.45
High
1.86
0.74
Tlalpujahua
Michoacán
5.99
25,373
73.97
88.4
High
1.62
0.73
Tapalpa
Jalisco
20.59
16,057
56.22
85.32
Medium
1.96
0.73
Cuetzalan
Puebla
21.09
45,781
84.46
91.73
High
0.83
0.70
Yautepec
Morelos
22.74
84,513
38.95
62.66
Very low
2.48
0.83
Cihuatlán
Jalisco
27.21
30,241
36.54
67.07
Low
2.88
0.80
Ixtapan de la Sal
México
27.53
30,073
52.36
73.63
Medium
2.2
0.77
Malinanco
México
26.36
22,970
59.89
89.25
Medium
1.53
0.73
Oaxaca
Oaxaca
12.18
265,006
39.85
51.95
Very low
3.15
Other little towns
0.44
0.88
36
Table A.2.2 Correlation coefficients for Pueblos Mágicos % of workers in tourism at municipality level, 2004 % of workers in tourism at municipality level, 2004 Total population of destination, 2005 Workers earning less than two minimum wages, 2005 Population without coverage of social security, 2005 Average wage, 2000 Human Development Index, 2005
Population without coverage of social security, 2005
Total population of destination, 2005
Workers earning less than two minimum wages, 2005
-0.37
0.03
0.17
-0.16
-0.12
-0.24
-0.30
0.36
0.35
0.58
-0.93
-0.65
-0.69
-0.81
Average wage, 2000
Human Development Index, 2005
0.77
37
Appendix 3. Cross section comparison of tourist destinations vs. other communities Figure A.3.1 Population Earning Less than Two Minimum Wages (%), 2000
Small beaches 100 90 80 70 60 50 40 30 20 10 0 La Huerta
Locality
Puerto Escondido Municipality
Tecolutla
State
Celestúm
NaGonal
38
39
40
Figure A.3.2 Population Not Covered by Social Security (%), 2005
41
Archeological 90 80 70 60 50 40 30 20 10 0 Palenque Locality
Oaxaca Municipality
Uxmal State
Chichen Itzá NaGonal
42
Pueblos mágicos 100 90 80 70 60 50 40 30 20 10 0
Locality
Municipality
State
NaGonal
43
Figure A.3.3 Marginality Index, 2005
Beach resort 5 4 3 2 1 0 Los Acapulco Ixtapa Puerto Nuevo Huatulco Cancún Playa del Cabos Vallarta Vallarta Carmen
Locality
Municipality
State
Small beaches 5 4 3 2 1 0 La Huerta
Puerto Escondido Locality
Tecolutla
Municipality
Celestúm
State
44
Archeological 5 4 3 2 1 0 Palenque
Oaxaca Locality
Uxmal
Municipality
Chichen Itzá
State
45
Notes: 1=Very low; 2=Low; 3=Medium; 4=High: 5=Very High
46
Figure A.3.4 Average Income of the Occupied Population (times the minimum wage), 2000
47
48
49
Figure A.3.5
Cumulative worker´s income
Lorenz Curve and Gini Index for Worker’s Income, 2008
1.0
0.8
0.5
0.3
0.0 0.0
0.3
0.5 Cumulative workers
0.8
Acapulco (0.36)
Cancún (0.37)
Chihuahua (0.39)
Ciudad Madero (0.46)
Oaxaca (0.39)
Tijuana (0.32)
1.0
50
Appendix 4. Tourist destinations and change in the social variables Figure A.4.1 Change in Total Employment (%), 1999-2004
51
52
53
Figure A.4.2 Change in Population Earning Less than Two Minimum Wages (percent points), 2000-2005
54
‐2
‐4
‐6
Municipality State Cosalá
Parras
Cuetzalan
Palenque
Comala
Coatepec
Ixtapan de la Sal
SanGago
Oaxaca
Todos Santos
Low
4
2
0
‐8
‐10
‐12
‐14 NaGonal
55
56
Figure A.4.3 Change in Population Not Covered by Social Security (percent points), 2000-2005
57
58
Figure A.4.4 Change in Human Development Index, 2000-2005
59
Figure A.4.5 Change in GINI Index for Worker’s Income 2005-2008
0.47 0.45 0.43 0.41 0.39 0.37 0.35 2005 Acapulco
2006 Cancún
2007 Oaxaca
2008 NaGonal, urban
60
Figure A.4.6 Change in Lorenz Curves for Worker’s Income 2006-2008
0.5 0.4 0.3 0.2 0.1
2006
Tijuana
Oaxaca
Ciudad Madero
Chihuahua
Cancún
Acapulco
National, urban
0.0
2008
61
Appendix 5. Migration and maturity of destinations Table A.5.1 Population Earning Less than 2 minimum Wages in 2000 by Residence in 1995 Occupied population earning less than 2 minimum wages Population not Population living living in the in the municipality municipality in in 1995 1995 Total population Beaches resort Los Cabos Acapulco Ixtapa Puerto Vallarta Nuevo Vallarta Huatulco Cancún Playa del Carmen Small beaches La Huerta Puerto Escondido Tecolutla Celestúm Archeological Palenque Oaxaca Uxmal Chichen Itzá Pueblos mágicos Real de Asientos Todos Santos Parras Comala San Cristóbal de las Casas San Miguel de Allende Dolores Hidalgo Taxco Huasca de Ocampo Real del Monte Mazamitla Tapalpa Tequila Tepotzotlán Valle de Bravo Cuitzeo Pátzcuaro Tlalpujahua Tepoztlán Mexcaltitán Santiago Cuetzalan
Population not living in the municipality in 1995 as a percentage of 5 years old or more population
23.11 62.08 44.82 28.92 37.59 52.68 29.19 34.21
24.67 62.56 45.68 29.72 39.96 55.90 29.88 38.26
19.91 52.20 39.01 24.82 26.55 35.47 27.48 30.40
28.13 4.05 10.77 15.11 16.36 11.11 25.06 45.25
48.76 59.62 86.84 77.55
49.68 61.26 87.39 80.12
39.89 42.24 76.36 60.61
9.35 6.63 4.90 9.82
79.67 41.46 89.44 84.70
80.57 41.02 89.93 85.10
64.02 45.54 66.67 75.00
4.43 9.52 1.01 2.62
62.31 35.36 56.20 65.29
62.56 35.18 56.86 65.99
55.32 36.55 36.67 53.89
3.17 9.04 2.75 5.36
63.34 50.34 59.19 59.39 77.05 62.83 48.29 67.48 47.80 44.20 55.95 57.44 62.04 77.35 55.68 69.31 33.20 88.86
64.09 51.16 60.02 59.89 77.39 62.96 49.09 68.46 48.42 44.97 56.37 57.82 62.50 78.53 60.27 69.86 33.24 89.44
53.65 35.14 40.53 44.62 68.15 57.63 32.89 42.56 38.74 38.54 47.28 43.26 51.42 56.28 33.53 59.22 31.49 57.61
6.54 3.69 2.97 3.16 4.03 2.92 3.93 4.18 4.72 11.53 3.36 1.84 3.98 3.18 15.00 4.76 4.92 1.53
62
Bernal 67.45 68.78 Bacalar 54.73 55.05 Real de Catorce 81.09 81.57 Cosalá 64.59 65.33 Alamos 67.84 68.50 Huamantla 73.33 74.18 Coatepec 67.17 67.47 Papantla 76.36 76.80 Izamal 78.46 78.98 Jerez 60.25 60.49 Other little towns Cihuatlán 43.86 44.71 Ixtapan de la Sal 63.61 64.03 Malinanco 72.75 73.66 Yautepec 56.97 57.85 Source: Own elaboration using INEGI’s 2000 Population Census
43.09 51.39 68.75 48.74 45.10 53.26 62.18 64.54 59.61 56.32
4.65 7.55 2.85 2.90 2.50 3.77 5.23 3.16 2.22 6.10
35.35 55.50 43.90 49.06
8.78 3.88 2.63 9.76
63
Table A.5.2 Population Not Covered by Social Security in 2000 by Residence in 1995 5 years old or more population not covered by Social Security Population not Population living in living in the the municipality in municipality in 1995 1995 Total population Beaches resort Los Cabos Acapulco Ixtapa Puerto Vallarta Nuevo Vallarta Huatulco Cancún Playa del Carmen Small beaches La Huerta Puerto Escondido Tecolutla Celestúm Archeological Palenque Oaxaca Uxmal Chichen Itzá Pueblos mágicos Real de Asientos Todos Santos Parras Comala San Cristóbal de las Casas San Miguel de Allende Dolores Hidalgo Taxco Huasca de Ocampo Real del Monte Mazamitla Tapalpa Tequila Tepotzotlán Valle de Bravo Cuitzeo Pátzcuaro Tlalpujahua Tepoztlán Mexcaltitán Santiago Cuetzalan
Population not living in the municipality in 1995 as a percentage of 5 years old or more population
41.04 58.32 60.04 46.11 58.87 66.27 45.04 58.27
41.25 58.37 60.43 45.71 59.25 68.76 44.88 63.16
40.53 57.33 56.73 48.61 56.85 48.79 45.42 52.45
28.13 4.05 10.77 15.11 16.36 11.11 25.06 45.25
75.00 76.53 90.31 81.76
74.36 77.65 90.79 82.40
80.88 61.57 81.18 77.44
9.35 6.63 4.90 9.82
84.81 50.63 75.68 82.08
85.09 50.12 75.68 82.39
79.32 55.09 80.65 69.72
4.43 9.52 1.01 2.62
67.50 35.72 37.24 71.33
67.43 35.45 37.21 71.29
69.04 38.39 36.01 71.35
3.17 9.04 2.75 5.36
73.05 81.30 83.32 76.44 87.44 52.08 77.91 82.32 62.39 44.65 75.60 88.58 72.40 81.73 68.78 61.27 31.10 94.60
73.52 81.85 83.68 76.40 87.52 51.92 77.89 82.51 62.48 44.78 76.06 88.72 72.51 82.01 73.02 60.61 30.37 94.89
65.93 68.48 70.93 76.90 84.87 57.86 78.10 77.80 60.99 43.39 62.74 81.28 69.73 71.67 45.02 74.59 45.23 76.44
6.54 3.69 2.97 3.16 4.03 2.92 3.93 4.18 4.72 11.53 3.36 1.84 3.98 3.18 15.00 4.76 4.92 1.53
64
Bernal 81.32 82.08 Bacalar 51.29 50.96 Real de Catorce 76.90 77.00 Cosalá 69.22 69.06 Alamos 76.76 77.00 Huamantla 79.76 80.34 Coatepec 60.34 60.21 Papantla 80.09 80.39 Izamal 58.19 58.19 Jerez 69.36 69.23 Other little towns Cihuatlán 69.85 70.09 Ixtapan de la Sal 77.17 77.39 Malinanco 89.02 89.37 Yautepec 63.00 62.83 Source: Own elaboration using INEGI’s 2000 Population Census
65.95 55.45 72.51 72.52 70.14 66.51 62.53 70.29 59.26 72.24
4.65 7.55 2.85 2.90 2.50 3.77 5.23 3.16 2.22 6.10
67.63 72.94 77.24 64.37
8.78 3.88 2.63 9.76
65
Table A.5.3 Year of development of the tourist destination Tourist destination Los Cabos Acapulco Ixtapa Puerto Vallarta Huatulco Cancún Playa del Carmen Puerto escondido Chichen Itza Oaxaca Ixtapan de la sal
Type Playa resort Playa resort Playa resort Playa resort Playa resort Playa resort Playa resort Otras playas Zonas arqueológicas Zonas arqueológicas Otros pueblos pequeños Real de Asientos Pueblo Mágico Pueblo Mágico Todos Santos Pueblo Mágico Parras Pueblo Mágico Comala San Cristóbal de Pueblo Mágico
las Casas San Miguel de Allende Dolores Hidalgo Taxco Huasca de Ocampo Real del Monte Mazamitla Tapalpa Tequila Tepotzotlán Valle de Bravo Cuitzeo Pátzcuaro Tlalpujahua Tepoztlán Mexcaltitán Santiago Cuetzalan Bernal Bacalar Real de Catorce Cosalá Alamos Huamantla Coatepec Papantla Izamal Jerez
Year 1976 1950 1974 1970 1985 1974 1999 1982 1988 1987 1996
2006 2006 2004 2002
Sponsor FONATUR FONATUR FONATUR FONATUR FONATUR Gob. Federal y estatal UNESCO UNESCO ASPI, Gob. Estatal y Sectur SECTUR SECTUR SECTUR SECTUR
2003 SECTUR Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico Pueblo Mágico
2002 SECTUR 2002 SECTUR 2002 SECTUR 2001 2004 2005 2002 2003 2002 2005 2006 2002 2005 2002 2001 2006 2002 2005 2006 2001 2005 2005 2007 2006 2006 2002 2007
SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR SECTUR
66
Sources: http://www.fonatur.gob.mx/es/Des_loscabos/des-loscabos.asp, http://es.wikipedia.org/wiki/Acapulco, http://www.fonatur.gob.mx/es/Des_ixtapa/des-ixtapa.asp, http://es.wikipedia.org/wiki/Puerto_Vallarta, http://www.fonatur.gob.mx/es/Des_huatulco/des-huatulco.asp, http://www.fonatur.gob.mx/es/Des_Cancún/desCancún.asp, http://es.wikipedia.org/wiki/Riviera_Maya, http://www.laregion.com.mx/oaxaca/especiales/diversion/playas/pto_esc/pto_hist.php, http://es.wikipedia.org/wiki/Oaxaca_de_Ju%C3%A1rez, http://es.wikipedia.org/wiki/Oaxaca_de_Ju%C3%A1rez, http://www.elocal.gob.mx/work/templates/enciclo/mexico/mpios/15040a.htm, SECTUR webpage, consulted on February 2009
67
Appendix 6. Regression results Table A.6.1 Descriptive statistics of the variables in oil, maquila and agriculture
State Oil Reforma Agua Dulce Atitalaquia
Chiapas Veracruz Hidalgo
Total % of workers population in industry at of municipality destination, level, 2004 2005
Workers earning less than two minimum wages, 2005
Population without coverage of social security, 2005
Marginality level, 2005
Average wage, 2000
26.26 55.95 52.49 39.42
113,902 34,896 44,322 24,749
46.55 56.25 49.13 52.28
56.01 71.01 51.45 50.33
-0.940 Low Low Very low
2.89 2.78 2.82 2.87
73,746 78,519 199,988 525,721 126,238
18.21 48.07 40.86 23.74 28.69
32.70 67.94 50.24 35.20 25.57
Very low Low Low -1.745 Very low
4.39 2.67 3.31 3.76 3.44
Cadereyta Jiménez Paraíso Carmen Maquila
Nuevo León Tabasco Campeche
Acuña
Coahuila
27.39 26.60 21.99 47.01 70.08
Nogales Juárez Tecate Reynosa
Sonora Chihuahua Baja California Tamaulipas
57.08 56.99 56.90 56.70
193,517 1,313,338 91,034 526,888
21.55 28.24 16.99 35.39
29.66 31.47 40.24 34.50
Very low Very low Very low Very low
3.80 3.50 3.90 3.31
Tijuana Matamoros Agua Prieta Mexicali Piedras Negras
Baja California Tamaulipas Sonora Baja California Coahuila
50.21 47.72 47.04 34.44 33.76
1,410,687 462,157 70,303 855,962 143,915
11.70 30.02 31.06 14.19 21.96
40.99 40.85 47.67 34.22 30.74
Very low Very low Very low Very low Very low
4.40 3.32 3.37 4.21 4.04
Nuevo Laredo Chihuahua Agricultural
Tamaulipas Chihuahua
30.13 23.05 39.58
355,827 758,791 23,734
29.94 15.11 67.18
40.33 26.22 78.43
Very low Very low 0.102
3.56 4.23 1.59
68
Table A.6.2 Regression results for external factors (all sectors included)
OLS
OLS
OLS
0.428***
-0.022
-0.020*
0.013**
(0.070)
(0.017)
(0.011)
(0.005)
0.049**
-0.006
-0.000
-0.000
(0.019)
(0.005)
(0.003)
(0.001)
0.950
-0.084
-0.047
-0.030
(1.10)
(0.281)
(0.184)
(0.087)
0.354
-0.001
-0.009
0.028
(0.385)
(0.098)
(0.064)
(0.030)
10.172
4.722**
0.110
1.416**
(7.065)
(1.808)
(1.185)
(0.564)
14.023
-13.254***
-5.085
-0.0.73
(18.927)
(4.843)
(3.177)
(1.512)
16.141
1.977
-3.820**
0.002
(11.221)
(2.871)
(1.883)
(0.896)
8.614
-82.712***
-57.527***
-39.038***
(69.388)
(17.757)
(11.648)
(5.546)
0.345
-0.507***
-0.310
-0.021
(0.291)
(0.074)
(0.049)
(0.023)
0.027
0.165***
-0.202***
-0.067***
∆% in total employed population ∆% workers in tourism ∆% workers in agriculture ∆% workers in maquila ∆% workers in oil Agriculture city=1 Maquila city=1 Oil city=1 Human development index, 2000 % of workers earning less than 2minimum wages,1995 % of population without coverage of social security,2000 Constant Observations
∆ in population without coverage of social security
∆ in workers earning less than two minimum wages OLS
∆ in human development index
(0.211)
(0.054)
(0.035)
(0.016)
-22.022
71.192***
61.787***
40.196***
(69.114)
(17.687)
(11.602)
(5.524)
159
159
159
159
R-squared 0.251 0.437 0.260 0.366 Notes: OLS: Ordinary least squares is a technique for estimating the association (parameter) between a dependent variable and independent variables in a linear regression model. This method minimizes the sum of squared distances between the observed responses in a set of data, and the fitted responses from the regression model. *** Significant at 1%, ** significant at 5% and * significant at 10%. Standard errors in parentheses
69
Table A.6.3 Regression results of internal effects (only tourism included)
∆% in total employed population
∆ in workers earning less than two minimum wages
∆ in population without coverage of social security
∆ in human development index
OLS
OLS
OLS
OLS
0.349***
-0.036
-0.030*
0.022***
(0.100)
(0.028)
(0.017)
(0.007)
0.560***
0.028
0.025
0.013
(0.124)
(0.034)
(0.021)
(0.009)
2.088
0.064
-0.280
-0.143
(1.282)
(0.358)
(0.217)
(0.095)
0.184
-0.129**
-0.016
-0.015
(0.187)
(0.052)
(0.031)
(0.013)
0.232
-0.199
0.114
-0.074
(0.855)
(0.239)
(0.145)
(0.063)
-3.373
6.098
-1.666
-0.653
(26.198)
(7.332)
(4.445)
(1.949)
-21.272
-1.273
-1.393
0.816
(28.430)
(7.957)
(4.824)
(2.115)
-20.290
6.915
0.404
0.866
(32.683)
(9.147)
(5.546)
(2.431)
-18.115
6.957
4.244
-0.745
(25.200)
(7.053)
(4.276)
(1.874)
Other little towns=1
Dropped
Dropped
Dropped
Dropped
Human development index, 2000
-45.218
-151.968**
-37.346
-51.737***
(255.241)
(71.436)
(43.313)
(18.989)
0.099
-0.479**
-0.097
0.002
(0.647)
(0.181)
(0.109)
(0.048)
0.045
0.256*
-0.024
-0.125***
(0.469)
(0.131)
(0.079)
(0.034)
56.779
112.998
36.417
53.181***
(243.256)
(64.384)
(38.945)
(18.097)
45
45
45
45
0.398
0.710
∆% workers in beaches resort ∆% workers in small beaches ∆% workers in arqueological ∆% workers in Pueblos Mágicos ∆% workers in other little towns Beaches resort=1 Small beaches=1 Arqueological=1 Pueblos Mágicos=1
% of workers earning less than 2minimum wages,1995 % of population without coverage of social security,2000 Constant Observations
R-squared 0.639 0.527 Notes: ***Significant at 1%, ** significant at 5% and * significant at 10% Standard errors in parentheses Notes: ***Significant at 1%, ** significant at 5% and * significant at 10% Standard errors in parenthese
70
