universal minimum pensions in Mexico

Mexico’s federal government has made enormous strides toward a universal minimum pension for its aged population. Social pensions began in 2007 as an initiative of the centre-left PRD in Congress, which was able to pass legislation with help from congressional representatives of the other two major parties. The pension was initially universal, given to everyone from age 70, but limited to residents of rural communities. In early 2012 the centre-right PAN president, in a surprise move, made the pension available to everyone, rural and urban, but added a pension test. Importantly, the test applies to individuals, not families, so a person is not disqualified if a spouse has access to an occupational pension. Also, the test applies only to contributory pensions mandated by the federal government, so beneficiaries are allowed to access pensions from state and local governments. In 2013, the newly elected PRI president lowered the age of eligibility to 65 years, retained the pension test, and increased the size of the pension to 525 pesos (about 40 dollars) a month.

In 2007, the 500 peso social pension was equal to 33% of the minimum wage and 5.7% of per capita GDP. By 2012, 500 pesos a month was only 26.7% of a minimum wage and 4.4% of per capita GDP. The increase in 2013 to a 525 peso pension changed these numbers only slightly.

This modest social pension can be accessed by every resident of Mexico aged 65 and older, regardless of the income or wealth of his or her family, provided the beneficiary does not receive a contributory, employment-related pension. The minimum pension for workers in the formal sector is more generous. The federal government guarantees that all workers who qualify for a contributory pension (i.e. those with a record of at least 1,250 weeks of contributions) will receive an inflation-indexed benefit that is nearly four times the size of the social pension.

The universal minimum pension has the full support of all politicians, and all three major political parties in Mexico. The only dissent is from the PRD (Party of the Democratic Revolution), which would prefer to increase the size of the pension, index it to the minimum wage, and eliminate the pension test.

Since there is no internal opposition to Mexico’s new social pension, a California-based research institute – the prestigious RAND Corporation – is filling this void. In a new working paper, RAND researchers argue that the 525 peso social pension is not affordable. The paper is available only in Spanish, but a note in English on the front page indicates “RAND working papers are intended to share researchers’ latest findings and to solicit informal peer review”. I am accepting this invitation to publicly post these comments. TdJ criticized two previous RAND studies on pensions in Mexico. One is a glossy publication issued in October 2011 jointly with Centro Fox and the AARP. The other is RAND Working Paper WR-903 issued in November 2011 but no longer available online. Attempts to access the paper now returns a message “This working paper is currently under journal peer-review and will be reposted once it is published”.

The latest study is not easy to follow, as the authors fail to show their work. The relevant information that they do report is listed in table 1 below:

The researchers estimate that approximately 2/3 (5 million) of Mexicans aged 65 plus years have no access to a contributory, employment-related pension. This is plausible, though the numbers are lower than usually assumed. The federal social development ministry (SEDESOL) thinks that there are 5.6 million potential recipients of a social pension in 2013, although they expect the number of registered pensioners to reach only 4.4 million by the end of the year. This is a sharp increase from the 3.1 million pensioners on the list at the end of 2012, when the age of eligibility was 70 rather than 65. (See Sedesol, “Primer Informe Trimestral 2013“, pp.14-21.) According to the RAND researchers, the hypothetical cost to taxpayers of providing pensions to all 5 million qualified recipients is 31.5 billion pesos, equal to 0.28% of Mexico’s 2013 Gross Domestic Product (GDP). The number of pensioners is forecast to climb to 13 million in 2029, with the cost of pensions increasing 175%, to 86.8 billion pesos (0.45% of GDP).

This implies that the pension is becoming more generous over time, yet the working paper does not mention any size of pension other than the current 525 pesos a month. The hidden assumptions of the RAND study become clear if we look at some statistics that are implicit in table 1:

The implicit pension is equal to the cost of pensions divided by the eligible population. Implicit GDP is the cost of pensions divided by the cost/GDP ratio. The size of the pension is increasing at the same rate as GDP: 3.4% a year. The GDP projection supposedly is taken from OECD projections, in 2005 international (purchasing power parity) dollars, converted into Mexican pesos by the Rand researchers. The projections are thus in constant prices. But something is very wrong. The GDP figure for 2013 is too low, if the constant prices are 2013 prices (or 2005 prices re-based to 2013). The IMF, in its April 2013 World Economic Outlook Database, estimates nominal GDP for Mexico in 2013 to be 16713 billion pesos – 48.6% higher than the figure impicit in this RAND study. Perhaps the GDP figures are in 2005 prices (though they are not labelled as such). If so, then the size of the pension is vastly overstated in 2013 (the base year), and even more in subsequent years.

Everything becomes clearer if we add in some population projections, taken from the Consejo Nacional de Población (CONAPO) – the same source that the RAND researchers cite. The population is projected to increase from 118.4 million in 2013 to 136.6 million in 2029. GDP per capita then increases at the rate of 2.5% per annum, reaching 141,254 pesos in 2029. With the addition of population, it is possible to estimate a useful identity to calculate the fiscal cost of pensions: t = r*p, where t equals the cost (to taxpayers) of pensions as a proportion of GDP, r is the proportion of population eligible for a pension, and p is the ratio of the pension to per capita GDP. (For details, see my article in World Development, January 2007, p. 28.)

The RAND researchers believe that they show that the pension is very costly, hence not affordable, not fiscally sustainable. In my opinion, these numbers are not large at all. Even 0.45% of GDP is a small sum, by international standards, to spend on social pensions for seniors. Also, the numbers are vastly inflated! The 525 peso pension in 2013 is 4.46% of per capita GDP, NOT 6.63%, so the cost in 2013 of providing 5 million elderly with pensions of 525 pesos a month is 0.19% of GDP (0.0442*0.0446=0.0019), not 0.28%. And 0.19% of GDP is a modest sum, indeed. Moreover, there is every reason to expect the cost to fall, as a percentage of GDP, if the benefit is not adjusted for price and wage inflation. Even if the pension were indexed to the minimum wage, so that it stayed at roughly 4.46% of per capita GDP, the cost of the scheme would reach only 0.26% of GDP in 2029, using the same projections of eligible population and GDP growth that the RAND researchers used.

The conclusions of the paper for the most part are to be expected. They follow from the erroneous (unconvincing?) assumption that Mexico’s modest pension scheme is unsustainable in the face of population ageing. The authors recommend that benefits be targeted tightly (following the Oportunidades model), and that workers in the informal sector be forced to save for their own retirement, so that they will not have to claim a social pension. As TdJ has emphasized numerous times, means tests are taxes that fall on the poor, and have many defects, one of which is the high cost of administration. (The SEDESOL budget for administering the federal social pension is less than 5% of the pesos transferred to beneficiaries.) Forcing informal workers to save has not worked anywhere, and there is little chance that it will work in Mexico. Even if it were possible to force the poor to save for their own retirement, the reduced consumption this implies would not be in their own best interest. In the wise words of Estelle James, a World Bank economist and pension specialist, “Extending coverage by requiring low-income informal sector workers to contribute to social security would not be in the interests of these workers …, even if the government had the capacity to enforce the mandate.” (Quoted in L. Willmore, “Universal pensions for developing countries“, World Development 35:1 [January 2007], p. 27.)

One conclusion of the RAND working paper does worry me. I am troubled by it, so I quote the passage in full, and have translated it into English:

En cuanto a la atención a adultos mayores es necesario que exista coordinación con las diferentes instancias del Gobierno Federal en México, en especial con los gobiernos de los estados, con la finalidad de generar consenso en términos del objetivo general y específico de los programas, la población objetivo, los beneficios otorgados, así como los criterios de elegibilidad, con la finalidad de evitar duplicidades y mejorar el impacto de los programas.

En materia del programa de pensión no contributiva federal es necesario identificar el número de beneficiarios que podrían estar recibiendo beneficios por parte de otros programas de apoyo a adultos mayores, con la finalidad de cuantificar el monto de las posibles “fugas” de recursos.

[Regarding care of the elderly, coordination is needed with the various agencies of the Federal Government in Mexico, especially by the state governments, … in order to avoid duplication and improve the impact of programmes.

Regarding the federal non-contributory pension, it is necessary to identify the number of beneficiaries receiving benefits from other programmes for seniors in order to quantify the amount of possible “leakage” of resources.]

Emma Aguila, Nelly Mejía, Francisco Pérez-Arce and Alfonso Rivera, “Programas de Pensiones No Contributivas y su Viabilidad Financiera: El Caso de México“, RAND Working Paper WR-999, May 2013, p. 21.

Perhaps I misunderstood, but I read in this passage a fear that recipients of the social pension might be ‘double dippers’. Why is this a problem? In many pension systems, ‘double dipping’, even ‘triple dipping’, is the norm. Most elderly Canadians, for example, receive a noncontributory old-age pension, plus a Guaranteed Income Supplement (GIS) and a contributory Canada Pension. The well-to-do are not eligible for a means-tested GIS, but many have access to tax-advantaged retirement savings plans and occupational pensions in addition to the basic old-age pension.

In Mexico, the only federal programme with cash transfers for the elderly, other than the SEDESOL social pension, is a small cash benefit (345 pesos a month) given to families receiving benefits through ‘Oportunidades’, a programme for low-income families with children. This programme is very small, and will soon be phased out. A major defect (other than the small size of benefit) is that the cash allowance is given to the female head of the household, typically the daughter of the resident senior.

State governments are reacting in diverse ways to ‘universalisation’ of social pensions. Fourteen of Mexico’s 32 states have never provided a social cash pension of any kind to elderly residents. Some states with functioning social pensions, such as Chiapas, Guerrero and Puebla, have used the expansion of the federal programme as an excuse to end their own pensions. I suspect that RAND researchers might approve! Other states, such as the Distrito Federal (Mexico City), are continuing with social pensions as if nothing had changed. Jalisco has reacted in my opinion very well to the expansion of the federal social pension scheme. The government of Jalisco previously provided means-tested benefits of 500 pesos a month to a small number of urban seniors aged 70 years and older. From 2012 it guarantees a minimum pension of 840 pesos to everyone from the age of 70. If this new scheme is working as promised, most Jalisco seniors from age 70 who qualify for a federal pension of 525 pesos should be receiving a top-up of 315 pesos from the state government. A logical extension for Jalisco would be to lower its age of eligibility to that of the federal pension – age 65.

A monthly pension of 525 pesos (US$40) is tiny for a prosperous, middle-income country like Mexico. Numerous state governments in the past have allowed elderly residents to claim benefits simultaneously from the federal and state governments. This is permitted explicitly by the federal government. There is no reason for ‘double dipping’ to end, and the practice ought to be encouraged. My fear is that the RAND researchers, and others like them, will not agree.


Comments are closed.