Two weeks ago I started a serialized essay on Latin America’s middle class that will appear every other week on the AQ blog. As I wrote at the time, Latin America’s middle class has received a lot of attention of late, including worry about its size, praise and high expectations for its growth and debate about its future. It’s also sparked a fair amount of speculation by businesses about its market potential.
But as I wrote two weeks ago, Latin America’s middle class is much more heterogeneous and, quite frankly, poorer and marginalized than many of us in developed countries would believe. In the last post, I talked about the definition of the middle class. In this one I talk about access to banking. And not to give away the punch line: it’s lower than you think. In later ones I’ll talk about wage security, education, access to health care, access to insurance, quality of housing, levels of satisfaction, and support for democracy. But more about this in subsequent posts.
Now it’s about the integration of Latin America’s middle class into the formal banking/financial system.
Consider this: according to an article that appeared in The New York Times on August 18, 2009, in New York City (notice how as a resident now I capitalize City as if to give it special meaning. Why I don’t know.) “in the world’s banking capital, 12 percent of households do not have a bank account” compared with 8 percent nationally.
That’s in New York and is a statistic that transcends socioeconomic groups—upper class, middle class and the poor.
Consider this: in Latin America, in Mexico, for example, the middle class’s access to formal credit averaged just over 20 percent (in other words, just under 80 percent of the Mexican middle class did not have access to formal credit sources—banks, credit agencies, credit cards, etc.). And in Peru under 18 percent had access to savings accounts.
While the U.S. shouldn’t necessarily be the basis of comparison (though it is for many of us subconsciously) look at the numbers: 88 percent of the total population in New York City are banked; only just over 20 percent of the urban middle class of the Peruvian and Mexican middle classes have access to formal banking structures.
What Banerjee and Duflo did was look at household surveys from 13 developing countries. By defining the middle class through consumption patterns (those with a per capita consumption level of $2 to $10 per day and then dividing it between those who consume between $2 and $4 and between $6 and $10 they look at consumption—rather than income) they and make comparisons between levels of the middle class. In Latin America they draw on household survey data from Guatemala, Mexico, Nicaragua, Panama, and Peru.
Here’s a summary of what they find. In rural areas in Mexico, whether at the low end of the bourgeoisie or the high end, the numbers were low for those of the middle class who had at least one loan (through a bank, money lender, micro-credit institution, or credit union): 21.5 percent for those consuming between $2 and $4 per day and 22.5 percent of those consuming between $6 and $10. And credit access was only marginally higher in urban areas (where you’d expect them to be higher): 23.5 percent and 29.7 percent, respectively.
Peru was slightly lower. Only 19.3 percent of rural Peruvian households with daily consumption levels between $2 and $4 had one loan and 29.2 percent of those households who consumed between $6 and $10 per day confessed that they owed a financial institution.
And in what is considered Latin America’s banking and financial capital, Panama: in rural Panama for the same ranges it was 6 percent and 12.6 percent. Urban Panama: 14.5 percent and 18 percent. Low in comparison.
[Data were not available for Guatemala and Nicaragua.]
Household loan rates of between 29.7 percent (Mexico’s urban middle class) and 6 percent (Panama’s rural middle class) would indicate that Latin America’s middle class either is underbanked or leans toward savings more than borrowing. Mind you the definition of a loan here can include credit-card debt, mortgage, a microloan, or even something bought on credit from a moneylender.
One assumption here would be that Latin America’s middle classes are exceptionally frugal and savings conscious. By these standards, though, even the most savings-oriented middle class (say in Asia) would have higher borrowing rates. Rather, the more logical conclusion (based not only on the low savings rates generally in these countries, but also their governments) is that this is the result of an intractable structural and cultural obstacle.
Here’s the proof. The percentage of middle class households that had saving accounts in Mexico, Panama and Peru ranged from 3 percent (Peru) to 33.5 percent (Mexico). These are by any standards very low rates. Let’s summarize: In Mexico, 17 percent of the households in the rural middle class consuming between $2 and $4 had a savings account, and 33.5 percent of those rural consumers spending between $6 and $10 had one. The urban Mexican middle class: 16.5 percent and 33.4 percent, not any better.
Panama’s rural middle class isn’t much different, though its urban middle class scores better. Rural: 10.1 percent ($2 to $4) and 33.2 percent ($6 to $10). Urban: 23.2 percent ($2 to $4) and 45.1 percent ($6 to $10).
Peru’s middle class seems to suffer from both inequality by geography and across definitions of the middle class. In Peru for those consuming between $2 and $4 per day in rural areas the percentage that choose to save formally was 3 percent, a number that increases to 18.4 percent for the $6 to $10 group. In urban areas it was 6 percent ($2 to $4) and 15.6 percent ($6 to $10).
What do we make of this? Several of these conclusions will be expanded upon in later posts—particularly when we discuss job security, access to the social-safety net (public and private) and risks to the private sector in an economic downturn.
First, at a superficial level, Latin America’s middle class in countries like Mexico and Panama (less so Peru) is surprisingly underbanked. This presents severe constraints on the limited capacity of the emerging middle class to consolidate their position and contribute (fiscally, politically and organizationally) to the broader economy and—by extension—the political system.
Second, in some places, living in urban areas matters. The access to banking facilities (with the exception of Panama) can improve the opportunities for middle-class representatives to gain access to the formal banking sector, though not by huge margins. This is a national issue. Whether it’s for reasons of lack of physical access, cultural distrust or lack of tangible benefits is unclear. But clearly, at one level, for the middle classes in these countries (and dare I say in many others) this is an endemic problem of scale, access and trust that differs profoundly from the middle classes of developed countries.
Third, there is a huge discrepancy between the lower rungs of the middle class as defined by Banerjee and Duflo and the upper rungs. Across the border, irrespective of rural or urban placement, the most important variable in determining bankability was the division between the $2 to $4 class and the $6 to $10 class. Now clearly part of this may be just assets—those in the higher end are better placed to save and borrow. But a more troubling trend may be at work: a stark division of access across the middle class. What this speaks to is the need for a more concerted public and private sector effort to reach down to these lower echelons of the middle class. And this requires an understanding—of observer and investor alike—that the middle class in Latin America is not homogenous and different members have different resource, access and structural constraints..
Next up: wage and employment patterns of the Latin American middle class. And later? Levels of satisfaction and support for democracy and levels of education—a point raised powerfully by Colombia’s Shakira
*Christopher Sabatini is editor-in-chief of Americas Quarterly and senior director of policy at the Americas Society and Council of the Americas. He regularly contributes to the AmericasQuarterly.org blog.
Guatemala City, Guatemala
Rio de Janeiro, Brazil
San Salvador, El Salvador
Julio Rank Wright
Christian Gómez, Jr.
Johanna Mendelson Forman