Worker remittances constitute a critical flow of foreign currency into Latin America and the Caribbean. Remittances exceeded $58 billion in 2006, greater than the total value of development fund inflows to the region. In at least six countries, remittances account for 10 percent of GDP. The market is so vast that it could easily accommodate new players and business models. Mexico accounts for roughly 40 percent of remittances to the region, and serves as an important barometer. In 2006 it reported only 15 percent annual growth and first semester 2007 data indicates virtually no growth.
This downturn is not as alarming as it may initially seem. Firstly, historical growth rates of remittances were inflated by the positive effects of gradual improvements in reporting mechanisms. True organic growth was therefore lower than the recorded averages between 2000 and 2006. Secondly, current softness in the market will revert within the next two to three years, because the strongest factor affecting remittances continues to be the demand for cheap labor in the US, and long-term demand remains strong. Growth will likely return, albeit at more modest rates of 10 to 15 percent, within three years. Neither the forces contributing to these swings in remittances, nor the duration of the downturn, are likely to establish banks (or any alternative remittance model) as a serious threat to the well-established money transfer operator (MTO) model.
The factors contributing to the scorching growth rates in remittances between 2000 and 2004 must be illustrated to understand why growth rates for the next two years will be substantially more modest. Improvement in the reporting capacity Latin American governments played a considerable role. Historically, the bulk of worker remittances were lost in the errors and omissions categories of most countries’ balance sheets. Responding both to World Bank initiatives (also known as the Yellowbook Reports) and market pressures, all countries in the region improved ACH capacity and electronic payment reconciliation between 2000 and 2004. This contributed to more accurate and transparent reporting of money flows. It is impossible to accurately quantify the effect of improved reporting capacity on the growth of reported remittances levels, but it’s safe to say that much of what was apparently growth was in reality reclassification of existing payments.
Another factor in remittance growth was the number of illegal immigrants successfully crossing the Mexican border. Not all remittance senders to Latin America are illegal immigrants, but many are, and it stands to reason that variations in illegal entry volume will directly impact remittance figures. The number of apprehensions made by U.S. border authorities usually correlates with the volume of immigrants in a given year. It will then take illegal immigrants some time to find a job and begin sending money home. In 2004, when remittances to Mexico grew 24 percent, the number of apprehension increased 20 percent the year prior. Not surprisingly, the significant softening of remittances in 2007 follows a year marked by significant upgrades in border patrol and fencing, and less than 900,000 apprehensions.
It is hard to predict apprehensions for any given year, or to know how this will affect remittance volumes. Initial DHS data for 2007 indicates that results will not vary substantially from 2006, which suggests similar flat results for 2008 remittance volume. If proposed upgrades to border patrol resources come through in the next couple of years it is likely the number of successful crossing will drop by 2010, and may result in the first actual drop in remittances, particularly the U.S.-Mexico corridor.
Compounding stricter U.S. border control is the cessation of railway operations on Mexico’s southern border with Guatemala. Over 20 percent of illegal immigrants entering the U.S. first cross through Mexico. At the Guatemala border, they made use of the railroad carts operated by the Genesse & Wyoming concession. This rail link was suspended in July 2006 due to hurricane damage, thereby obliging Central American immigrants to walk 300 kilometers from the border to the nearest rail station. This negatively impacted remittances in 2007 by interrupting or deterring pass through flows of Central Americans eventually crossing the U.S.-Mexico border during 2006.
Falling remittance volume from the U.S. to Latin America does not necessarily mean that the remittances dry up all together. In many cases they switch corridors and take on new forms. Case in point is the Mexico-Guatemala corridor. Historically, this was a transition point for Central American workers but is now becoming a remittance corridor in its own right because of the increased time Central American immigrants are spending in the region, either as a consequence of logistics or the closing of traditional border crossings on the U.S.-Mexican border. InfoAmericas estimates that over 50 percent of the remittances sent home from southern Mexico are in cash and therefore are not being formally registered.
Countries such as Spain and Japan are also emerging as significant providers of remittances (See Remittances Start Slowing Down and A Yen for Remittances). This year the Spanish Institute of Statistics (INE) reported that Spain hosts nearly two million Latin American immigrants, of which just over 30 percent arrived since 2005—more or less starting when U.S. border entry became more difficult. Spain formally assimilates Latin American immigrants much more willingly than the U.S., and this encourages immigrants to consider Spain despite the greater distances and more complicated transportation logistics.
Traditional intra-regional corridors are increasing in importance. Remittances between Nicaragua and Costa Rica, Haiti and the Dominican Republic, and Brazil and Bolivia all show growth in excess of 10 percent for 2007. Remittances along these corridors are mostly cash-based, but they are gradually adopting formal channels.
DEMAND FOR LABOR
Demand for cheap labor in the U.S. impacts remittance flows more than border controls. Higher inflows of both legal and illegal immigrants between 2000 and 2004 reflected strong demand for construction workers in the U.S. The softening of the real estate market in 2006 and 2007 explains a good deal of the interruption in remittance growth so far this year.
According to the U.S. Construction Outlook published by Reed Construction Data, the construction sector was replacing 92,000 vacancies and creating an equal number of new positions from 2001 to 2005. Between mid-2006 and mid-2007 the sector lost over 100,000 jobs. The net effect on the labor pool is therefore nearly twice as severe because new positions also ceased being created.
Fully one-third of U.S. construction jobs, or 2.5 million positions, are filled by U.S. Hispanics, or more accurately, Latin Americans with varying degrees of assimilation. U.S. Hispanics are the fastest growing ethnic group within construction, and they are also the most under-represented within the sector’s labor unions. Estimates made by InfoAmericas and other research organizations show that a further half a million jobs may be held by irregular Latin American workers who have no job security.
Assuming the informal sector is hit prior to cuts in the formal labor pool, this leads to conservative estimates of a quarter of a million jobs lost by the formal and informal Latin American/U.S. Hispanic contingent. Various studies calculate the incidence of remittance sending to be 25-35 percent among U.S. Hispanics in the construction sector, with an average yearly send of $4,000-6,000. All told, this would represent a drop of about $500 million in remittances during 2007.
This is a temporary situation. New construction starts are expected to begin recovering in 2009, and the construction industry needs Latin American immigrants to fill existing positions. The retirement of the baby boomer generation is creating vacancies, and the number of 20 to 44 year old American men is diminishing. The average age of the U.S. construction worker in 1998 was 32, today it is 38. Non-Hispanic men are increasingly less inclined to take construction jobs, and this will increase demand for Latin American labor whose average age in the construction industry is 28. Similar trends are evident in the agricultural industry, where the participation of Latin American migrant workers is even higher.
ACCESSING THE CLIENT
Some pundits speculate that a downturn in the market will spawn greater competition in remittance services, facilitating the increased participation of banks. Although the market is indeed adjusting, it is doing so in a manner that will continue to limit the role of banks and their ability to access non-banked clients, who represent the bulk of remittance senders and receivers.
One of the primary limitations is the banking infrastructure itself. Banking penetration in Latin America is growing at 7 percent per year and 65 percent of the population remains un-banked. Most populations under 100,000 inhabitants do not even have an ATM. Financial infrastructure is expanding in the region, and this points to potential incursions by banks into the remittance market.
Nearly all banks offer wire transfer services that are safe, expensive, and take many days. Wire transfers also require an account in both the sending and receiving bank, and are therefore best suited to transfers in excess of $1,000. Most remittances average $320. Over the last ten years banks responded by signing-up traditional MTOs to serve as embedded agents in bank agencies. This was a cost effective way of not having to develop their own remittance platforms and customer service capabilities.
In the last couple of years various banks announced in-house remittance services that compete with MTO products. Official identification requirements hindering opening of accounts by migrant worker in the U.S. was overcome by accepting consular IDs. The direct price of these services is generally lower than traditional MTO costs. With banking penetration on the rise in Latin America, this combination of factors seems to forebode a threat to MTOs, particularly if one believes the current flattening of growth rates heralds a long-term trend of consolidation. In reality however, the impact of banks getting into the remittance market will remain limited for quite some time.
Banks must first overcome the suspicion with which they are viewed by the middle and lower social classes in Latin America. Amongst senders the situation is worse because although they can now present consular IDs, they remain suspicious of exposing their informal migrant status. Over 90 percent of senders that used both MTO and bank services prefer MTO services because they are faster and feel they are better treated. This last preference makes sense when one considers that many bank branch managers in both the U.S. and in Latin America are not eager to mix comparatively affluent clients with a socio-economically poorer class on the same service floor. Indeed, the purpose of offering remittance products is not to compete in the remittance market but to provide products that facilitate new customer acquisition amongst the un-banked.
According to the 2005 Journal of Payment Systems, nearly 75 percent of banks offering a remittance service product, offer it only to institution members. In Mexico, only one-third of remittance recipients have bank accounts and in El Salvador just one-fifth and of these, 86 percent prefer to pick up cash at a teller window as opposed to depositing it in their account. This preference is largely because of the indirect costs and disadvantages of using a bank for remittance services. These inconveniences range from the cost of opening and maintaining an account specifically designed for remittances (most normal checking and saving accounts do not qualify for these services) to the wait time (24-72 hours) before the money is available. Furthermore, bank customer support centers are under-prepared to field remittance related calls. Ultimately however, most remittance receivers rely on cash purchases and see little or no use for a pre-paid or debit card, which are not accepted where they shop.
Competitive products from banks could upset the position of MTOs, particularly if the market is contracting in the short term due to lower demand for migrant labor and tougher border controls in the U.S. Realistically, banks are not prepared to enter this market competitively without using MTOs as agents. They are not geared to be efficient channellers of small value remittances and they are not nimble enough to work with new corridors that accompany fast-moving migrant labor flows. In this regard, MTO offer tens of thousands more points of service that often go into rural areas where no bank exists. These networks already exist and can quickly be adjusted to service new opportunities.
Jan Smith is director for the financial services practice of US-based consultancy InfoAmericas. This article is republished with permission from Tendencias, the magazine of the InfoAmericas.