Migration and gastronomic employment for gastronomic MSMEs: the data multilateral banks already read

Why does a profitable restaurant get turned down for credit?

A loan file collapses over payroll, almost never over sales, and that is the most expensive lesson this sector pays for. Consider the 22-table restaurant in Lima that lost a 40,000 USD working capital line in March:

it was billing well, it had a queue on Fridays, and still seven of its eleven staff were paid in cash, with not one verifiable month of employment continuity. The analyst had nothing to score. Against that wall runs a sector that, in the United States alone, employs 10 % of the workforce according to the National Restaurant Association 2024, and that across the region hires fast and asks few questions. The figure deciding your access to the financial system is not average ticket: it is the number of consecutive payroll months you can prove on paper. Cheapening payroll for one quarter makes the operation more expensive for years, and the number that proves it sits in waste.

The real cost of hiring fast and asking little

Some 7.7 million Venezuelans live outside their country according to the R4V Platform, and the kitchen is usually their first job because it is the door that opens without papers. That tacit deal carries a deferred price: when the learning curve restarts every quarter, waste climbs, and ReFED 2025 documents that 70 % of foodservice waste comes from food the guest left on the plate. A new cook over-plates because he does not know the gram weight. Multiply that excess by 180 covers a day and you will see cheap hiring gets paid twice, first in product and then in the margin you meant to defend. Six percent. That is the productivity ratio of the regional MSME against a large firm according to ECLAC's work on structural heterogeneity, while the equivalent European figure sits near 40 %. I got this wrong for years, blaming the difference on hours or on willingness.

Productivity: the gap effort will not close

It is neither. Whoever does not record does not learn, and whoever does not learn buys badly again the following Monday. ECLAC itself measured in 2024 that under 4 % of firms in the region use artificial intelligence versus over 20 % in Europe, a figure confirming where the bottleneck lies: in the absence of systems, not in people's capacity. Diego F. Parra insists at Masterestaurant on one simple test: if you cannot reconstruct last week's protein consumption in five minutes, you do not have an operation, you have intuition. Banks have spent half a decade trying to replace hard collateral with observable behaviour, and the informal restaurant simply offers none. Roxana Maurizio, regional specialist in Labour Economics at the ILO for the Southern Cone, argues that informality does not yield to inspection but to registration incentives that give the employer something tangible back. Credit is that something.

No time series, no alternative scoring

And here the counterfactual is worth running: had that Lima restaurant banked its eleven staff eighteen months earlier, it would have handed over eighteen data points of a series —turnover, hours, declared tips— and the analyst would have had surface to score. The 40,000 USD line is approved or denied on that series. Formalising costs money, nobody disputes it; not formalising costs the capital that funds the second location. The owner cushions his own risk by releasing staff when demand drops and rehiring when it recovers, so the cushion is supplied by the migrant worker. It works, but it charges interest. Every cycle restarts the learning, pushes waste upward and sets SDG 8 back in official statistics. There is a genuine tension worth resolving: the business needs flexibility because restaurant demand is seasonal, and yet stability is the only thing that produces the track record banks demand. The bridge is contractual, not moral.

Turnover as a cushion: who pays the interest

Registered hourly contracts, with documented shifts, give the owner flexibility AND give the system traceability. The BLS reported in 2023 that 36.9 % of people aged 16 to 19 were in the labour force, almost always on split shifts; that format already exists, it just needs recording. Between 60 % and 70 % of workers in hotels, catering and tourism are women according to the ILO in its Sectoral Brief on gender, and in Spain the Anuario de la Hostelería 2024 puts that share at 54.3 %. Now contrast those numbers with entrepreneurship: UNDP calculates that 65.6 % of new e-commerce stores in the region are led by women, and Women Entrepreneurs Grow Global recorded that they started 49 % of new businesses in 2024, a five-year high. The conclusion arrives before the argument: whoever sustains the operation is not whoever reaches the capital, because an informal payroll history transfers no credit reputation to its holder.

Women in the kitchen: majority on payroll, minority in credit

Formalising the floor and kitchen payroll builds, along the way, the file that a station chef will use to request her own loan. Three scenarios, three different decisions from the same figures. Small restaurant, up to 12 staff: your goal is not formalising everyone tomorrow, it is closing SIX continuous months with at least 60 % of payroll banked, the minimum window an analyst uses to build a trend. Mid-sized operation, 13 to 40 people: attack turnover first, because with 70 % of waste originating on the plate (ReFED 2025) every restarted curve costs you food cost points; cap the ceiling at 32 % per dish and measure who is plating when it spikes. Group of three or more locations: consolidate a single time series and negotiate as a group, not as three loose files, particularly if 75 % of your traffic happens off-premise, as Circana estimates. There delivery data weighs as much as payroll.

Where these benchmarks come from and what they miss?

It is worth saying plainly: these figures come from public multilateral sources and sector associations, not from a proprietary restaurant sample. ECLAC measures structural heterogeneity over formal firms, so the real productivity of the informal MSME is probably worse than that 6 %.

Data from ReFED and the National Restaurant Association describe the US market and work as a directional reference, not as an exact equivalent for a venue in Bogotá or Quito. The R4V Platform updates its 7.7 million with different lags by country. And world hunger, which SOFI 2025 places between 638 and 720 million people, reminds us what all this formalising is for. Use them as a compass. The only benchmark that decides your credit is your own payroll series. No time series, no alternative scoring. Commercial banks with MSME portfolios have spent half a decade trying to replace collateral with observable behaviour, and the informal restaurant offers none.

Four differences a programme officer checks first

Roxana Maurizio, ILO regional specialist in Labour Economics for the Southern Cone, keeps arguing that informality yields to registration incentives that give the employer something tangible back, not to inspection. Credit is that tangible thing. Turnover cushions the owner's risk and shifts it onto the migrant worker. When demand dips the venue lets people go; when it recovers, it hires again. The cushion works, yet it charges interest: the learning curve restarts each quarter, waste climbs, and SDG 8 slides backwards in national statistics while the P&L still looks healthy. The gastronomic skills gap is not about attitude, it is about accreditation. A Venezuelan cook with nine years at the stove in Caracas lands in Bogotá with no paper proving it, and the market pays him as an assistant. Open Badges micro-credentials close that gap within weeks and lift effective wages between 18 % and 30 % without eroding margin, because productivity rises ahead of cost.

Four differences a programme officer checks first — in practice

Short supply chains and circular economy stop being rhetoric once data exists. A venue measuring its food loss and waste discovers in week one that it throws away between 8 % and 15 % of what it buys, and that single figure is enough to renegotiate delivery frequency with a nearby supplier. SDG target 12.3 gets met through the till, not through conviction.