Essay · Tourism economics
The Local Dollar
Why overtourism is a value-distribution problem, not a crowd problem. And what an operator who builds AI would do about it.
Barcelona is raising its tourist taxes again. Venice now charges day-trippers a fee just to enter. Amsterdam capped cruise ships and stopped building hotels. These are serious cities trying to fix overtourism, and every one of them is treating a money problem like a crowd problem.
See Sight Tours has run small-group tours since 2006, across about 22 North American cities, more than 200,000 travelers a year. Stand at the curb in any of those cities and the pattern is obvious. What residents resent is not the number of visitors. It is that the visitors come and the money leaves.
The leak nobody names
Tourism economists have a word for this: leakage. It is the share of what a traveler spends that exits the local economy instead of staying with the people who live there. It is the most important number in the overtourism debate, and almost nobody says it out loud.
The figures are blunt. The UN Environment Programme estimates that of every 100 dollars a traveler from a wealthy country spends on a vacation, only about 5 dollars stays in a developing-country destination. Five. The other 95 fly home before the traveler does. For most developing economies, UNEP puts import leakage at 40 to 50 percent of tourism earnings. In all-inclusive package tours, it estimates about 80 percent of spending goes to airlines, hotels, and international companies headquartered somewhere else.
Overtourism isn't a crowd problem, it's an economics problem. When most of a destination's tourist spending leaks out, locals carry the costs without the benefits.
That is what breaks the deal. A resident in a tourist city pays the full price of the crowds, the rents, the lines at their own grocery store, and sees almost none of the upside. Of course they want fewer tourists. They are not getting paid.
It is ownership, not headcount
Here is the part the volume debate misses. Leakage is not driven by how many people show up. It is driven by who owns the thing they spend money on.
The cleanest proof comes from Bali. A peer-reviewed study compared hotels in the same destination and found that locally owned, non-star hotels leaked roughly 2 percent of their revenue out of the local economy. Foreign-owned four and five star chains leaked more than half. Same island, same tourists. The only variable that changed was who owned the room, and where the booking fees, the management fees, and the profits went.
Change the ownership and you change the leak. Volume barely moves it.
The new layer: digital leakage
The old leakage was foreign airlines and hotel chains. The new one is the booking layer. When a traveler books an experience through a large online travel agency, that platform typically takes 20 to 30 percent before a dollar reaches the local guide who runs the day. Viator's standard rate is around 25 percent. Hotels run 15 to 25 percent. We have paid those commissions for 18 years and run payroll against them, so we will say it plainly: the industry stacked a digital skim on top of the old leak and called it distribution.
The reframe
So here is the shift the industry needs to make. Overtourism is not a volume problem. It is a value-distribution problem. A destination is only as healthy as the share of each tourist dollar that stays with the people who live there. Call it the local dollar.
Seen that way, the standard playbook is solving the wrong equation. A tourist tax restricts volume and does nothing to redirect value. In practice it often just adds a fee that flows to the city, not to the local guide. Visitor caps shrink the pie instead of changing who gets a slice. Both treat the symptom, too many people, and ignore the disease, too little of their money stays.
This is already happening
The fix is not hypothetical. The two models already run side by side, and the difference is measurable.
When tourism is built on local ownership, the money stays. New Zealand is the clearest case. Government figures put Māori tourism at about 1.2 billion dollars of GDP in 2023, up from 975 million in 2018, across roughly 3,600 Māori-owned tourism businesses employing around 15,000 people, and those businesses on average pay higher wages than their non-Māori equivalents. The revenue lands with local owner-operators because they own the business. That is the entire mechanism.
Bhutan does it by policy. Every visitor pays a Sustainable Development Fee, currently 100 dollars a night, and that money goes into a national fund that pays for citizens' healthcare, education, and conservation. The fee is the retained value, by design.
Now hold those against the extractive end. The all-inclusive resort and the cruise ship are the same model: lodging, meals, and entertainment are pre-paid to a foreign-owned company, so the traveler rarely spends outside it. Barcelona took 2.8 million cruise passengers in 2024, about 1.6 million of them day-trippers who never stayed a night, and the city is now doubling its cruise tax because high-volume, low-overnight tourism brings the crowds without the value. Venice, with roughly 50,000 residents, absorbs 30,000 to 40,000 day-trippers a day, and introduced an access fee for the same reason.
Local ownership keeps the dollar. Foreign, all-inclusive, platform-mediated ownership leaks it. That is not a prediction. That is the data.
The structural fix
If the problem is distribution, the fix has to be structural. You change where the dollar lands by default. And for the first time, the tools to do that are cheap and in everyone's pocket.
Short-form video is the cheapest distribution a local operator has ever had. A guide with a phone can reach more travelers in a week than a minivan reached in a year. The open question, the one that decides whether the local dollar stays local, is who captures the booking: a faceless aggregator that inserts itself between the creator and the customer, or the local who made the video.
That is the bet at Tripshepherd. Use AI and short-form video to route discovery and booking straight to local guides and creators, and pay them when a traveler books through their content. Not because it is the nicer story, but because it is the version where the dollar lands locally by default instead of leaking by default.
Which brings up the guides
There is a loud argument in travel right now about AI and tour guides. One camp wants AI to replace the guide with autonomous, human-free experiences. The other wants AI kept out entirely. Both are wrong, and we are in a rare position to say so, because we run the guides and we build the AI.
AI should not replace the guide. It should find them, train them, surface them, and put a paying traveler in front of them. The human at the curb, the one who knows which café is actually good and which viewpoint is worth the walk, is the single thing automation cannot fake. That human is not the cost to optimize away. That human is the product. And that human is exactly where the local dollar is supposed to land.
AI shouldn't replace the guide. It should find them, train them, and put a paying traveler in front of them. The human at the curb is the one thing automation can't fake.
Overtourism will not be solved by capping visitors or taxing them harder. It gets solved when a traveler's money can reach a local guide, a local creator, a local business, by default instead of by luck. Keep the dollar where the trip happens. Keep the human there too. Everything else is just managing the crowd.
Sources: UN Environment Programme, “Economic Impacts of Tourism” (UN Atlas of the Oceans); The Travel Foundation, “Creating Equitable Destinations” (2024); Suryawardani et al., tourism leakage in Bali's accommodation sector; New Zealand MBIE / Te Puni Kōkiri, “Te Ōhanga Māori 2023.”