Bad policy: short-term handouts trumping long-term growth

How would you go about entrenching poverty in a primarily rural, formerly agriculture-dominated tropical island economy?

Your land is blessed with two monsoons that almost split the year perfectly in two, interspersed with inter-monsoonal rains. You have diverse agro-ecological zones, and what's more those two monsoons approach your island from diametrically opposite compass points, spreading the water wealth. Your people are inheritors of centuries of vast public irrigation works that have conquered any imbalance in water availability, bringing liquid gold to the most fertile and agriculture-friendly parts of the island.

Let's assume you already have highly fragmented landholdings due to weak land markets, previous cock-eyed land reforms and a lot of inertia around more meaningful reforms. People with handkerchief sized sub-scale plots can't sell their lands due to unclear title or worse, arcane regulations. Rational land allocation is distorted by regulations that forbid any use but agriculture, sometimes of specific crops (ref. self-sufficiency) - you can't even  grow something the market might actually want.

Coddle local producers to the point of inefficiency with populist import protections, price controls and astronomical, unsustainable input subsidies (which by one of my calculations almost completely wipe out the GDP contribution from paddy); then cut them loose and bring in cheap imports when prices spike just ahead of the local harvests (or near elections) - depressing prices for everyone and killing incentives in local agriculture. Once in a while try to bring in a guaranteed-price centralized purchase scheme to repeat dramatic failures of the past. Big farmers want to cut their losses, small farmers just revert to subsistence, turning away from the vagaries of the market. Value-adders lock up their expensive (imported) machines whose output cannot be sold at a fair price thanks to the controls.

Self-sufficiency is one double-edged sword to swing around too much. The food shocks of 2008 provide an easy stalking horse for all kinds of protectionism taken too far. All kinds of marginal lands start coming back under cultivation, some of it just for subsistence, all competing with the "high-potential" zones that should be left to cultivate commodities efficiently - and cheaply - without specializing into higher-value crops or away from agriculture altogether.

There's no easy fix of course. Most of today's marginal lands were the agricultural heartlands of yore, sometimes by accident, sometimes reflecting the changing state of the art in technology. Like in Japan, agriculture, particularly rice, is embedded deep in our souls, and can't be turned away from easily. Few other alternatives exist - development in held back by isolation in its many forms; roads, technology and education have taken the longest to penetrate - and make an impression - in those same areas.

With an end to conflict, the clock is winding down on inefficient land allocations - if we don't get smarter about what we grow, where, farmers around the island will be faced with harvests they can't sell. Can we get markets working in time to keep the right farmers growing the crops best suited to their lands (and all the handicaps we can't dispense with overnight) - and to the market opportunities out there?