First and foremost a fence needs to be built along the entire length of the US-Mexico border except where terrain makes it unnecessary. A 3.5 mile stretch that is currently being planned will cost $58 million. This would mean a cost of $32.4 billion for the entire 1,950 mile border. However, since the fence exists in places and will not be needed in others, around 1/4 could be eliminated. This would make the cost of the fence around $24.3 billion. Assuming it needs to be rebuilt every 20 years would lead to an annual cost of $1.2 billion. Its primary role is to limit and not to stop illegal crossings. In 2005 there were 1.1 million border detentions. With a fence I would think initially the number would stay the same as fewer people try but a greater percentage are caught. Over time I would expect the number caught to drop significantly to the 400-500,000 range as this and other policies make it not as worthwhile to illegally cross into the US. This is the main priority in my attritional strategy.
Step 2 is to start punishing those who cross illegally, even if only mildly. Currently they are simply fingerprinted and put on a bus back to Tijuana where they can wait for nightfall to attempt another crossing. The next part of the attrition strategy is to detain people caught for 1 month for the first attempt and 1 month extra for each additional attempt. Assuming statistically that 1/2 would be first timers, 1/4 would be on their second attempt, 1/8 of their third, etc. then the average stay would be 2 months. It would be silly to detain everyone at first since the numbers would be great. In the first year or two it would be better to either randomly detain people or only detain those on their second or higher offence. Not a horrific punishment but still something that will discourage most repeat offenders. When detention center construction catches up with the lower number of detainees then they can all be held. While I would expect lower numbers, it would safe to build capacity for 600,000 annual detentions, or 100,000 beds. It cost a Pennsylvanian county $24 million for a 300 bed prison, so 100,000 beds would cost $8 billion. Assuming the standard prison expense of $30,000 per detainee-year then annual cost would be $3 billion. These first two proposals deal with enforcement on the border. With that taken care of, enforcement in the interior can be considered.
Interior enforcement has two main parts, fining companies for hiring illegals, and requiring local government officials to report illegals. As with the fence, the goal is not to fine every company for every illegal they have hired. A handful of companies that are the major employers of illegals would be selected, audited, and hit with the current fines for the illegals they have hired. The fine should be raised, at least five-fold, and the process repeated. Companies will inevitably decide it is not worth hiring illegals and most will cut back. Not all, but enough. As for local government officials, they should be required to report any illegals who attempt to enroll kids in school, are pulled over, or otherwise come into to non-emergency contact (e.g. domestic abuse) with local officials. Many will of course refuse to comply with this law so they should be punished with loss of federal funds. Nothing quite focuses a bureaucrat's mind like losing their federal funds.
Now that it is much more difficult to get in illegally, people are punished for trying, employers are bit more reticent about hiring illegals, and local officials will turn illegals over for detention and deportation, we can discuss what to do with the 11 million here. My plan is to offer them $3,000 if they turn themselves in and go back to Mexico. It would be a deal few could refuse, and those that do will be caught eventually (and be detained and deported without a payment). Fingerprints, or other biometrics, would be used to make sure no one tried to get back across and claim another payment. Assuming they all take the deal it would cost $33 billion. It is better to have as many leave voluntarily than to have to track them down and deport them or just let them stay.
This last idea is not directly related to the others. It could be implemented and have an effect even without the others. The policy regarding Central Americans has recently changed. In the past we would give them a slip with a court date and tell them to appear for their deportation hearing; which, of course, few did. Now we put them on a plane back to their country. This effectively ends their attempts at illegal immigration since the trip is so expensive and dangerous that most do not make it again. It is such a daunting prospect that many Central Americans are now trying to pass themselves off as Mexicans so they will simply be sent back across the border to rest for another try. We could do something similar with Mexicans by flying them to a city in southern Mexico. The expense of getting home, much less back to America, would be so great most would not attempt to cross again. The annual cost for this program would be around $1 billion.
It is unnecessary to deport everyone here illegally, to stop everyone crossing illegally, or to fine every company hiring illegals. Those are strawmen built up by the pro-illegal immigration side. With an attrition strategy the price for each attempt (currently almost nothing) would be raised. If currently 1/2 of all attempts are successful than statistically (assuming no one gives up) everyone will eventually make it. If we reduce the success rate to 1/3 then only half will make it. If only 1/4 make it each time then only 1/3 will ever make it here. We further reduce that by making people give up after the first or second attempt by mildly punishing them. The ones already here are dealt with by limiting their job opportunities and ability to function as American citizens do. Add in a cash payout to leave and the problem of illegals already here can be mostly handled. A few hardened illegals may refuse to stay, but it is easier to deport 1 million than it is to deport 11 million. The total cost of these proposals is $65.3 billion with an annual cost of $4.2 billion. This is pocket change for an economy that will produce over $13,300 billion this year, especially in light of the importance of securing our borders.