Normally — a word that’s of diminishing utility in Washington — votes on major policy proposals, such as the Republican Obamacare overhaul, would be assessed by the nonpartisan Congressional Budget Office so that lawmakers have a sense of how much the bill would cost and what it would do.

In the case of the American Health Care Act, though, the most recent CBO analysis dates back to late March, the point at which Republicans gave up on trying to pass the bill the first time around. Since then, there have been several major changes to the bill in an effort to corral enough support from wavering Republicans to move it out of the House. What will the effects of those amendments be? The CBO hasn’t weighed in, but the GOP leadership is pressing forward anyway — perhaps, a cynic might suggest, because of how the CBO’s analysis of the initial effects stripped away support from members of their caucus.

So in lieu of full CBO analysis looking at how much the legislation might cost or save the government and how it would affect the number of insured Americans, we have cobbled together an independent analysis that will, hopefully, give some sense of what the bill might do.

The initial analysis

Since the policy at issue is an amended version of the original AHCA, we can start with the initial CBO analysis.

The very first iteration of the AHCA would have reduced government spending by $337 billion, mostly by cutting subsidies for those with expensive health-care plans. That savings comes despite cuts to taxes related to Obamacare.

By 2026, the net effect of the bill would be that 24 million fewer Americans would have health coverage than if the Affordable Care Act — Obamacare — were left in place. Nearly 1 in 5 Americans under the age of 65 would lack coverage by that year, compared with a bit over 1 in 10 today.

A substantial chunk of the cost savings (and reduction in coverage) is a proposed cut to Medicaid coverage of $880 billion over 10 years. That’s a reduction of about 25 percent. About half of the increase in insurance coverage under the Affordable Care Act was a function of expanded Medicaid eligibility. Incidentally, as a candidate, Trump pledged not to cut Medicaid.

The effects of the reduction in subsidies and the reduction in taxes would not be evenly distributed. Over the long term, the CBO figured that insurance premiums would drop — largely because some groups would be priced out of coverage.

Moving from scaled subsidies to a flat subsidy would mean that older and poorer Americans would have much less of their coverage subsidized, increasing the likelihood that they would let coverage lapse.

The shift that results from tax cuts and subsidy changes would increase the financial burden of those earning $50,000 a year in income or less — and benefit those earning more. Those making more than $200,000 annually would see a disproportionate gain.

The cut in taxes assessed earned the AHCA the endorsement of Americans for Tax Reform.

As Republican leaders scrambled to try to secure votes in March, they introduced several amendments aimed at increasing support. The CBO analyzed those, as well. The number of uninsured would still be 24 million higher by 2026, but the savings to the government would have dropped to $150 billion. Insurance premiums would only be slightly affected.

What the new amendments might do

After the bill was initially pulled, two amendments were introduced which have brought it back to life.

The first was an amendment backed mostly by the conservative House Freedom Caucus that, among other things, would allow states to apply for waivers to change the mandate for covering preexisting conditions — that is, medical conditions that exist before the issuance of an insurance policy and which might indicate higher long-term insurance costs. The amendment would similarly allow states to apply for waivers to define the essential benefits mandated in insurance coverage, allowing them to let insurers sell policies that are narrower in scope, potentially decreasing the cost of the policy but leaving those insured without coverage in some circumstances. This was the MacArthur amendment, named after Rep. Tom MacArthur (R-N.J.), a moderate who negotiated the plan with the Freedom Caucus.

The second was an amendment backed by Rep. Fred Upton (R-Mich.) that would add $8 billion in funding over five years to so-called high-risk pools meant to offset the higher insurance costs of covering those with preexisting conditions.

President Trump is joined by House Speaker Paul D. Ryan, congressional leadership and his family as he formally signs his Cabinet nominations into law at the Capitol on Jan. 20. (J. Scott Applewhite/AP)

It’s these changes that haven’t been assessed by the Congressional Budget Office. The Upton amendment emerged only this week, so it’s even less well explored than the MacArthur amendment but outside groups have analyzed the likely overall effects.

How many people have preexisting conditions and would therefore be at risk of increased coverage costs? (An important note: The MacArthur amendment would only allow rates to increase following a lapse in coverage.) The Kaiser Family Foundation analysis indicates that the breadth of what constitutes a “preexisting condition” — things from cancer to sleep apnea — means that 27 percent of Americans under the age of 65 fall into that category. That’s about 52 million people.

The American Medical Association released a statement on Wednesday criticizing the reliance on high-risk pools for covering those with preexisting conditions. Such pools “were not a panacea” for the 35 states that had them before the passage of Obamacare, with high deductibles, lifetime limits and big losses according to the Kaiser Family Foundation. The liberal group Center for American Progress estimates that the high-risk pools under the AHCA would be underfunded by about $20 billion annually. (Other estimates indicate a larger gap.) The Upton amendment adds $1.6 billion a year.

AARP estimates that annual costs for those with preexisting conditions could increase to $25,700 annually — relying on those high-risk pools to offset the costs. New analysis released on Thursday estimates that the funding in the AHCA would be enough to cover only 5 percent of those currently enrolled in the Obamacare exchanges who have a pre-existing condition.

News outlets have reported a number of other salient factors.

• The New York Times notes that the cuts to Medicaid spending may not only affect health coverage for low-income Americans, but also funding for special education programs in public schools. An alliance of school administrators and school psychologists has publicly opposed the AHCA.

• The Wall Street Journal reports that the effects of the AHCA’s changes wouldn’t be limited to the population that has coverage through an Obamacare insurance exchange or who are on Medicaid. Even those with coverage from their insurers could be affected, the Journal’s Stephani Armour and Michelle Hackman write.

Under the House bill, large employers could choose the benefit requirements from any state — including those that are allowed to lower their benchmarks under a waiver, health analysts said. By choosing a waiver state, employers looking to lower their costs could impose lifetime limits and eliminate the out-of-pocket cost cap from their plans under the GOP legislation.

In other words, employers will be able to pick criteria from states that reduce their costs — and reduce coverage and introduce lifetime spending caps for their employees. According to Vox, 91 million Americans had plans with lifetime spending limits before the passage of Obamacare, which removed them.

• When the MacArthur amendment was first unveiled, Vox’s Sarah Kliff found that Congress — which has coverage through the Obamacare exchanges by law — would be exempted from the effects of the amendment. That loophole still exists, though MacArthur has pledged to close it in future legislation.

Paul Markovich, chief executive of Blue Shield of California, told California Healthline that the AHCA “could return us to a time when people who were born with a birth defect or who became sick could not purchase or afford insurance.” Markovich called it a “moral imperative” to guarantee coverage. Certainly insurers would like more people to enroll in insurance plans, but it’s unusual for an insurer to want to see more costly patients sign up.

It’s important to reiterate that the long-term effects of the AHCA are uncertain. While CBO estimates aren’t perfect (they famously overestimated the increase in the number of insured people under Obamacare, for example), they’re meant to provide solid guidance for members of Congress to inform their decision-making.

In the absence of a CBO score, members of Congress are left looking at a glass of water and trying to determine if it’s half-full or half-empty. In this case, it’s not clear how much water is in the glass at all.