Making abortion illegal doesn’t stop women from having them. You may have heard this argument before, often from someone wielding a symbolic coat hanger at a protest.

It’s surfacing again as the United States reckons with the very real possibility of the dismantling of Roe v. Wade. Supporters of this argument correctly note that women have been ending their pregnancies for centuries, and laws designed to stop abortions can drive women to get illegal ones.

It is time to stop making this argument.

As we debate the future of abortion access in the United States, reproductive health advocates need to hold policymakers accountable for the full range of consequences of restricting abortion. It is not the case that in the absence of legal abortion, all women who want an abortion get one anyway. There is no magical inevitability of abortion access.

Research in the United States and abroad shows that when women are denied legal abortions, many carry their unwanted pregnancies to term and give birth. In the University of California San Francisco study I lead of women who tried to get abortions but were turned away from facilities across the country (the Turnaway Study), more than two-thirds of women who were denied abortions because they were too late in pregnancy carried their unwanted pregnancies to term. Scientists in other countries where abortion is legal—Bangladesh, Nepal, South Africa, and Tunisia—have found that among women who are denied legal abortion, about half get an abortion somewhere else and half carry the unwanted pregnancy to term.

States that pass laws making it harder to obtain abortion care show a similar pattern. Abortion restrictions which cause the closure of clinics, bans on public insurance coverage of abortion care, and gestational limits all prevent women from getting abortions. When abortion regulations forced the closure of 19 of the 41 clinics in Texas, the abortion rate went down 13 percent the following year. Estimates of what happens when public insurance no longer covers abortion indicate that about a quarter of low-income women who would otherwise get an abortion if it was paid for instead give birth. And even before the imposition of most of the current state gestational limits, 4,000 women gave birth each year because they were turned away from clinics due to advanced gestation.

Nobody knows how many women never make it to the clinic because they can’t raise the money, can’t get a ride, don’t know where to go, or are deterred by protesters. Making abortion even harder to get will reduce the chance that women get abortions.

So why do advocates keep saying that women are able to get their abortions even if abortion is restricted or illegal? The evidence comes from comparisons of country-level estimates of abortion rates in countries with legal abortion and countries where it is illegal under all circumstances. The average number of abortions is roughly the same in these two sets of countries—34 and 37 per 1,000 women, respectively.

But what is not the same is the demand for abortion. The unintended pregnancy rate is significantly higher in countries where abortion is illegal—probably because contraceptives are also difficult to access. That only 48 percent of unintended pregnancies are aborted in countries where abortion is illegal compared to 69 percent where it is legal indicates that many women have to carry unwanted pregnancies to term.

There are serious consequences to women who are denied a legal abortion. The Turnaway Study has demonstrated that women who are denied wanted abortions are worse off than women who receive them. Compared to women who receive a wanted abortion, women who are denied experience more immediate anxiety; more serious health complications from pregnancy such as hemorrhage, eclampsia, and death; higher likelihood of continued violence from the man involved in the pregnancy; lower full-time employment; and, despite increased use of public assistance, greater poverty. Being denied a wanted abortion also reduces the chance that women achieve aspirational life goals in the next year such as getting a better job and finishing school. Nearly every aspect of her life is compromised.

And it is not just women who are affected. Denying a woman a wanted abortion makes it more likely she will have a child that she lacks the financial and emotional resources to raise. In the Turnaway Study, we compared the outcomes of children born because their mother was denied an abortion to children born later to women who received an abortion. Children born later to women who were able to get an abortion are more likely to live in households where there is enough money to pay for basic living expenses such as food, housing, and transportation than children born because abortion was denied. Abortion denial also affects maternal bonding. Women are much more likely to report feeling trapped as a mother, resenting their baby, or wishing for the old days when they had no baby after abortion denial than with the next child born after receiving an abortion.

The biggest problem with the argument that abortion criminalization doesn’t stop abortion, apart from it being an oversimplification of how difficult it can be to access abortion care, is that it implies that no matter what the government does, women will find a way. Policymakers should be on the hook for the full consequences of making abortion more difficult to get, consequences which may reverberate for generations. Some women may find a way; young women, poor women, and other disadvantaged groups will disproportionately carry unwanted pregnancies to term.

Reckoning with abortion in America after Roe v. Wade means understanding the real physical health risk, economic hardship, and destabilization of families that comes from depriving women of the fundamental rights of bodily autonomy and freedom to determine their futures.