Boycotting Israel is a peaceful and effective way to get Palestinians their rights, the same as the boycott of apartheid South Africa

The fuller answer

Start with the case you are making

Two states for two peoples is the settlement that answers both national claims in the land, and it is the position to argue from. An end to the occupation, full equality for Arab citizens of Israel, compensation for property lost in 1948, a Palestinian state: every one of those is available inside it, and people on this side have argued for them for decades. One demand is not available inside it, and that demand is the whole of the disagreement.

Say what is true before you get there. Boycott is nonviolent, and it has a long history as the instrument of people with no army and no vote. A boycott of goods produced in settlements is also a distinguishable position, held by people who want two states and who buy Israeli goods otherwise. Americans for Peace Now adopted it in 2011, and its president said at the same time that the organisation “continues to oppose boycotts and other forms of BDS … against Israel and the Israeli people as a whole”. Conflating that with the full programme is expensive. The moment you do it, everybody in the room holding the middle position stops listening to you.

What does the movement actually ask for?

The 2005 Palestinian Civil Society Call for BDS sets three conditions, and says the boycott continues until Israel complies with international law by:

“1. Ending its occupation and colonization of all Arab lands and dismantling the Wall

2. Recognizing the fundamental rights of the Arab-Palestinian citizens of Israel to full equality; and

3. Respecting, protecting and promoting the rights of Palestinian refugees to return to their homes and properties as stipulated in UN resolution 194.“

Read them out loud in that order. The call names its endorsers as representing “the three integral parts of the people of Palestine: Palestinian refugees, Palestinians under occupation and Palestinian citizens of Israel”. There is one demand for each, so the third is no footnote to the first, and the constituency behind it is the largest of the three.

Why the third demand is the argument

The first demand is about 1967. The third is about 1948. Those are not the same dispute, and only the first concerns territory that a two state settlement would transfer.

Paragraph 11 of resolution 194, adopted on 11 December 1948, resolves “that the refugees wishing to return to their homes and live at peace with their neighbours should be permitted to do so at the earliest practicable date”. Those homes are inside Israel as it stood before 1967.

Say Article 1D yourself, before anybody says it to you. The 1951 Refugee Convention does not apply to Palestinians receiving UNRWA assistance. It says so expressly: the Convention “shall not apply to persons who are at present receiving from organs or agencies of the United Nations other than the United Nations High Commissioner for Refugees protection or assistance”. So anyone who sets the Convention’s cessation rule against UNRWA’s registration rule is making a category error, and will be corrected in one move, and everything true they said before it goes down with it.

Now use it, because Article 1D also says what the arrangement was for. The exclusion lasts until that assistance ceases “without the position of such persons being definitively settled in accordance with the relevant resolutions adopted by the General Assembly”. The drafters wrote a holding arrangement, pending a settlement they expected. Seventy-five years and 5.9 million registrations later, the holding arrangement is what there is instead of one.

Then the narrow point about the text, which is the one nobody can take off you. UNRWA’s own instructions define Palestine refugees as “persons whose normal place of residence was Palestine during the period 1 June 1946 to 15 May 1948, and who lost both home and means of livelihood as a result of the 1948 conflict”, and add that “the descendants of Palestine refugee males, including adopted children, are also eligible for registration”. Neither birth elsewhere nor citizenship of another country ends it. Over two million registered refugees live in Jordan, where a majority hold Jordanian citizenship. No other refugee definition has descent written into it.

Einat Wilf puts that as uniqueness: “For Palestinians, uniquely, refugeeness is an hereditary trait.” Do not take the word unique into a room with you. UNRWA answers that derivative status for children exists in other protracted situations too, which is an argument about practice, and once you are having it you have stopped talking about the text.

The Congressional Research Service records 5.9 million registered Palestine refugees as of 2024. Israel’s Central Bureau of Statistics reported 10.178 million people at the end of 2025, of whom 7.771 million were counted as Jews and others.

Nobody knows how many of the 5.9 million would move if the right were granted. Anybody who hands you a figure is guessing, so avoid repeating one. What you can state is the arithmetic. There is no version of that transfer, at scale, which leaves a Jewish majority in place.

Expect to be told this is a partisan framing, and answer by naming somebody the other person respects. Norman Finkelstein, no friend of Israeli governments, walked through the same three demands in 2012 and landed in the same place: “if we end the occupation and bring back six million Palestinians and we have equal rights for Arabs and Jews, there’s no Israel.” That quotation reaches us through Ali Abunimah, who supports the boycott and was writing to disagree with him.

Quote the movement’s own answer before you respond to it. It “does not advocate for a particular solution to the conflict and does not call for either a ‘one state solution’ or a ‘two state solution’”. That is accurate, and it does not settle the point, because a demand can decide an outcome without naming it. Accept all three in full and you have described one state between the river and the sea in which Jews are a minority.

This is where the argument backfires if you push it too far. Delivered as an accusation, that these people want Israel destroyed, it fails, and it deserves to. Delivered as a question it holds: which of the three demands are you supporting, and what do you think the third one produces?

Does the South Africa comparison hold?

Better than most people answering it are willing to admit, which is why it needs meeting rather than waving away.

The consumer boycott of South African goods began with the Congress movement in 1959. Chief Albert Luthuli’s appeal founded the British Anti-Apartheid Movement. It argued that “economic boycott is one way in which the world at large can bring home to the South African authorities that they must mend their ways or suffer for them”. Boycott therefore comes with a famous success attached, and refusing it in principle is not an option available to you.

The disanalogy is not about the tactic. It is about what the tactic was for.

The Freedom Charter, adopted at Kliptown on 26 June 1955, opens by declaring “that South Africa belongs to all who live in it, black and white”. The demand was equal citizenship inside one country, made by people already living in it who claimed no separate national state. When the demand was met, one state carried on within the same borders and everybody in it kept their citizenship. Nobody’s national existence was the thing at stake.

The Israeli-Palestinian dispute is between two national movements, each claiming self-determination in the same land. Applying the South African model to it means treating that as a civil rights question inside a single state. It answers the national claim of one of the two by dissolving it.

Most people who reach for the comparison intend exactly that, and they are entitled to argue for it in public. What they are not entitled to is the assumption that the parallel is only about a tactic.

What can a boycott actually do?

Be careful here, because this is where an advocate gets caught. The temptation is to answer the effectiveness claim with an effectiveness claim of your own, and the evidence will not carry it. There is no agreed figure for what the boycott has cost the Israeli economy, estimates in both directions come from interested parties, and a well briefed opponent will have a study that says the opposite of yours. Rest the case on the demands, where it is strong, and treat this section as a caution rather than a weapon.

One documented case is worth carrying, for one narrow purpose. SodaStream’s plant at Mishor Adumim in the West Bank employed around 500 West Bank Palestinians alongside Israeli Jewish and Israeli Arab workers. After a boycott campaign the company moved production inside Israel. In February 2016 the last 75 Palestinian employees lost their jobs when their permits were not renewed.

The movement’s reply is not evasive and you will hear it. It holds that Palestinians “are forced to work inside settlements in sub-standard conditions because of Israel’s deliberate destruction of the Palestinian economy”, and that what is wanted is “decent and dignified jobs within the Palestinian economy”.

One factory is an anecdote and not a study, so say so before anybody says it to you. Lead with it and you invite the reply that you are defending settlement labour, which is a losing position and a fair hit. Use it for the single question it carries well: who gets to decide whether those jobs go.

What about universities and culture?

The academic boycott guidelines published by PACBI hold that “all Israeli academic institutions, unless proven otherwise, are subject to boycott”. They also state that “mere affiliation of Israeli scholars to an Israeli academic institution is therefore not grounds for applying the boycott”, and that an individual’s academic freedom “should be fully and consistently respected”.

Two things in the same document are checkable against each other. The guidelines exempt Palestinian-Israeli projects only where the Israeli party accepts all three demands and the work amounts to “co-resistance” rather than co-existence. Dialogue between the two sides is excluded by design, and that is the sentence to put in front of a committee that believes it is voting for solidarity.

The most persuasive voices on this are people who share the room’s politics about the occupation. Michael Yudkin, the Oxford biochemist who has argued the case against the academic boycott at length, opens by saying he opposes “both the policies of Benjamin Netanyahu’s government towards the Palestinians and the occupation of their territories”, and then rests everything on the principle of universality: that scholars are judged by what they do rather than by who they are, and that boycotting an academic on grounds of citizenship breaks it. He also makes the practical point, which is harder to answer than the principled one. An Israeli scholar denied a referee for a promotion case is unlikely to be comforted by the news that the refusal was aimed at an institution.

Alan Johnson, arguing the same side in a published exchange in 2016, puts the political version of it: “Boycotting a civil society to express your opposition to a government is unjust too”, and the boycott “creates a sense of siege there, reinforcing reactionary trends, damaging the peace camp”.

Be accurate about the other side of this or it will cost you. In August 2024 the American Association of University Professors dropped two decades of blanket opposition to academic boycotts, stating that they “are not in themselves violations of academic freedom”. Cary Nelson, who led that body from 2006 to 2012, called the change shocking and said its policy could no longer be used “as the gold standard for academic freedom”. Quote Nelson rather than pretending the reversal did not happen. Tell a meeting that academic bodies uniformly reject boycotts and you will be corrected, and you will lose everything you said before it.

Your three moves

In this order, and in your own words rather than these.

“Which of the three demands are you supporting? I’d like to take them one at a time.”

“The third one is about 1948, not 1967, and refugee status there passes down the generations. Norman Finkelstein added all three demands up and said: there’s no Israel.”

“I’m not asking you to defend Israel. I’m asking whether you’re arguing for one state or two, because the demands answer that even if the movement won’t.”

If the person says they support a boycott of settlement goods only, agree with the distinction and stop there. That position is not the claim on this page, and arguing with it hands away the credibility you need for the part that is.