How big will deficit be? Depends on the economic forecast you pick

''Assumptions'' is a key word in trying to understand why Congress and the White House are so far apart in estimating budget deficits.

The latest forecast by the Congressional Budget Office (CBO) calls for annual deficits in the $140 billion to $160 billion range over each of the next three years, compared with a new White House projection of a $115 billion shortfall in fiscal 1983.

That's a whopping difference, with major implications for interest rates. If the CBO is right, the Treasury will gobble up a great deal of money to pay government debts, putting upward pressure on interest rates.

The difference widens in what are called the ''out years,'' with the White House forecasting a $95 billion deficit in fiscal 1984 and $74 billion one in 1985.

Both the White House and the CBO command distinguished economists, who work with basically the same raw materials - rates of economic growth, interest rates , inflation, unemployment, and the like.

Why, then, do their computers churn out such different results? This brings us back to assumptions, which can be affected by the biases experts bring to their interpretations of how the economy is likely to perform.

''Ideology, not economics, has dominated policymaking in the Reagan White House,'' says Barry P. Bosworth, a Brookings Institution senior fellow and former presidential adviser.

That ideology - an expectation that huge personal and corporate tax cuts would spur savings, investment, and robust economic growth - consistently colored Reagan economic forecasts with a rosy glow.

Anticipation that the economy would come ''roaring back,'' as Treasury chief Donald T. Regan once put it, gave rise to a variety of assumptions. Tax revenues would grow, thereby shrinking budget deficits, unemployment would fall, and so would interest rates.

The CBO team, headed by director Alice M. Rivlin, took a more sober view. It cranked into its forecasts some assumptions of its own, namely:

* Government outlays under the Reagan program would rise faster than income. Spending would remain constant at 23 percent of gross national product (GNP), while tax revenues would shrink from 21 percent of GNP now to 18 percent by 1987 . As a result budget deficits would grow from 2 percent of GNP in fiscal 1981 to 5 percent by 1987.

* To finance a deficit of even 2 percent of GNP - that is, to make sure that the US savings pool was large enough to accommodate both government and private borrowing - the nation's savings rate would have to climb by 40 percent. Herbert Stein, chief economic adviser to President Nixon, and Federal Reserve Board chairman Paul A. Volcker, among others, argue that such a savings growth is highly unlikely.

* Finally, even if the economy did respond as White House supply-siders predicted, such rapid growth would be inflationary. White House projections, says William Fellner, an adviser to Republican presidents now with the American Enterprise Institute, call for a 10 percent nominal growth rate of the economy. This, says Mr. Fellner, ''always was inflationary'' in the past.

Not surprisingly, CBO forecasts irritate Reagan administration officials, who call them too pessimistic. They complain that US financial markets respond to the CBO view, not to White House optimism.

True enough, US bankers and other lenders - wary of huge US Treasury borrowing needs - have been slow to lower interest rates. Fed chairman Volcker notes that the Treasury may have to borrow $100 billion over the next six months to keep the government going.

The CBO estimates, adds Volcker, ''are in the general area of what I think is probable and in the general area of what the markets think is probable.''

Stung by experience, the White House has grown more restrained in its economic projections - but not restrained enough to satisfy the nation's investment community. The financial market, says Republican economist Alan Greenspan, operates on the assumption that the Fed will have to monetize the debt - i.e., expand the money supply - and that inflation will average 9 percent over the coming decade.

Meanwhile, according to informed sources, Murray Weidenbaum resigned as President Reagan's chief economic adviser partly because he could not agree with economic forecasts - including the size of budget deficits - now eminating from the White House.