This thesis is a first effort to develop a grey-box model of a macroeconomic system. This is in contrast to current black-box modelling approaches. These black-box modelling approaches result in models where the variables and parameters have no economic interpretation. Companies who develop macroeconomic scenario models such as Ortec Finance, have identified this as a major limitation. In this thesis, an economic-engineering approach is taken. Engineers use the laws of physics to develop grey-box models. Economic-engineering theory is based on the analogies between the dynamics of economic phenomena on the one hand and mechanical phenomena on the other hand. By applying these analogues, a structural method is developed that translates the National accounting and the circular flow theory of macroeconomics into the laws of physics and engineering concepts. The engineering approach leads to three main contribution: 1) A bond graph model for the United States economy. 2) A linear time-invariant state-space model derived from the bond graph model to perform time domain analysis and 3) The Laplace transformation of the model to perform valuation and analysis in the frequency domain. In the financial industry, the use of the frequency domain for evaluating and modelling economic systems is a relatively new development. This thesis demonstrates the potential of an economic-engineering approach to formally use both the time- and frequency domain for macroeconomic scenario analysis and modelling, making them intuitive to both engineers and economists.